sg16-avd

Download

sg16-avd@lists.packetizer.com

August 2000

49 participants
146 discussions

Re: [H.323 Mobility:] APC-1902 - contribution from Intel
by Reddy, Paul K 22 Aug '00

22 Aug '00

Hi Jaakko, Please review Intel's APC-1902 contribution which is at Pictel site and provide your comments. For you information I am enclosed the APC-1902.zip document. thanks, Paul Paul K. Reddy Roaming Services Capability Manager Mobile Data Services Area of Opportunity Applications and Services Lab Intel's Architecture Labs Intel Corporation Office Phone# +1 (503)-264-9896 Mobile Phone# +1 (503)-807-9564 Email: paul.k.reddy(a)intel.com -----Original Message----- From: Jaakko Sundquist [mailto:jaakko.sundquist@NOKIA.COM] Sent: Friday, August 18, 2000 5:53 AM To: ITU-SG16(a)MAILBAG.INTEL.COM Subject: [H.323 Mobility:] Comments on APC-1932 Hi Radhika & the rest of you mob people, Since I'm not coming to Portland, I'll send you my comments on APC-1932 here. You can discuss these issues further with Mr. Rissanen in Portland. ** Comments on chapter 2 of APC-1932 ** The LRQ mechanism will, of course still work, but I think that for a true global mobile system, the mechanism is VERY inefficient and thus a better way of locating the called user must be standardized. I.e. generally a GK has no way of knowing under which GK (or set of GKs) the called user is located and thus in order for the LRQ mechanism to work, the LRQ must be sent to all GKs in a domain. The domains may include many GKs, thus the LRQs would create very much traffic also adding one GK to the domain would necessitate all the other GKs to be informed in order for the LRQ mechanism to work. Furthermore the originating GK does not even know whether the user is located in the same domain, thus the GK would have to wait an arbitrarily long time for an LCF and if the LCF does not arrive, inter-domain communications would be needed -> how long should this time be? I really do not see, how the VLF <-> HLF communications across domain boundaries BREAKS the H.323 model. The H.225 Annex G states that "all inter-domain communications are executed via border elements" (not an exact quotation), but we can either change that statement into "all inter-domain communications must be executed by using the H.225 Annex G protocol" and use the protocol between VLFs and HLFs or we can say that the VLF and the HLF must always be co-located with a BE (to create elements HLF/BE and VLF/BE). As for the question about the "distributed" vs. "hierarchical" model. Again, the VLF, HLF and AuF are just FUNCTIONALITIES, i.e. you can co-locate them with each other or with a GK or a BE. Thus, if you want your GK to be able to directly communicate with a HLF, you can add the VLF functionality to that network element thus creating a GK/VLF (or perhaps a GK/VLF/BE), which can communicate with the HLF. Thus the current H.323 Annex H architectural model does not stop you from utilizing either the "distributed" or the "hierarchical" model. These comments apply on some parts of the chapter 3 also. ** Section 2.1 ** I do not understand this at all. We are producing an enhanchement to the H.323 protocols and in doing this, we have identified new functional elements that are not present in the H.323 standards beforehand. So, the interfaces may not have been a part of the H.323 standards before, but after Annex H is done they will be. As for the specific interfaces, I don't think we need to specify the HLF-HLF interface (in fact, the interface is not even present in the architecture diagram) and my opinion is that we can leave the AuF-HLF and especially the AuF-VLF interface unspecified in the first version of the annex (thus assuming that the AuF is co-located with the HLF). I have proposed to use the H.225 Annex G protocol in these interfaces, but there can, of course, be other options as well and I would really love to see contributions utilizing other mechanisms, such as the IETF's AAA, etc. My opinion is, however, that we must select and standardize the use of some specific protocol to really enable interworking between different manufacturer's products. These comments apply on some parts of the chapter 3 also. ** Section 2.2 ** I think we have a linguistic misunderstanding here. I would like, if some native English speaker would comment this, but I understand the terms in the following way: - Visited Domain, etc. should be read: "the domain currently being visited by the user". I think that this kind of interpretation is commonly used in technical language (or am I totally wrong?). - Visiting Domain would sound to me like: "the domain is visiting somewhere". - Serving Domain is OK, and is already used in Annex H. - Target Domain is OK. In the present draft I have used "new Domain" instead of "target Domain". Target and source are words that are commonly used for the new and old domain, etc. (E.g. in GSM handovers, the terms "source BTS" and "target BTS" are used. If using the words "new" and "old" is confusing or misleading, I ahve nothing against using target and source. These comments apply on some parts of the chapter 3 also. ** Chapter 3 ** Most of my comments on this chapter are already captured in the above comments for chapter 2. A few short notes on the other issues: - You're right, the security check for unregistration is definitely needed. - Also Connection Termination needs to be handled, although I imagine that there is nothing really special about it compared to the "normal" H.323 system (that is why it isn't included in the present draft), unless supplementary services (and perhaps some billing issues) are taken into account. Those are my comments, I will discuss the issues with Mr. Rissanen so that you can have a meaningful discussion about them in Portland. Have a nice meeting everyone! *** Finally, just to annoy the editor of H.323 Annex K: Ice-hockey isn't enough for us Finns anymore, now we are beating you in football (socker) also... Finland - Norway 3-1!!!! ------------------------------------------------ Jaakko Sundquist * +358 50 3598281 * Audere est Facere! jaakko.sundquist(a)nokia.com * ------------------------------------------------ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com begin 600 apc-1902.zip M4$L#!!0````(``!(a)%2G31+\I.;$````>!0`,````05!#+3$Y,#(N9&]C[%P) M?%1%FJ_N=)).PAL.0^2F2$`)=-HDW!%'0(a)Z20*!).CHP'G2Z7](/^HC]7B=$ M'0<</&>=\=99&74]%M9CEH&%T14U>/(3#UQ=1A=!\,)!CJS#C"`#F7_5>Z_[ M=:<;(L*@(Q7^J5?UJK[ZZKOJJT[(YK?Z[OBWWP_:2>+*CTD*.=J50=(,?>SY M9K/6Z$/(O^#9A,>C75U=K&L9VM<!UP,W`#<"-P%=9\IWMNQY=`.Y@F18"+'V M>SZB613T2.<2\B/2N+!QH3)6&4NZE0Q+#AE?0TA.J8DC-)R0>V$0[KAQA]#' MS*:KJW>D+]FS7B[EW\L'DDC-D(GGV:A[H;8/5=\9ZVP#A<N&)J\GH[Y;&_\8 MZDL,[WN76X@/[.8.4]NLSD$]6:MG:_UZ_=HXC$&]9YS*UT'45M0?E%KXKK>C M)G&U`[5]NH6\B]I<IM9;4(]$?56EA1R`Q()5%O(YVE^COS_I7O1]D_$)7J)< M9B?$FTK(*^.X,J/CM)K1967S!+7=.C"VUO>G%];NTR<J%WU>?,WH?T!4.HHY M2F?[$%7N>OL#31[Q9?I09G6$.+1Z_D"U]FIU/)_Z?G1ZC(^K>T7YF36=D-7E MB%1?IY"]>-X]F)`#Z,\!_0'DQ(N^GKZ?--38(J$O+C`]E-5ABK>[#\:KZ^Z; M:B'[$F_]-)121UE!T93"8L&47NULH$[1)[J#?G\X(+E=BA0,T'K%%?"X0AY) MUMJB6PF&*!7JE;"GG<X(!<,MM&BBC<X-BS(?433NO*+Q@C#77E1<4#3^O**) MM,[5TA(,*6(X)--:452D0+.-.M#C`VT;G1,2FX,!&VVH+[71XJ*"X@FT--P< MEA5:7%A8F)Y>'G2'_6)`*1&<DN(32P2A?DY#75E%27J$^RK[N.)QM#;8*/DD MI1U,AEHEMRC3L(RU:'6%L_)V67TMTFH'Q;(T(+91<;$B!CRBA[:$@DK0'?3) M0G5`$7WF]'2A='J]LZZTS%E",YQ>2:;N8$`)28UAOD>,;PG*6$`)4I?'0^/6 MCY"CC2X9Y#'C&_!@%X2,C`Q!Z$4=?!F7C[J]KD"SNIRZ5&D@("ZF5=03E8U7 MI$U!GR_8QK8L>UT^'VT4&7>@C7F3[<7V(KXJ>RJF(a)N`40WXI`.IE7E?(Y5;$ MD`0-NN42ZE+4K8K4(\IN/#*2"A:05<%2%]^H),HVJHAN+[,67[23!IOX:+^Z M645;QTXK83GB8I>_Q2?:^`C]%=N@D05(.^SS4"G@878HTC:O2Z%*>XM*6^," MB\\-UE-1<=NIRT#*%:!RF%N<(`@1$6J:8B+WB[+L@CAA2>7A$-L;+!`+A[B) MJYS-F(FAHH>+/`2?:`Y(5XJQ/+MD6L5T$>G`[ES=-UW=%#L-MM04#`>X4J`@ M1L*&>8%@R(^W1DZ82;A%3S@D1O4I!5J#BT2/G58%V\16,42E./)>.)L8D#7J MW?CIOCN5(KUH5J6-5LVJY"92VE!)855<FS#'Z@#VYL%B&.YA&@HTAR79&[MP M4RCHUP02U%95%XM:I5'XU(5M:7[D86MT%Y/&^NA:9[Z-AEL@$/"-]=U!;+S= M1B4%IA#`+NHJ9E0S;W56SYE-:ROJZTMG5#`NM*TJ7D2I9B\M==BIT,#HQ$B9 M;=@5QEALV&TP@1BJ=16.6?,84R%1"8<"JD^Q4;5.N^I[,%]TP$YKG6P<XXSM M.2%O4D!61)>'&3.76%W=7.KVB:Z0KSUB]'+,#IC])Q8/8[^;\IAV-1-@^O6U M<AD+=<9].W3KDN%?8H#QK0M7QJ(E"!%86B,/C_0C#(C,*1'/H#ION\S=GMM) MN\"%`!*J1EQ)=0*2H\>"8'Z,6MADS3)XI'5)S(0Y143-%NS:%<!"T?U7E]MH M=6U%-1W=YI7<7BU@<(MGP=7CP;9D9C8R\Q&/RB2;4H\I<KB1AS6\JB['(#VL MA<0KV"E&6Q",_*+"Y,"VS_C0*.JAC<D:YM<H^H*!9J8I=0O87%BU+[<HM293 MOPWO?:YV=8/ZKC5E0S(J*2YSK^A>Q,)%B+]K`0=B`'QJ3.ALPXN:F"5HW7S+ MHZ,G#[:<'PE#[.7M<K>I,"=7(X@K-IT+?7')0)2-8KZLVE8HZ/(S:U#%`[5H MIM6*8R/HAP;YHH$@MUQUF'ILS8P-`EI(X]L,-C+5Z\<-7Y2Y9C`D7<F-EI., MZ8'C23Y993L@-JOK)W9=F5N0*N=N87H.$ZQ.641:PF3`8RG,65*@^;9`Q`HT M((a)EMG%D,MM*&I$E./DJC4*6K6Q'].+%<H7:=,UE1MU==3NEH)].A*GXYJ'&E MAR`6"51QJI386*,[N"'C`IA&U"W4@,5,6+,8[F3>H%]7G&:_XC'XY]%>-Q6- M3*LD2^QLB*I?X<<]PH9*B7DV=*!Z5:O+)WEXG$/TP=L0<RBVQ0!G33=ZQ%=& MAV\\:I+:4-HDA>"O,>%<=V+,#XF)WMCI=-TWF/1L/&""FFZ'HS&SL;W%)6-0 MOBI3QI`>V$H;G%45LYW59;IUS6VHJ'?J"@4M!-H&9KA(G96P+*1'\BRU([W4 MS;EI;-<M(#V=CY\]QVD\AT1/=&;W5_6B%IBD0!,+/I'5N>`:@XJ71F8S:?,5 M2F-.N63+TNBZI=]@S0;=6:,K0,2^=FV9D[2S:/;U35>*[9[.%1PG$5B9$FSF MQH@;`/<2'H+@@Z-Y&+49#R#65P&W9#%,<_AH6L#XU=S"Z%AL"W&<MX1#ZEW" MY78CR^*Q3Z6!(XIEQ"Z$9AS*JB4RNIH;L"$LT?5)_%1N8@DDI]@F0?LL9,&] M7.P<L.O9E<'264H0B=YQ'+'$0-9B*C/]-A>CIGMD-P=@X342#Y@#Z(=-2&QA M"8T<"<Q1DIHWM;`,$%M4CAV\HX>CC:KG/6-8/T:E2%1G*8P8:(^[KD2V'^$' M`I?X6D:V;2I1QEVX18U,[(TOJ)N8\20,<`''':.VF(F-$L\`93HZFK]4.[2T M15=Q)#C;U,#-;C/YW&-E[>2AH[%8`IEH03^?=DL=])1"C9G&;`K9GG9H0I<1 M0<;&3XB'AV:JI9S,'<%A=#07(a)#_,XVZ!.E4,,1G;HM?=R#B>/^D^(_,`:(N8 M8ANN4"P_T"-O1/ZRY&>:2BA-1/^`R.R3'9910;$QQ[2<TRD;_=:!4[5%83,2 M<!HQQ:I9E=$L;/38&3/S$0E$+@B7(4\X5O87%4MRPS'X#+]>G+`PV,5"J-6( M5B(]9!=K:VQ!AZ#!RJO88KAGR"++OI7X"X;$K^M.;T_O&#_8R\6Q[Q:1%%"G MR$XE52`LGD6<3)N!H2VBFT=*-EF+>I%CY>1>5&($9SMS;_F>W5MBU/?#NK(H MH3!.J^/<6O0IJE\EN+!(J@,EN-($]5G&WC/7F#/7F!Y=8X28FTR"B\P_^AYS M*F\CWZGKB/[);5,XQ)3QO;ROG/3KRIG;RC_O;>54WE-T9_HV]Q1.`UK3-&9( ME1.LW^AR+XHZ:M)+CB!$'JSJ[4;]%JUB;CQF6L8VX%8$AROLHS/MM$[T>-CM M@?WD&2]#3&Z,<1NMJ1Q7,&[2)/93\:(B.MM>8:?%$W`DEK:*@3"\MTKR^>3& M8"AHHW/JZ-UTRJ2BXO'\!^DTRRGZ2NC8(CIZ0N&X?%H\<7S!E,E3)MIHI6MQ M7/_DH@F85(&XCAEGT:IYCHJZ6=6S^=T`!TE)"_BT+[*'&)O3),:EW0WEF[(3 MO^@OS)86>24?+1=E;PLT+";96W%!,?O%@6^^M]BM39E46#!YPOCC;RW`V;+[ M[1Z=L2C3@G!Y@B((!1"(@R6/])(Q2"3K9E14SJFK+772[.+^M$`08G\:'_D% M@-J@1_2I::7Z\WGV\WRANXWUH$_+7I/U)CZ?P9B:X\;TS9@Y-GGJ*P@7Z1W, MV...5<-ZW3_\P]1CD(7/]7#%GN_DFT@HF=QBTX;$P[7M)<PQ>JY/5H[_JSAG MRIGRCR_L-S\MA/\&:":0!4P"2H!YP'Q@&?`0\`BP#E@//`L\![P"O`KDI!)R M-C``&`B,`NX"[@;N`WX+O`8<`(a)X#?P.RTPCI#^P$NH!AF80,!V8"7\%A+NY- MR/5`=0XA/P'F`5Y@(;`("`!!(`0H0!AH!:X%_A-X;!!H#L%ZAPZ0SGV[R6>? MD&U_?.?-PX<V=?SW.K(:7T^L>&+%`[_!-_($6<$:OXI\H:-_RD(FH3DU9C(7 M:+":_%7$DC71NE3O^M$$ZU)#DU@C3UD3HMWI7BL)5&58TM%75M./9*(^GST. MP/[5-0IK<HWK6,IJP'>&2GWZ&%.:^L1G3F0S\5:(O&7-%+08?3_625.[20J^ MI6OO,]1U0#MEHCJRK(:R7\C4=)\5I_.'-7W_#EBKZ?TY3==]-3U_:;2C/QD; M.XR-_^U98^V[IFACI?%-\L9]QL:M21M7D?B23<E9CV23;"!GI$EX98!Y^MDK M1J4-2EEA&9_F2-^0WC&BHZ.#=)BS"`8ZH@-3B:ECA)G]HF_*2',,B;@WW>90 MPEIGCS29[_81ZY9"]LO!SK2Z<Q@[:285^G.B/LY)M_4R6._4;CV.2,^\H22. M6^$5*S$_F-HQ8AH,^-OHG_GY(,W'[]'\?+GFY[J/?PB_WF'P<9*.V&`EY!S@ M)N!FX$N@$FSO!?8!^S.BL:#&$`\NZJW&!#?@`10M/MP`W`\\#1P!C@(Y?0@I M`#R`""P&[@#>!-X"/(a)(.`8/Z$C(8&`(,!3X'=O7#FNQ7PX^P<IA_/_+5X;^0 M/Q/2R?6QFW_?M7,K(5O(.^3-C40?&0-"7EB_+FITF=85Y@P3B[*1YQ3#L]GP MS/5N-3<2WB2IUL=16<V#B?4QULZT/FJ8\6AD1JKZ;-U@BO:PYQ3S-))B'D32 M884IIM1,<XJ)6-A#Y:G0XS9B*#$^_HZQ\8:Q\:*Q\8RQ81!@'($3:9Q<:K]/ MVH@6""_S'))5^:A9?]I@@J"$5S(M:KC9`%%":KV7/9C:=]F"K[G7/I@ZC>!? M[V7P-H#]-QOKH-@YN<9XDGZG-S5!3#'6"?NR3Y;??4T,Y6#2QL=)WVQ/^N9] M8^/=I,/>/H'&O1\8A)-\F+%,-HH]=<L`"R@D@G5+)4FYLVM(UB,YMER2=F-6 M-TH][$OA"^JVL\*<VAAYTOH>-_<?3%)'DK21:OLQA/K%,):S+'.L"S.DK&=[ M:3;#CC23T2))WY,=,YO[(UX8V7]QN4'$,6^2-UY<>0)S"D]@CK&Q(6ECO;&1 MW.63^']61'.]EVTV<^_>;$Z+ZI/_%[K39%4D6]>9%ZA$SCP#J,J)S;DE+>?V M`7XM[VX!KM!R;Y9WMP&+@7;@2N`JX&K@9\`UP,^!)<!2+3__!;`,N`ZX'K@! M^%U.-&\?.D3-W<<`8X'S@5I@-B`"2X"7@)>!;<`!-@ZIC@VX$)@/W#54_5]> M_P%T`#N!CX"/@4^`3X'/@%U`ZS"<?<-AWQ0Y.0NMP#Y@/S`V#W3S6(P[T$VT M7\8DP?O)[D_(3I0/X\%?O__.&Z_"M)[!F;V2G=/IK%-[3C,\IQJ>+8;G%,,S M/_LCY[C>&\T;5JIYPY-F=OH/TO,*;945^BHUIT/WAXTB^ZNQ$2/??[9&C*7$ M-$[AHKR8!A%CW$?,26'YAM=$AIP*7SQB7/M$&C'G>LR;/QL;^Y(.^^($&JN- MN4!/YRQ/-B>FG.:<864D9U@9R1F>5,^=[&&:'NV(;Q<`/P;*@!G`+&`V,`=P M`'5`/>`$&H"+@9\`\X#YP$^!2X#+@,N!!8`+:`3<0!/0#'B!*X`0(`-K@+5` MWY&$]`/.!Z8"%P#U@!/X-7`;D#N*D#Q@"E`"G`],!?S`$F`@<M]"H`BH!/X` MO`Z\`7P$?`P<!`X!?<[%&H`#F`MXSE7_S_.A`X<._"GI%_F(["!;R=M;W]S* M'[9NVDHV\J^7(WBYFY*.7RQE-7WX=YS(%O9!$[$4UE3C,F^V+DU9N/TL*-IJ M8F\;K";#ISE58TQ"]#,D]"3^U"KQ)U.D_%2<G<=(N+Y;B5T/&^N-C;4]:W0O MZ1%O9'E@U#=/;_Z7+-<IR(N-">5`!5"9I\:&VCPU-K"XP&("BP<L%K`XP&+` MY9KONS6?EX"%P**\'I__/YS,X%2?_ZP8SO^5AO/_5.O:!_B!`!`$6O*BL?]J M1(a)LQ/0-Q.Q,8#E!#7+\0F`8L`GR&&+\4N!.X"W@,>!SH`#8`[P'O`WN`O4`6 MXGPO8,`YW<^&&4`CX`9^!EP#W`S\$G@(>!A8=YPSI"_.BW[`*.`<8&J",Z4) M\&IGR['+D8/_OU\M>_;O)_MW\R]RS"]]Q/XHE11S,TDU4R*8S"83$<QYQ/J" M_H&(a)56]A%,3_TQ/5T;?1S;?.#Y,WDH>6F$8/J<5X4LR;F/3]1!CM/*EODC>( MJ1F@P`B2FJ?Z_PLF8LI3_W3/D-/E*Z?0!LXTXAHQ3J$6;A6HAQXK;HE:[&HV MQ*_DA$_+ULZ4$RNZ_K.*26:A*;MP3+JC-[$"PU?]PDY7O5`Z8E7`D@ODW7IS MZDA@U"I"SEXUR3P`."[U,^5[4]B]LIQXM):N6O7&-3;UH=1W4QY(23C14'H3 M2^=4U*LMF9WL+U^9V&>6F824D:Q.]A>6+*6,WE)37URYGS>Q/S#695+_1E:? M3HM*8DDQ2>^<#5Y2+&OOW3;Y"_]OGGK[RFWK=I9FK^TB8\Y2.2KG0^\P+1@Q MN#,7??$C6^R$W'20D*\0(BCN(a)SN0_13^3R^RY)!`.OMAZM'%RY^8EZYL>7GU M`[5+'/TZIM\QIJ!\.B[8N4?J1EENW[S(VV_^^FF/7/K)9[D5-W>.?Y!>^Y#U MROR<\@K+;;D;*_K,'4;-N7AS;]?K5^]9_W[MXEUKY6=W?;IIRO-K#N]]\^@U M:S[<]_J^:XH<%KI:&KC$[-BXY&A_9_8O-SX][KR6M9_?./.V3;GRUFL*V[:^ MMNSNBIS-\T>X:<TFAV7)C4O<II4/3[UN[2/6YM]NO6-K8,>^E1_-O^CB[>T; M>O][S<8-^25_6Q!Z]86?'_WT5[<<O/"NW)^U7]I__U^/_K&H\WGEF?9M@U8- M7//DBWV7NX,[LJ\:?LL:07IJ;W6*+^TOZ]Z:NWI;U05[2E]^:_ZO=VQ?-[7W M?PGM3^U]Z>.]T^YO.K?SDO/_]8$UST[9M&Y7_KK[U]WS\=R<Q2N/W)2V^[VC MK[U]^=-I_W?A.X[[9N7YS;?\X='/\Z\:O&;/I*=>>NZ&QR]8?E[;KD-=]\UK MOG#-PE53%\P;=LFM]UQ;MKKDO3%?7'O/IW]G[[O#HDJ:O7LB0\Y1<A(4$1`0 M$)1@1)($14$1)0@"2A04`RR&ES4KB@F5(.:$HJZ"("IF45$1!3/F%0,F4&]U MGV&8M.SKWN_^\SU[>.K,H:O[U]755=W59TZ?V32Q/%-K\8JY3/,DTRD&_7]; M[Z_Q^TA5F==&Z<VFCP?%L>?$J3_>T*S\[)KTT3M1CQ+NJB^[NR;]SJW!RU<& MCS-;M/9WMQ!'W8^K9E>O.NB2WS,DV^Z@08/IU-]JWN^,/?ER]L3S]Q>__*:9 M1V_9^FCZPF&YT__(KV/^1\9[=?96<W\S6P^IA0>+AT_9^^FR^L4RD]-E]7,6 MZ`\R<70>E?`A_5C)XX16CGQMWJ"K?_Y9.#S64NZPN<],/\=UGM8'%QA&#CI0 M,=3QTX3*>.4O,^[^?MXGRNAJ\J@[YTHB^]Y>ONI>\-XD];=+/KRD]SQ_NR/_ M]1.G@UF]SR=GGO1Z,FGE*99QEMN?D:-\C"2V;SFQS>G&CL7%?G8[ALF:Y#;. M<@@[U:_L?O9FESF2<C7S'\QVT#O!&<3J]\#3BCE'OM'=^*R[N874X=:=ATX; M79>XO]8K<\3!=S%W%A0]43'-F+9QS+/]RUY721Y4>G9!)C-HP8.>UVDG\[ZE MC^PU:<2BZD;4X)9QX\%,*^6].V_MJSZ^<RXGKK5HX0>3N?UFG;OZ>J!9UAEO M6S6[^0>6F\\[JR%3O4E?PNRHI&-UT+XG*B9'.8KI+*E'LF;9)SUL;58/MOH\ MF1;`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`NS,U MXZ:]>7CTU.-U(>8+[T]9_G6&5GC:->66/D/*;_T<R@CJ<V>7X\T7]3<OE026 M;?RVZ[+^89V(RAO?4CW?[7KS`(J-E7.9J%UXM-"\)LBK3U/8Y:-_O+ADYQUV M1CGW_+>OYGMWM7UK'!W6_G+#MF_'3&LW^'<L-S]WH,4^Z//TT(?G[1/S)^O$ MAN\^/3G7>.*:0]7>UY5]+YA9H+DWE:?;F59;;N0,FF0_^,:LNN`QJ9SQ.28J MV4TJCRL2)[PIM(DR4+5A/&?7E[[=HA"_\*W=Q<Q%-^RFMF7O:$#L5^?O25X= M'I`Y)O9QW([<BN*5L6-&Z_5Z]E6I(.X&_>=E^S-VY8H_FNRKEL6-OQ?]^%W^ MGB71KSUF3EU\,6!W89O$X3V]=`_?3%RVY$Q\X\\UC2?C>CR:]"K]5$I(\<ZS M#^UO]K^]O>28:],E\Z:U83>"D^=M6[_IX^*O\];D>]M>C2LJO1QM>TV[>$6+ M1IM]Z8ZRKRO6;[BH,BW!SX2,0VH95QL]I-K*Y'SKS`9LLGEC.//LWFUQQ^<W M'QKPPFA6?N6ATN>+*P8?>1W?RW#HM*PS1[?>J]U4["R_9N?,S4L>G_%?>_M\ M3<>.@OR>Q?\I..A];'M426Q=OM[\H1OILNGTTQ^<%SV]O45JVL*D]3-&OD\U M.SVD(^[>;O?`.I5'O?VR3RZXWJ$8^,3.+^_9\_S$UN5)IWN/T7FZ>]>]U,"B M\MB;UY5/.NU;>^'$X`OO`X\M7#A"<4[H]SA#Q\"@O2_W?+38ZC)4_?5`I]@C MUWR.%]@W/CUTX4K)D#2;RX=B;*O32FWE/X_/B59O/NUTRX\1DJUU5^KC]B5% M]M^NOYP2,JKPC;OYA//F,#XT%=5?M*\;,U[V9>S%ZYOCK&,CWTY,GJ6S(;A6 M>5M[OR!;N<_)AV+.T5*>?9CY0R?[Y`O)@RO'5*L5R&Y.=LJL:7JXL?GSWL2, M<]?NKGSZNY+B-__[_=\^'Q!]3$4Z<N3B6T6']TQ_OK)<M:Y]8N314Y<?VI:^ M7-*+,?!)='B???-9)Q@%3Z;IY_5])<ENH57>9ZJ\>I@I>;7]X*'F#8_W[XZ9 M?]BKKY>59D.C1=65*(;INN\-%T-/#]QFU6:7J.6[J<UB^.TO(<7/JDUCHY;Z MZ[S]SX?G#G;6;_T'?-`V67\@]F[09-NE"UXVA%]8$?7ID<('TT_]/EC?67S& M<EO/<MO+^RMW+`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`KWK5<SRSGINSM.O3D!%=X_ MFA>U*F7S$,^=EUUB>RCWDI[OON=CI..XL*4C'X8TLX]_L=T].2LC],[B)\.\ M0WS:CM^I_OI^6+\-CLDM@]:M:Q]_8.83BW1E]:I>F6H%8RZ-N>3O7*)1JK%M M]NQ$`^>LQVWL&P'WH[WKR^DGBHX45:2/]7DTZEB6WDB5/:I+W&(&!Q1+R]Z^ M';,RXOKD@\L?M1W^,B_BQNSW,D,N;MXQ[U/3@2:SVO(M&\IUQY?9O<B-]ZA] MX_KMDL^-Y!=#QNV8VK&[8,R--]Y:4]\^="O7_;Q_6?*L`0OT#GPXMO?//;%I M>FL/JSN<';K`\$WJTO+28X=<JO+>C%Y:MB_1Z'B^7MF-HZMTSIRW'^O8./V/ M]Q&-N1LCBS(.1-X\5AC=$3/S_@]/O7$/=,J-YR;L:J4='C`IY+WDAG//7NM7 M3KIKW3@N-'A22EN"?Y^O:W17O*D/NE6GO\[#]T+P>J/SA]S7WXC:,716B<:F MU?=;%V0N&[5?[^(K[_WT>U6Y"FYY#R;LI,N_//XQ?"=M;J'#\V&I;4V-2H_, M1MS?6=.VQTGURD,EAT6OVC:U:Y[(+DT9GS31L#F:_>+N(673V*5:T;T>3@K0 MZ+^A1_6H#[/VW\_)?K,-5<87OC.Y?.IJOD1B]HZBL2?7%3!UK*X83Q[1+S## M]/FL&:9_WO[\<4#KRF:_IF\_K_E\>WO2\.?538&W:H>/RL_9$V&<%G`HSV)9 M@HUD>[O-R(LC=/?\L#+\>6R<R3XUQHB&8=K^^W^7\3]\[8+S^:*3BP.*'R/- MW[^KS';>&/]C]=NV"Z-VWC.1G_V^X7A5PXGR#O/`SU(F=LW^!LH+S5]9:61[ M&*BVIZ>5VK`'N#FEY;S+]3H3-^W5A7M_].A1T';IM$KDR`^+RURN'^I3L[;L M\*6/PR57C7.+O"SI[UPT^8JDS0/_Y6O&M-S:Y)U>GY0=?F'-2TV=#R7Z07>= MOKU?N<91M^;]CK09^T;7IR:;[7JE4?0J)F6\SO$`Q2'%>\?G[-TUJL>JZ!X% M%>-&?8ZV_./)@<^*JVZK%51HE5Q;DC7T\-D)DDMJDK26J7X(7'PVK/BLP13G M275AEA)-[[FRG=8?<:=QYM#:3,<INSPE*T_L-IEI/'_$Y<BC(_*?@!3&7JV: M?8]<VC-@0T>3B]:LPTGKQ^@D]]<LTEC:%,GY5+_X]K!K0\H\E!;7+=)\5C;O MM]\U%@:B[Z'![9/JM^4TKJFSMG8J7Y0T^$M0\O1[<U^_./FI0F;+K96#_Y.; M9G/6:!3[&@I;NO5'Y9]QIY2#+F_?['!HRY_7"YP\#0^WO?8Q"?)M*;`-]GV: M5?C62QY:F)%67+ZSX('+N4C?NO*&$Y_N#1PC[RM3VGN5Q[!EP7ON#YFCJI4Q M_Z#5Y@LU2;E9D8%R!]L2#N8\"J?994=OU[RT9;35BL8G&8K2RMM-(\Q,_5/S M=Y<'?/3I%U3K6^1O-=ZQ9=6BWJ53'>\79>8G?0J-O#RZ_*MNX1'/%8<K+^<= M2XU6UCT<YSFJW/;JT5M&WG=M,H^-OABW;/JF.7=M$_I/V>TTM/:$FV]5'L[; M%A\^X+O]\TTI/<ZFG\W3M6XT?70A;QRL3&[_4;%C=TA>_&\+SXT^\XR"737H M\/&3MKJUY;H+]\N^/CNZU=K^QJP4HX:<=SLGJ-Z:[+BO)?#S1//9M:L7S@Z6 MK*RK]`C6D9OL[5#HL/1`QKEIJYN^ZQ9VS"UL;,G0V-46?'2Q9*6+9]'%MF,S MTP\.7=U2$FJ<GG=QIMJ`P4\'2#>[?-TWN7!C[:VD@*O+%-)6T]<O!!LZELUX MJB6W)FSZ\HTW>U19'[[<I[T=F\6T-NU/?N5I!>_RAPUP7+AD=H;7Q61=EU/% M+3[E6?4U_9Q:LNH;-J<<BTQHNO/D[9.3[;MJ7*]NW-YS[,J4DH@7@RN*B^;4 MZ*1ZWU#UC5UZI7C!^J)7U18[9<O:O!W'?'CGW*M(\WKD4(FU0[1#3Y^ID@@> M^KARY=DALE]4ID[5KK]1<=_O\VRCIKR$A4X?5Z%9Y^.];@5Y75]GVFMCW[G. M97/[>(8D>20^/[\D0J5_7LP]?[7DGF,O5T08+5LMG[X^\]+(IZIG+FXTN;/1 M7-IA?-1[A?,/!MKDKS)9<;SWL$FZ6@O.EVTZ%*@]L#0F:='0<^:>>5/K1QC6 MUUQ_&1/?YG"[??".4MU`?]_"FT4*LJ..7TO*,KWX>77)XS4S'?:V#O[H\W+C MPI*,UYL>!DX(23><:=2VX$W!@0#=I#N;+LQH_*PJFU5A&QH_M(KV_<2UHC*] MQR\?_N?`FYD_XNO'O]*03R^."?&<,"MA\ZZ%P_9(EKII1Z</+56J-?>;<'=] M>L07L.1EX[1GC"YO+CAP8MC;GU,.=XS?D%OH,*?:S5U=PF"(U*0#6_S<ULI) MN:NR/!:R,TPUEJ^O7S#<VG[U6;V)$]P6U2L-4#$(<.MO\,=P16G+J.G+L^]T M>(0V:R[=W&ZTU+*B9NV,^Y\+VEV>-<VXZJB>8-(ZY\K292WQF1X%;C&+R_[, MN+@C9%GITRQ'\Y%1[G_F:X=:-IQQW-7O^IAE[GYOIUI.3WR?XN84F^CA7]%K M@U//OFV%*>>OWI[IT_P@YG9&V82>$2&^$U>5YGJ-#(A^:O`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`Y9HJE`22@21<!?!G)%,2@!I<#_F#<934/QD/N!$!*^:1@/"#&0 M*P7R./TBWK_'__I00E.0%**A$"2/R&,!0)X(N;[[2<>?B(U\0-E)I)OB$-7I MU-/E<HBW%Z$*/UBN@O"&-E\H-`IHZO`GS/CADIQ01'.E`7(HPMN0AD,'AD.' MXLZ,1OK(&EE`_=;`ZP,RL(QI;&,:^8IR!%VY^`'-U13J4>%6PGN8G=3&0':< M+#`+.E($<D7X+K0PN@V@#R?G*7#&]YWI2,J41J&[`CH-@<DQ7#&"DUCY^I&R M_4A9!I2E4V7/(5=3\`)7Q`39J=I%R]J2LK8(_VH$$\HR<%E74X;L60F:%H-6 MRJQ"D)M%$/#]=UT1!#N"@,\C4``@64.:N#S#D1VIA06U,*E:F/RUN$(O_K6< M]J06>X+`!@06A<#B1QB')`C".,BE*8+0GR#@<S*XJ`7(.AAY07H*Y$@!JXDD MV!*`S::PV8+8G&ZP'0`OBJ!&05HT#`YX4!#&Y@"V!(4MP8\=C"0)=C#"O_4@ MC.T(J"D$.P72)Q$T861)0.90R!Q^9&?DYOK^YU;D3'(-AMQ1W*$K!3#\X#H) M*)J<IX-V]-%0\",\A/$?ND@+T6ATA']"@H4"D2_R(&W&W\3@WT0Q1*I($?B& M`GQLCVRP2+S-#^=1008B>6Q('D7BJUA:9?`O0YJS0![L=WHP`LCG:#$4<UHZ ML&&#7W/]C1YD6&IR[A214E.,E/V)E,J$KR&&;T_X*H2O+H9O1_BJA*\FAF]+ M^&IPMH"VXC'$`N&OE4:`%B-`W^E<RXL![>*VJD,["MD(F8+F<&Y3A)^YX\]M M3_)I0+Z%+)Q/[B_RV9%\FI`OCHGU*T5&,$.1?-9$0BTXVZ,?OMC.<`V2(`^_ MG86!92'HB1Z\W5S*Q5M@`)%$6`;IOY"!ZC]M**--P_ED_B(?90LZD`]_+VN* MS%P+T4,:SB<-/H@EP#;M`Y89#W)$@CWBPQ(9$#Q+N)85P!/T$=P^/3B/1OI8 M!_31I"78DJ>1R3B)^&77?_IP%4%&,TV$=]CB;6X*O&F"[+$%1`DT@-H(#77H M`:X2/0%*&/#JIG"G\/W7B2N<&DD\*@D"!#P*#D'8%PT0PYA&5<VW6XX)LY,: M^($IZ,>`COU`B=>.!*[\^LB?^'$D&6,2(,S`8P`'X0=SAPQG(W>D`]+:T-V) MUD1+!Q(=8(a)]7149X-H,.IT&'_R2;?A"M"C@VT*<TY$ZW(1C\<ZL^KQ^H48*- MC(EGXA^X^2$?3O>G8X^0@5HZQZHPZ.%(Z*5(4@[[A@GQ=Y5B,_@O%DK1T"1Z M+.'PE^J2<Q!8C#+DQB5DS]HP*'7):=/ZTP,8"<Q<5@GH[A[GBZ2R=%\93]DH MN2SA7VJB].H`=671)]'QR"4M5%?7:"J)>F*MX+Z!UE&;X&V)E$FD;1(")8T0 MGG7-<&^R<2$KN@JTC8%M!UF14BOI5J1_</NQ?>-R:<0KY9`YWZB&1TD+*+&" M7DBW$%/"AFBH%Y0H92KF)/]`4.L/^:WT@W0S,;DIC^L-N6W9.'[`?7.*[B2B MY4YKQJVPX+4";POWXT8O/J05U^D^B+*GKOE-N&=U(#YB<1$Z-W'S;Q_%O:!$ MVIA%?\]MHR">$9'/$N&P2H7,"5CSB70ZHJQ*,+=(a)K_7%.PM5H"!"I*?!8QD. MI+_<0+X$8DU4#<K0,]#23E%Q57Y(H,\,&58$D[\DY778<JR[2B,<M^`2-@Q7 MPN,OP2^?,O0?7YVM(*D&C:MA9X)`1Y2_"WI;.(DKJ'&$?\YG0P_C\)"R`AK# MGV%&=.<&\_AT;H_$\%I,Z0]K7I$QB6%!="F:L[.%=#*W8$T8,&R()N3$Y.9O M'36GV)!V9#-LB.XH-.%>ZD^->P]H7'U;<,M0UM!5IM/WI:`O.WV$S&#8HQCV M)/<0E`BV$$[T@J,6K&5'_MA`1EG#M/H4MRTT5$C:PN2S>37P#&H0EN79*4U" MJXKR&+`@$G%3;9DLU!9Y-$"P1WFCA26W/RU);9-)3^)5B0+T,)4=MPC/KPQD MS[.Y@T0Z%J\N(U*'BZ"E<K5&:=J09D-*X$7I9+`/*V(_`P5;8V=ZBK(R!_`0 M&GI/?())9MUDHF$=&%_E<WPYBCD'/@O,@R5L7PZL8?"O+/[PA?8P!Z+.>3V* MS'33R&_HZ2*\`SR7I9B3]DU$E[FLJDX/46*ZDIJIJ!/7;`!6)9]3PE;,"?E* M175\M7<P2]AD!87;2_6&(=,)4?Z,9UD\!B3#'Y[%W8F-")2O5UJG2J0WIOR3 M:4QJ]^'.@]BR/!"9\I;@*8^R+'<F95DC(4<&FD$\,8+4H0S1<Z<=RIYETW$M MD(PFD'+^S`FD)]Q@5>$#_8Y'=VNP3TH?9CRII/K3$Z!=]SC*TIZR6?)'%=\J MFZI5\0[H;_+*$Q6$-\,DX-TU_M3\R/0GUB!L\<(CL`&,D%T:]1/4:#O,DZ>X M_8EG,]*?+#0*;+/+%H9!^7M@"\:"MM`/YE?>>A;;]U9&(<.2]$8\U!Y))*)B M,3GP+Q68!XEN+VHC/$(%03RS@HE'*$G@9I"1)(G4FX"F(NPY`WO1W'O!S"Q+ M(CX')!Q!XO%+'=;Z\CD!#(%5_3PXB!7V$!.=.R(\-HU$B(RQUMPQ5C3VC4$X M8O02B7V)1X.+4B/T=:8SP?,!=!QQ>U/1TS28CO=TX*Y6*3XD&L^I$.M]QWQ$ MK)<.MHJ]5@]0<%L4<Y[\$%N46]8,<H/7$MVQ>5ZKSXT&?!'>I:.8D_^#LN!W M##K+GEL+ML$>X&WR.0YTP5K8Q+4[:X@`FP7O9&'[48)RV+,S^.)$P8B/BN6U MP&H8Q@Q<N4@,JT)BK<Z]6_A-/_BM0E&4![/P_*(*5II*(E=]T.$T@HUOF'6. M+'@5&@>?,XCGJ8$/0%TJ,,+^%_6,H[R=-8[T<QQ/7_@>3AS1B0FT%_!(UTWG M8-U1FMDBI'\%HAU?3R;@,PE^+VQ#+'=6+R%L&X)-K84"0=\-L&BBVNM/VJLH M-*<G@_U-(V,@OA=A`:W&[31`05@N,CML8?VW,FVCQ@?6-H1''/X[!/QK$'SO M8`KWK$]6#`DHA8\S6>CN@N#ZA3\_'B-&\^)$9Y#164A&9<1]@92O)PPN0%A, ML$.LNR06CG+Y8V_^U1&".6\,8%-OB0`U,&%=TO&CJQH84`ZASCAR"A4YL*80 MNPW@2HGGEAAB0\E$XNG<^SWX"OML,&6WP1**.<-_=`:Z`@T0W^]NK)4B_=Z/ M]#L5:8^%?G>5IN)5:U8A"X]7<@+VIT]L&H\W%MQX@(W&0?4XAJN'N1LB`%8] MG*7YO-R=W*N)(^,K0EM0"):>W'H<11<8!8&+AIR=A^/1M7^X8Z.C#G7>%;DS M\'<'#1T1V+[QJX<,\<E.7_0D/72*A>_02@NT19\O_E!#H8@TZ3>Z0#]($RSN M2\FXO7V=]+8DN0$O>.<*1[?CL7:4BR-`J1'0I]H\+/86EJOT`X5Y:OK:50:4 MI.J(O'(,#R!LZ&@V3V@81WOBNQ-D'%6$603?Q(_GC7[BUMY,LO8>3"1\S\+? M9$@(>+PUJ2\,9I5%2*37A-I)S<UT]D""V[EJQ^4G@H5QP&RMO@OZFV!Y+RK6 M8GLA'"]T69(;R('U%$[NK^#9>C+)$8ZZ0_-![F">AFP?HG71N0'WWB1*ZV)F M`96_&IW9U.@\!#`F$\_LBN!U(85X:.<0Z$H3P%7'D8*P@Q)9HZ"O(()C8^-7 M^-MY#$>-$;@F#_P&,N$^T43D_75J:,P%"?"I.H\AIKQZO$!&&-F)AF5$[(-? M,Y$87P1<!E$!'@7G@$:#M86S'5#G/:LX[MR'OU3J)E:2L41^T#=);$M$>9=H M>_&,1`<?H>9Y-1JUYY%&6K@6/M;3J/_H',1WK*?Q_X?OC.+8`]MC(!`>M^8! M;0;:#70:J!%1&]<4H:0ND#V0)]!$H#2@Q4`E0,>P*H&>`'T1D`8?^'>^U9C4 MM6"JJMA4>;&I<F)3I<2F<L2FLL6F,L6F*HA-51&;:B$V54MLJH;85%VQJ=IB M4PW%IAJ)3>TK-M58;*JYV%0SL:E68E-[,;NND=CC_]+FVCN%X9`+#ME+PB&; M1RASX+#(F2SN.,1A.<0].)+D+$7.9*,Z1X:<R3MS.7+D3.X[<A3(69&<E<A9 MF9Q5R)G\2CI'C9R[-("EP?[0J16:B'ZZM"B<RNS\AP\/8^$7I.%KC(3S"?HY MSM%UJ'5#]ER:Q$=X3L(5XS=SX!UWU#<D5!I^6Q]^8QM^:Q=^<P]^>PM^@P/> MQ8]W\N+=G";<G7H8'^^^P3^3CK^CP.M%'83(_7Q]1$4M6%!C+C9^$Q3>#8[+ MXN"F!\*1,X71%R$RUY+[;08J/_/F88NY`/004>.:<O!J2`6Q6JEON-BMN`ZZ M#'G3`'>\I`CCT+C7>HC6:D'X$JU8#@939ORF@L?[?X1?B1NU+,OL]96?:*D> M`>3UFBO2?:=`_B>I"*]2/^.)21[16YWD<>,X9!L2%AH[BCQBM#;(4=N4S$CE MDJT*J/.@,Y@L-DV"8^!&1]3V):P,;*HX_W#X+)2ATLD0*H65)MV*#70>P1+< MRD3,$2F2:V+N2):[Q8DF@F<D!D]1"$^/#X\A@*<H@B<G!H\IA"?/AR<H'U,$ M3T4,GI00GB(?GJ!\4B)X"F+P)(7P%/CPF`)XDCP\=QJ%Q^#B=6XS&TSVG<VE M85UW;C-3(((a)4MIP0-L9SYY./R<53!CPV5[XNR63_4C*.B&2L7Y1,YF\D8XN1 M3()76OHO)9,0D4SB%R63^AO).&(D8_-*2_ZE9&P1R21_43+.WT@F)48R%J^T MQ%]*QB)(/GR227.1`(a)!)5ZQDUVAX0!.4CRU&/GY4F7^$RA*#&LB'*LM%G0RH MQF)1K]"P9)W8&F1X5^#3@R`VO__*<['Y_9<AY+\,/@S!\8`A(JLB,%S_H:QT MH7J$M:O$Q?XU[=+^!E7Y'Z$B,:C\>E45HU>ZD%Z5^#`$QVVZB.5K</'X+5^6 MAZ3&AR0A@"0K@J0E!DF&AZ3!AR0I@"0CTD9M,6V41[E4.A=/E0^/+8`G+X*G M*P9/#O!T^?!4^/`$=28GTK-Z7+Q?ZUD#OAIH2-1>]/\1JK$85'Z_,>"B_A._ M,>'#%F>+AF+T*KQ=6Y</0U"O"B)XQF+PE(7P#/GPZ`)XRB)M-V'^\S&C-U\] MG?$=?V^9,O^)=_?B0V6*0>WYCU#[B)&57Z]F8O2J*J37GGP8''+=J5=5$3QS M,7@J0GBF0C)UX:F(X/42@Z<NA&?.AR<X?JB+C$067#S^D4B:AZ3^ETC2(OW1 MAXOT:_UAQ%>#.$^W_$>];,&'RF^1G:WNRY6U'Z#B&6`>PC-!+DD7U*4^'Y+@ MO*LDTC=68OI&30C/C`^/*8"G)A*YJ7/Q_MO(38</NU.7_'B:OXBG]3=X/7X1 M3_MO\'1^$:_'W^#U_D4\33%XV%>4N.M74ERNRQ8IC(LX*[>7F0SL)8,8-%*^ M"XF&E]!8-`7N$IHD<J(a)EM!4274*[MMG0AC+QFS8HXBVA.91!CP6K'0<4`A0* M-!YH`E`8T$2@<*!)0).!(H`B@:*`HH%B\?.!0$E`R4"I0`-A/3P(R!7(#<@= M:!C0<*`10)Y`(X%\@/R`1@$%``4!C08*!AH'%`HT'B@,*!QH,E`$4"10M!YU M!P<'>]O!EZJA7(a)T9ZI.ZMX&#2YRB"*M?#57JD^+@D!:GO(-UMH8N]4EQ</#L M`%&8&ZAE$40FUI*='%ENF41IJ@S^I#@RW'JJ):D<U;PRTES9%#E<23B='+SL M.:I%W65)-T!H!:.3@Q=8&!^GX,]KZIT<O(@[JL@MH\%?1H(K&RFCRE\&+TF/ M<KAEY/G+L+A2DS(R_&7P`AM;%RZ#;P!VE6%PVX-3\&=7&7Q3(Y%!W85,E*'J MI3@*F*-*O078$UIZ@X>FPN7@4-<2]S"/@T-MFBS%&:7"CX9O[F#'\8%Z1D$/ MQ/`X.*14[$'5<PWJ6<K&'.P9.'C=JD1QPC7Y.3@$WXHHS@%6)P<?>")["/B7 MX#I?@9^#@VTW-L7Y(L7/P0/$.5GJ;NUY^-RMTLG!0Z49A^*8<_@Y>)#:KDMQ M\&<7!P^'VU6Y'%5^#A[('9@4!W]V<0RY[<'/XEP#N2QY&L4WL0Q@L,&_*(%U M_HFG-SQ)98%,:7!]`'+H,3JUH\>M9Q.BI,>?5!D<+JY0HWK;'7PPC5=/'VY+ M\>"'=:')*Z//U8X)MSTF/(X!MZ75B-)!-8]CSI4:]P^6NIDGM1E7`LQ)T@=[ MX'&LN!+@]@1#64V>;#C8L0:)E\)U.-CU!!X'AY0K0+9MW'JV\23`@6RGC2;J M4I\4!X<2[R0H'>"R^)/BX%`2VV@EHG)4\LK@^[+4G6A9\N0IWH5@BQP@<.B' M;%!_^'-`=FB$/O5>3GQX\ET_1=3+JV,`8!K028C"3P&9,*B>9H,U2@$U`[T& M.@;"5@#M`2\]#I0$/J8(>(%`04`A0*%`D4"2X#%27-+CDCZ0(9`S4!Q0O$'7 MNZ&Q3!(P!7=/^/D?X?^8B$ZN)7B?^%@$UGH=Z`90`B2E`:T!,@(?M`:R`0H" M.@5C6"U07V7H::`OT*ZO0*U@:!^`3L&8>EM+L*W_;=N^P@I'`=IE"C3`2+BM M$A#`4"3!^]020U0KQ;65O_S/GRK$"O"S@*GDF9B1R!)1WZY36^DV0*#AA$)Y M6^VZ-MF%\DJ%\>4/)4\234(QB'JN>3!Y2BX<K@+(]Z=IY!MO_,1@**1@#OXN M.)S[?%\824LB9?#S=!AO!`I$080S#.8">[C"WZ1'$`0_[O,[<5R<,*B=_RFX M,&0%MFP#Y6S@"O^)+S.-/`>50IX'FD;2J%K[P)F_9E$T-\#S@'S6R!&\QQJT M1SV/^:]._]_HU`K2_]7IOSK]_T6GUER=CH%:?$!#OG`5`/\'HB$@BQ]!347X M27+J^?-IH+5(\E0>?I(NG/R']S=%(7TQ]8=#GHA_^_1?/_E7I__J5$2G&__5 MZ2_I-/Q?2_T_M-04>N;/GS^W.;TT_2F\3136_;.!=_;Y1DEQO#G`6WEDVU-Q MO+G`\S-:T"*.-P]X%Y[/?"^.EP4\RR:K.'&\;."-_FAQ11SO-\Q[+>,HCK<( M>'^R;GF)X_T'>)DN^U^)X[7_Y![RPG$_>#Q1,O1$+YOQT#B>'3NE3A,!H\G M6H[9#8_5#0]_[2+)E!;+D^!>B9.%TPU/DL<3Q93JAM?Y<S;B>#(\GFA]LMV4 MD^N&UYDDCM?Y*)DXGF(W/*5N>,K=\%1X/-'VX:_0_\I>U+HII]X-3Z,;GF8W M/*UN>#VZX6EWP]/A\9#002-?2^>$642(*Z?733E]'D^TG$$W/,-N>$8\'A(Z M:-RO]\7S3'@\44S3;G@]N^&9\7A(Z*!QOX(57ZY7-^5Z=U/.HAM>GVYXEMWP MK'@$'30D'4W/)MN>/VZX4EIXD</))GB>%/B2I#<\$/^XGCOPNN0'WVTI3A> MLNUO_\/>E4`U>6WK0$((`2($"!&9QS"%0<`(,0PR3X8YS%,89(P,$1DM(D6+ MJ&AQJBAJE5I$2Q5M]594KJ+5EBI:EE6K2"U2M3Z?Y5K'2T2!X-ZMNMYZ;]VU M7ERNUO7]WSY[.OOL<_X_?V3:FH:_@;"$N9=E/+</5$#8"ETYV;4#LGT05JZ> M+^O_9.V/$!:4LD86LV%IU544J\T))F(R0Y5*B:WBC6LAS)#V"_$;8]O+<5.A M4>P8D4TZXZB5!,762#F,=%D[="F$9<M7DD8K.JBG,.TC%%L^9P7)\WB/$H0= MUME,PFP?<<`Q1^87)$HTKPG"/O,Y2LHKZI>%L"[A&1+F3R>G85)U?-U1")L= MZ"Z'Z7+<9)6<)H(Q8F[+=80V98$V*#R6PW2YZA!(7M\VLAV*@X;+3V09@@+H MSW\E/R>WG6"Z0]A]^[GRF`W//&+D,=YM4IL\IJ>B=[L\+6_C/R#LWIRUE(=) M":9OVB!#X/J>I`S9[=X.\39XWZ#T7J<Z0QA=;YB"Y9F<ZDP%;-Y^:KU)X3[= M%]!EM#_+(%'S2H1)H`W&"M1>HZ??0AC74YF:M*GG#PB[R-6DZ@:<;8<PE^+% M5,(]VQ`($]MMH/)"&@60GGIE)ZF8[:<6#:)8"W.[HH.O!P/"2H2=BFV#W7P( MXU1U*V+Y8A$^C&+9I<I*F.V+[`V5L%RJJ(I1LC(;J8>P5O-!)6R\K-S#RBG\ MQVL@GQE&'E-61'CUA#C:8Q$/7!]<9XAH,(a)C/4K28UA$B`.M+$W4O#8L#4><4 MC7I4*1'"/M;\B8;9=]#W)HUWE+$.PM86^D[S_\WJ`(3)\#*GW7IT_C28+_2Z M:8.#QC80MB2X>5I[9V<$A.F%5ZHPOJC[#O+USYP:%<R&GQ5VJ1#2A(,@5OBM M2O,O`U1(YO?R,JJ8S&ZY$`D&UL%"1JLJEF=_9LO1L?HRAS>#_D#,'H1T>6QB M2=<>;-.'L+XR$=W^9),8DNEE6T_'<JD]>@6=FF(/[E-MM3ZAQQUM\H<P\_P6 M.I9G*=.):OSYG)N0GH9B%36"=G\IA/&#EJCY[V=^!,F4Y[6H<7+W;X.P-6:[ MU4)/ZW9!6#RM1PV+7X;O5;6M:;X;(5WN%O/5F??IH$S]Y'AUS':GN`QU++:\ MLEKU!S[5S1!V@]&LCM4L)=)6]0(6XSF$/2W;K5[0X-8`V1!MKZ*!Z3FS6$_C M2NYY$L2+4A9H8/G2HOTA*O-G]DX-L0;S$H25+3BJ(6"3/X#&<W9P96`UA&B2 MQ5"22["$>-["`XSF.B:XCB5Q;C&P-2<@_AZC<P4I%Y)9+V>L^?`/PA.(E^?C MJ8GEDEE1AF;_5_K1D,RKY&Q-<G>]$H0=L5JEV?>9:#\D\P+1AXGY.GM.--/W MG\[@7.EP:&":"@^!<TPG9S<JLUNKEXG%W3ALA%GSKSN-$-:>ZC8=ZX59G&?3 ML1KRIQ=-"^MM`O2X6IB>%I%KM;BK^RY#V)W<'U!>E8?V#*P_.Z&7/..[#1Y% M$.8<K:*-]>7"A*7:[)5;`R",'710^\[6<!:$_1;MIH/9'I_FH]/91AN&XO?A MHAR=TB\[0)FG>4TZF'U/8]?I](P<.0O)-%#?IY-^6&LAQ+L8=50'6\,;A1:Z M'A\.G(9D.B9%Z&+Y><UJEV[M1TTZ$&]>?*<NMK=84-FMB_52U?.&=#%_MJ9S M]`*O$L#Y8#LC7`^;T]>GB_6P-7QH5IL>-IZ,V8]Z70%.#A"63?E53TDTU`UA MJLK6^I@NNR/C]0\9+*^';-!U.:6/K1U;K)ZB,MN4R`8;/^7>A;!\=U4##OLQ M&(=;1HX&M1M2]D.ZK!,(#+#YIR->C&*T>;4&]4$/P#K_R.2\@:>"TT$(N^EH M:A@7\'08DMF?'V:(S=LVXXV&F%_662L9U7:$QT`85];*".N7CGCV&?4&>:2# M.>\[8+2M//0FA#4'NQIC=5<F*\T8B^W"H#7&<@@O7;_+N'&3T`+R64^NM4G/ MC5NI$';8.]:D=U$"'\*V+UQH@NDIJ[02Q<IUFU!,2?>@"18':]7_-L'FBF5R MHRE6ZPQ2OS7%:O)0/M<,\]D.HP`S[*R$/Z?(#,O=C]D+S;!>XX7PH1D6/X[0 MAX7Y17E^,`OS2_*,4!3;X9K'PNIN8>XBUE:;3O!^U<W87UE8K8OV?(YBK=D* MYI(a)NEV98FF.UG,;9;([-S<=!O>98W!NGW3+'^HFE!L/F6$VFSW]@CO6*Y[SE M+##L#V\OB_(@>B<T']QL0RRP^)E9Z5MB<?]%.-MR6">I&I+9G)=LR:I=?`CB M'20H6&&^5DK*L4K9N7,`PC8F?6Z%Y6Y<8+M5^](Z,"=.B/9:8?9YED=:8SW# MARD5UNAY.7.9=>;2C5MBIT*C&+?H8VO1KL7@/%H9_ZDU5C\C3>AL+.==/'Y@ MRR.Z;*4]86-^B68R;+#ZLC?1R0:+[7Z#5!M,YAESD<TU/OD`%/?AJL4VF*_/ M^3VSP?H>OP6RMFR"YS$(.^XRU[:GX1JX9[:R+[>EU@A$X#Q*T++3;^]?!^E9 M:.5FA^D9K+S1#CN'EA,,V<4MN78;DCEWFK8]YK.U)HGV6&W](a))-EC^ER(.P3 M^[[&02MHO/+B%GOT3*!L!ZK+X^CSZ'B.J8MF4N]RP7RA"N_-Q-:<#E*80TZ) MX3)(S_/%9QSJ"=7@GKG13L,1T\4JBN%8;BUZ#/'25$,<L;I;6K3!$3N#-R9K M.V%^T=`3.:VZFQ<!C;?(MLYI&L);4+;*"5W[R9N<!'0/\&RTPG"3$[8G,<_; M[(35@D_D.YPPGU64?SH+LV\+X0R*&9J.S&J@>X/GY8/J!`ZV5CTVRY^-R4QP MWHMB+H(O9_=?\;6$,''Q^=EW$BA@C72(V.'<6<_]`<).%+<Z8WZ9*]_F/!:C MJ7]&8Z1YS1D]"S+2=<$P5HS8Y=(%3_`>IJQ?BPN6@_&A@RY8W#6S?W/!\GJG M&HV+]?-VV69<S/9EV6QNT;</8Z#8V@9]RCUR<Q58RT^'1L]ISO1="&'W'5+G M8./=-EF,8J6Q>^9@_DQV7LG#L`BUU2BFR&GB8>/I:'?RNI=^4P?94"(XRC,! M<X)`L(P_S0O6N(a);N#7.$CWC8NOF!BY,KMO9?,BUVQ>I$D6"MJ\-J&EA[9@4> M<&UL,P7/2CH4K[@6/`K<!V'"63==T7LOFK/=,#VO,_S=TONR6)`NWQ.'W3"9 MEXO\W:__0@'/#LNH(>X%\QM*()X2.]I]<43[:@@[RRERQ\ZO6VS[W;&<8.K\ MB6)?VAAX-%R1Z/DF%J5EYX'9I^?AY.&KW-$!82?BVSW$M2PP#HRX@+G8?+]( M'FU!*=O@,YV<+,]`B_,Y$#;;R,+K^,E5NI#,%<Q97BPF+P/"!%Y^7IA].79[ MO$*%))C'O>B%Q6&6S6TOS#X[/LD;&\_$N=X[[B#S(H1]97#+&UMOV[*/^=3? MU0)Y5VRO^F!QS_=YX=.[1Y`'^5,CR]$7X]G)K?'M$I1\`F%'2II\]YGS=T$R MKWCN]/V$Y`D_!V&AXH>-Y^WGZI>U<1OX/,K=X"@_K(=NB3KMAZTY#/]?_;'Q M&I-U`C",ZF,:4*K_2!ZR3S'=,@"+[:5*OP#V!1'XO"DIJPSEA;NN#\!ZC4*' M`P$-$27@OOA/PPL!F.U$?DP@M@=:5^D0U'3Z4"AD7XA!9!!EB0.8U_5\Q6!L M']=%B0G&>K>4O/3@(Q>U-2&LR+DW&%M7_A&I-@_#]I&=YJDA_OS1^,"\IY<: MP3/'&U7WYZ'[CA0"'_/G:7(B?R1WU54(^UV[D$_/6'X,&N^.YTH^^LQ"["H^ MEH-:,9?XV+I9G40)<;#8Z0"-MSY\3@CFZ["%'X5@,E>'4D*9NP8W0EB=CGHH MIR?)"QI/6]$^%)-YN8H?6O3+^EX(.Y?>&8K9OD[]?BAVEL?(a)&H5A\;.QL`Z[ M991)AK"0O+PP"F,$?-9HH]ST\)(GU0J(a)/\4!X1Z=Z10P?NY-X9CMMM.)$6W7 M[<'["`I^(1&'LNY]#HU')JZ/2`AA"B'>W(JN"/T4PST0]CQD>B3F3T]"<>2= M;6QPO.',&5&8GGI5<Z,PF;M3+D5UW_8'ZPN%=#T*\XL::R@*BU]5X$@4]NR= M-B%`\.`H:2^$^;`C!=AXGQ7E"#`;U"BE@NH?=X/G&DKQRP5UY[Q/0+Q]\ML$ M_N?#09_M-=XCP.RKC;(a)D.+*X'IQC']#*HC$]O=Q[HS&9%T,8,9CMROKF,9Q2 M%1%DW[+"L)C2KSM&(&QE254,IDM36$V,Z>\<\!F>IHC&&.RLQ$-N?TR;\#YX M#VR+%SD6.\.MI\3&UL2L5X6PA.+66'YL\`^03'G!V5C,+Z<7G(W=9]AT$L)B MZ2-QV+I9S*+%=_+N+8>PZ@KO>&P-'Z9LB,?L*PS=&]_T_,X^R`;OJJ_CL;7C M2?2I>"PGKN>?BL=L;YIW)AZ+T3F>;`+&NY_DEX"MQ6GDH`1L7V5?%(7*W$P3 M)V`]PSV]SQ-&,G-60W[Q<F$E8OD9KN^0B/70Z1'<1,SVM+B\1*[U(RJ$#26= M3<36L88%PXF8#0.>-DF8SQR8)4E8_);.KDT:T/JN`[+]HO/Q)$,+YT40K[^\ M+TG_1?,R"#,DSDXN.5$.[N-"+#V3]VW[#NQWL\0[D[&>J$3Q4O*0NV<XQ`LV M"D[!_'+49VD*EM?.*JM39)<_!>]-N/O=20F\OQ;LP0;3%%.QOB=REF4J=L^T MB,5.Q>+0&E2=BN69?T%S*I;7'VL>%F+WSLC,8T+TO,?EI#!N<(<:I.<=E0$A M%H>?<A\(L9HU:ZY2&G8&GQ67E(;I^:#X8AHF\XKLS31A2O@12,\CI%MIGM=J MZR#>,J%J.C8?5)W9Z9A?.#,BTFDYM^C0>+^+'Z9C\9ON/9*.W5-<5>Z5@<7O M094H`[L_MK:0D8GI:65JF(G5B=ABNTQ,3]O,/E3F5],TYN],[6N%,/F*&?-Y MV;P>",O-+YZ/V?>CPIKYF"Z-%LTH;S[M[/S+/)8::$.V>19F@^6L[BQUQCWP MO+6,D9&MB/"$?'$VIF?7@@^S,3V[K#:@V)=9OV7KAO:80[HL4GZ&\EHR7F0W M;^E:!?%4*<0<3,\TL4H.EKO<.`;*B\SVR,'VVAJ6:3F8GJ>F[<K!GG$1\7[+ MP>9TNL+]'"Q^V^((N1A&#U;)Q6RHC\&Q?>YQJ,S?O1?F8M\+^9*Q)!<[4RT) M:<KM,V6#:YPBOR<7Z\_N6);EC5(a)]!L_\&RI*\[&SIPN)F_*Q.O%%[F?Y7;DT M74AF=-#5?*Q.;%8+$K'K>FH@;%Y8G`CSV0U?D8CYZU`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`ZY]K2'X<DR)/ MEKQO@_TWC*URXXRM<F_%X)#&&1S26S'&?I7V):-7=A)CT@]2CG\4`3OF^8_N MW4?_YK,(DA^E_&NJM$'O1)6V[)VHTB8"5-P[F3+CU$R9M_*.-`,:S$UZL&Z) M5,E+L6NZ1_7D/IL4-#>9UV]>F7)EG9SD2M*K*R>N?OW"%>FK`RF2J\GC5\M. M9KQ^UXH4(X\D85"D&$0IUNM7K4QF]2E+6-0W6"1IYNNWK4QBKE>1,)5`IMP4 M]NN7KDRPX]0D;!K*)D^5,&VJ!'V&1(+*7TJ0?T.*ZA0I@TR)%/K?2J$07K^< M`O[($"S>/\E:WS.C";+2R3;!8!)5:_AC#":1/"7I"-))-\'J&4TAROA4!Y*/ M()U\$\PNLFI-R;.73!&%C"0A03H))]B'1I-]Z.E+=K\"^2^2D2"=C!,2]E%5 M:P1C$EJ5R'^3E`3II)R0TJJD6M/SY*44%1KY+9*3()V<$Y*VTE1K.&.2KJN0 MWS))"=)).B'M^NA42WK\4MH^-?+;):M$(A'(*")07D>3\*^7'2*PB+S,7,+K M/'R#HC&>N8=>O,Y<1XJ,$X4P]L/`8S]?+37RV\\=^5&R_-C<Z4?GSH2EML\F M62JY\.4+L$AO3AZ$(KE.HJ+,*]K4V8/0Y`@3@1JG3ID^"%6>(!WC";KT_"&. MK[.O\G9TG1VE*Q#>3)%)(J0F$")"$1`AR;+)8B;/(.)XL_`J4T<KR*@8942, M)%FE1$V:0HBH:7\A2I+WTN(FYM!$\CYZ-"EY5?]&'/O5:\LDP1]/+I6Q%TE) M_LV:]!.B!H2QET&]^/_/?\3G/Z6`0V6-#%1$1PIAM+!*-09O\"8JZ41!'OM] M=GGI,OQ_1R:Y29,S)?^D3)@KN=)\;)@I5_:.-W1CI?3URO3^$J$K_Z>UE-HV MO"&3.DGF6(`EOS0^MA;I_Z^RB._$DBRNR>_!DUR=,LHBO3,K^>44>O>QV&-O MNWSGL=Z=E?)R<K_/6._.2GE91-Z%-9:'E'<>*?F]6!+]%-YKK'=GO=]8J:,L MZK_9^Q+X*(KL_^J9R61"`B3<-Y'[C.$&08B&(]SA!(a)%)(.$,)$"X07`)BBXJ M'HNLU\*NKNBZBHJ*J[\5CP6\_J*K+KO*BC<JNG@RW0W=_ZKNZ9EZ/=V3[IEP M^KY\FLF;Z??JZ/=>O3JZRC5785Q<,^/B8CE,=<U5&!<7RZ$.R,50U[8QRAW6 M@$;YA#IY$G'RU6W$Z"%:7JV,BE*]E<RK983J#%8=`YM^C$`ZA]BB4XG=F?;9 MY@MRA`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`N/H/,5 M4)/>H9MT@=?OM)G5>4]0WG2=]X37[V8X1.??0(UX@V[$&WQ^MV,ANHQT6NX3 M>KG3D_SQ>`&K^HO7!>BRLOT9%8=$35:VW^_4_IVU6W9Z7=-".7.'I1,0&RX^ M_^U6:RISH+:Q7G,#,,S+AC/Z=]!&>KFQVFJD%]&[33U"GX1DGA,GDC!K8LK@ MM)XM/%9$$<"ZCD@PYSQG]F/,EN(3*DZ"K/:^>[<07I"V6XCX;HN83K\_AQ9M MGUZT',-WAQ?KVOAN6!WHNR\LW\UP`<0U;OV*U=!E*N>$H^;(JF2,-O%!7K?9 M9G#9(EZ,'<%H#]!=RW1XJ9@3$3V)WNRR9I@UP=TX8\/.H)O.H#,[97#I>LX. MB]\]2Z(a)BHWDL6.*=!43#=6BXW=%PXS#<\V$J5C`)L%\.&&<4$=N&)E'=>DC7 MK4D.;2B*Q;"A4$HV-A3%QMM09!C3$:LY=`Z$V1-;M6X1/3NVGPM[(4C<J];C MF"9+>);^C'0<K=H?AC/>+#*<W?Z(5?ZL5GOW[T!"'L%8X^6<J[*>M#^L(Y&> MM,[II#]MS^VT5VTOP8UW\%MX!UV*6P=A+RD>'^&W\!&ZM'C=A+U$Q_.0)K.O M?#DX3HSC!-X9"OT8XN@C7>S:B-/)OX)E5'&PA/0BBL7Z'3%GRZC.LY%6=T'; M64_,GL/M;#744*^FFETC6FW)@S,>%_*,1S5='HE";(USN.F#LSDSW/+AXMKR MX3QQ?F?`-Y^=V3I7)6*(8Z#%679TV&7'9CX]@//IV+I4R5K(A$?0G1O=N<SE MV79(<2^+C#.C"<ZEG+W75<,]9&=LO#<B!JMEYSJ*U=U03Q0[OJY:A>]`Y$`7 MWCALOKJ/L#3?M/#RQ"Y$'Y9@GX0:-(-+TPJ-Z43S6!0,!#XV[2RA%C57T#AZ M^^+<>,UY2\O%X(%DERTMC,#=M[1<_)V3&F=+>T:B[Y4UJS+ZSJOE,/IF<!D, MODM_-O9)-<#V_^9I=FXX3YOO9WOD\[2BZMM2Y=2NXV.%9/N^\K^SLVAX.KG_ MSMH\?8.J(a)B4/AKQ[0O(,.FBBQ]>!].X0S?:2YN6Q\R2MY!_>_YF7W<_VW>9_ MWY+4"M2J<7_W`_K][(QPJ]^WAGXWZ.,AFNV/;W5_WD']=W96'?_[?_,^:<G3 M[+Q=GF9[@?,T.Q.:IP_4VLPKJW86%T\;Z=\?2I_M)VSU.WE5_[W?P.K5^-_7 MK;\JC:?9N90\O7'4?2!]=K8H3[-]\7EZ"YE6@Z?9N6(\;>1G6B@_9M?I(1`^ M$QTUX6BB#?E[7X7/S_>:3IO]KD$7A'XW^\F`B3:Y0@(JDP)Z.D)`Y5+TR]L/ M1+(SMGB:G6/"TT;^]X3RQ\ZFYW]??,TKS7AZ;-IJ4$1V9@-/L[-Z>/I(]Q&@ M"MD>T#S-SB*UHM-?U_/#SHSE?P_3;^B_L_/D^=^_799?QXHN"-W/SB#B?V=G M:O#TR)FW`Q59VVQ;:YYF9UWS]$NMMP*58>>Q\W36R&>:\#0[6X^GV?DI/,W. M*>%I=G853[/]]'F:G;7!TU,:U+N4I]E>UCQ]=>Y_@$ET++RM#4\G970#*OAJ M_VU->9J=(\;3[.P-GGXK6>!:*_T<19YF9_KP-#O/A:=/=,T%)L+.(.%I=EXZ M3T^J/KDN3[/S?7AZ_95-&O,T._.8I]EYHCS-S@'GZ>M7E8#R+^V^$=0?.QN` MI]F94SS-SG/AZ>S&XYOS-#OOFJ?9.<\\S<ZGYFEA?G$K<'_^<B"?G4O&T^P< M*)Z^MO=\H)_LS!R>]J3=`O2=G6/*T\WZ'LCDZ<=F]`3ZQO:BYVEVGAM/]RX: MTHZGV7[_/,W.8^/IC^O=!^R9G2_&TU^OWP#2?Z71H08\O6!U=>`BVW;*!/'* M>_WJ`Q?+SG+B:;7HI[8\S<X^XVEVEBM/LW.;>'I`XS+07K$S2WB:G?7.TVS/ M>I[>-;LWT)>#"SN#Y\/.`./IMN5S0#S"SL3E:79&)T^S<U5YFIUIR]/L_#&> M7MHIIPM/L_,A>+ISQH\@?PL"7X+\2R,/M2<<V-F15G3%&[!]W==7I]DYC?S] MQN]-^NF_LW/@^=_-[;';]KZR]MI(?V4_F-]W373VY9#>$J)!IX,"*`MW_['0 M_2!8(EI_```X7XI:)AH$TQ3`N"B`,Z6H9Z+-70A@;!0-370C$PV<,05H'"F` MLZ4`P0?1M[?ED6FB+S'1+4PT")8I(a)#.E`,I+`1I'"N`,*(`S(\:H>P2J#4;0 M_DDFC=[8,ZV?`WF:",;++$V$C(K/%#!F=O2$E;A=5%P!F6<IKG%8G#YJ`<1] MK(GST=Y0YBU%[G.CLQ?48:D75\+>-)I]AL9>0?M6F818LK,1TYH:^\/:B&GN ML%H$B+@>U,=G,43%S,DS0$S.`2=B+.HC(N8>*J*`E-H6*I4;!K80$8R1@Y@/ M-)*#\0>=%$*LK"ZLL%L3;:V]T84+#XVQ005#1-JK<19P1J6Y8YBKB>_F7A&N MMQ7_"A6Y0[IA@-]"I'G]O5'2=-JM.DH_K;(1_<Z:Y3,HH2(*%"^QVT'6"Q9# M"NP&NQ*8L8^*WM%SDV4E17*GCPVYU!"&)J_KC]ACD_.8+^P90M+>J`(A#/E] M(PKG;,.VZ!(ASA36]HT\9F<OMS"N-_K:.Y%H!0;;9QD)M^O'_+3/&]L/K?#$ M<+1VJ.CGQ/CU4X(a)L7-Q11^P6OJ.#QM[O<GOV)AZ#O<)+/>!IB]2W4_:"B<F6 M59L93GTSK9>CL?SS3RP7Q"=X+,1$[XK>^2S;W),$<7:@#ZJE4$5@4VLI`?:I MYG@W".39GD2[[O^\MC8B_VQH?,?X9-_M\;PD"+M);BLR:/\&*_%G$S-?:TO@ M)(0=/(^$_OB&=K..T^L_M/OP`;UNHV6[/53NO]&K5Q]">M.K#[VVT.LF>MU, MKWQF'SFT42,M-C,Y^>1P%YW>,XE]YI.G0Y_D1OWWIT>SMBV?>$)TJV7Z[Z^M MUC\S0Y\M0I\'5_._LZ&0(?ZMM,H]'K\WR9?D\;*P`W2B#=\\GG8W%M*@?RG) M)*/HYPKZ.98&O`M)(5E$?^]!Y;!7]/@>MJFWNH']-XZLHCPS*2<;Z.NFIY[J M]WD8;%._@BRAZ1?R/'Q*E>28D"D]MPH;4F@Z)-F3%/#X]'N2`Y$<ZAK;6>VF M9I)L=4_V,]DO9S^677FYAZ33W,!A>\O<-,HLH=<2RET[<P9IF/D0@3D<K96J MX:FDH?7WY>4$8CR'8K*`E-,<+**R\FF-+-/R-I9^/X?^74+SQ::`>O6@)68U MF^*DQ)'R]G58NWDTO1*RG/Y?3DLQBZ:J\U).C^`1`KY`LC9X8J4#DRC'(IK; MHM`G*W\+&O%V"6QIH+<36@Q`REOT&/Q&QQZS=]#/F21(OQE,969=RL9-V$B$ MCWC3V?#RDO:$]#_.OF6#BR=#6;PD>E(<X1@/^/<E[R%[Z`-KELW&F1H25M-L MO.<$]52[ZQIZ4!E>I=R/DB=,WS+=6RZHJC%>E4%]'%LYT8MTI?]G4B]13'6* M:9:NY[WH59?D:AZ@A/[=AEY7TGN8KA<39E6YVIWE]&_F6=I3RRBDFJ7K5Z>0 MS2[3)!K?#-)L;PFEBNEWLZA?ZJ3)T=/H%+*I$BV5I81-,S2B^9JKI5E();%[ MYVI>H"B4AW942BF-:HHUS68VNH3F),/$-8YD1?$B$`@$`H%`(!`(!`*!0"`0 M"`0"@4`@$`@$`H%`(!`(!`*!0"`0"`0"@4`@$`@$`H%`(!`(!`*!0"`0"`0" M@4`@$`@$`H%`(!`(!`*!0"`0"`0"@4`@$`@$`H%`(!`(!`*!0"`0"`0"@4`@ M$`@$`H%`(!`(!`*!0"`0"`0"@4`@$`@$`H%`(!`(!`*!0"`0"`0"@4`@$`@$ M`H%`(!`(!`*!0"`0"`0"@4`@$`@$`H%`(!`(!`*!0"`0"`0"@4`@$`@$`H%` M(!`(!`*!0"`0"`0"@4`@$`@$`H%`(!`(!`*!0"`0"`0"@4`@$`@$`H%`(!`( M!`*!0"`0"`0"@4`@$`@$`H%`(!`(!`*!0"`0"`0"@4`@$`@$`H%`(!`(!`*! M0"`0"`0"@4`@$`@$`H%`(!`(!`*!0"`0"`0"@4`@$`@$`H%`(!`(!`*!0"`0 M"`0"@4`@$`@$`H%`(!`(!`*!0"`0"$0$U;3_A:COO>F$U*P()&=4;%?J[D\B M0B#97]I.NU&[V1/-4E=CV>S)J#BJLVSV^'L$A)X!DCNL'AD]S$/&T*NJA'CH MS5DD16.RR_XA*B;[M";FD`=DGPA>:Y:<U#!+3BIC8?=E42:#+<F2;7?M,-ON MVHPM*<2613,:9DVV8LUL$F;-;,)8DSG6+)I^A#W%@CW8.J/BH*RQ!ULS]A03 M>Q;Q\2)2HT7,S0J+F)O%1*1:B,BB1>+%5(\2D],[HZ)`TJNN-Q-3W49,%O%# M437-HG;GA$7MSF&B:L80E45K#(K+,(G;/"RC(AC4U6D8$Y=1B;@L$B")H$I, MY%S!SJ:(_FB$C(I\/?.[!7\L@](+FT,+NR]D4#:%K<S$B(4Q.[$OG:_`FU&Q MXY3&5^#U.S4NG?<$Y4W7>4]X_6XL2^??X,NHV*!;U@:?WZU9Z3+2DS(J3NBV MD)[DC\>F=#D[DL(VM2/)'Z]!Z;*R_1D5AT1-5K;?[\2:F+SZ)!JIN@<2PNW, M7,$_>IA`54/0[6#7B=CZ"%GU5B%D?C8:V<`;UN`&7FN-]!&C1#KL=/(@?;X! MQ:A3/^'K(<QIJ97[:`VNU+6Y+.`/$',-1KBM]')O(*/BF*Z7AU/\U:*X?5"" MA6;NKI91,5F7L"O-GV8I(<DD)5HW=Z6%6X[T&OX:ME+\9DG1VEDCHZ*W+NEH MNC\]IJ3D*&EF_3R:'M;UO%K^6I5*"Q#[4`+5#-7L8E,S:U&<FAWTQ6ZJ#4&H M=1>&UMEU\^I::%V,[A83H'JBI+A1Q2XDFPKJ$F+TV8ARH(J&*%3&7YTRAGL= M'CO]<:&*>C[4*$$N5-&`X/.;^XT6*JAW([)1"5$)40E1">V5$`ZUQ%9"VZ`. MQU]P_(7JFN82<J"F-:9DW0"3UYAJVN=*CP#1E*,!50Y"E8-$E",S-"":H*IJ MH:-M5R8G,BY:F:KZ-3;WJDH,OCA4-<(;GZIR_'&K*B\C$54%<A)452@K<575 M:M=V\)RUS'O5F"US[*DA;."Q@;?1/2CM3([VH!*B$MJH#92&2GB>*R$QH:Y9 M"7=7<SJ4>K'K82+`[CI!'<9AHS.AAZ&N9O2POED/]_G][('SNF=FJ6O6O;T! M9YIC+0A5T$(%]9@'53#$8RZ#8Q5DC#Y4POB44&NSXNBD)]$J2XITTB_>9AUU M^8+195U_$M;EB]`OZT!=QN"`$^)0";N@$OY:E3!-ETC,./M*V!65\->JA%4X MZAE5[<Z5T$OZH`K^6E60E=57%6_*)20DLG33N1A2-6)BYL;*.+T6QNF0J<(; M9JKP0B:?+5/4&X(Z4R)@23E_TR69UDNR_J@?/!$1`!'):XC5>5[MZHQ8Y(:7 M5^69B+-*[/+OMWCF5-_"&=+<E,_N+<_COO!C/^Z+>LN3LEF_Y1G%%K4*BK%: MON4YWA]V:^/]MF]Y^J)'\VW8;=="Z2(LWO)\/CGLP)Y/KO0M3U^T?[814^F* M*%^T@]9%M4H)OY;9*L7Q6YZ^:`]M(\[I>VF.[(-J._!"T3&!.238D>2/O'EA M<%7Z@D]9P/2"3YBSLM7M.:D6J]LCW)5,$M'&UGJ2B),0>[:=-K+VL^V\E)B- M_N[:E33Z0%*L1G]#/0>-/I06H]'?V\A%HQ_7FX[/G$B0U9V?/WL<T1&K33/O M)2S.;1^ZZ2PU6G9M"V320Z)XZR$V8N_!$#-$B:O$"!T)F!NQ8=5K?U-21L4` M_9%M2M+WG:B=:7&W"7$)=*D!L<O,O&?!]WKK5<-<9M;BQ8ZDH])K3RJ'?:G+ M`F&1M%4,E3K!AV8/NT&KFA:"]3Y->(TT8]*=G%W`&67$1L!IL%D'G%%LSK<5 MB6)UMZU(%+N;Q?>ZB-Z^<*/>VQ?_MB)18N+?5@2N:W8:<%I$!S;BG"[$CZLO MM*@2IQ/5F@JDL\7-SL$*;V_O-GWTY#"K%9/EACQQ6HXWQ':^64Z4OIXSRXG2 MSPO;<A+P_`FP)L5D[>0+LW;R^4&3T)GX8[*>\W#>":HTOK3+8DR'8GYT<T/\ M<3QUR-H[\NAZLT<7H*P!G;6TDE3AMF@6(U;VK(=3PJR'4]QEF'620ZR[TMQE M>&7-,.O*FNXRG%<KS)I7RUV&&]0-LS:HZR[#Q^J'68_5KRS#!BIM0(a)^?L.!R M!5<A>N=$7%8"<4)<=N:)R12C-V%^JU![D-I;A5NUE&)W5Q-S3W:P>UE1+\QD M6IA=^J.:['!;P"B6*HC?PZP8A9!S$H4P-QX:,-9WMZS2`6/0,MM;<C'EN-MI M&&`XR=B)^6PY8)OG(<;(M,=CQ\'J:(1JU)&'<"/2-ARPA?.$1Z'M<P4;-D]X MU-D^5ZP]"^6*MF<>?H39A@,V8Y[P2+)]KF#KY0F/%MOGBC5:H5S11LOCX)T7 M]\K$\_&H-"QS.<QR]OIQH3K`$9!?RPB(G6K5TD3>0:N]1*_V.[1A=Q=C;<XD M5-9TU[)X_-%2>#T,#T68]-"9)*=J6<M"&:*EN=%29Q+=*FTM"]6(EAJ/#CN3 M'*]*.Y/N1,,97`[4"38L"<]%5$2:A`I/?*DET/F*ZM+8MRF64;EA60Y9JJ!- ML;-EC,HOQ*C\UQ.NH&J=-QV^V(BMJC%6,;K+1)(-1PR+,"K/(0M:Q*_4(F*# M)7A&(AG+?FJ"TZGV8%;D>MUPE1L=>]PA%OUQ.S,Z5E4A-KVJG!M=U`$H[HPN MO4:8/;U&?$;'AFI"1I=7*WZC8^,W(3$-ZB9F=/D-PT:7WS!QHV,GOH3$Z2>^ MG.'93[MW`O7<W$]9I^FL]PN&5L:.NZ-8L"G`IL`"">@L8ZWBQ2=P7,B1^XUB M,12]>8@-%?W"4_2X%Y_8SW3.Y68Z0XG#NQ(a)L-!?G;!CL.,[5G(TY5P9T]3E- M+=2K6^AI4^(a)7YSHAN_2B-.80OH."[Z!4[JO.U*PC0H?[^HMMKYF4(X\8'CX. M>XTZ%,JQO=+V(0%[M=U8T+&]4G]1!?9:Z8OBCNV5MDQ5:*_N=B"RT*C(FTC\ M!*+`Q:Y5O""VA@7KE1V(B3N.Q0%V%F##@HN\?\5Q=@*CA[$1QYQL'&\ONBM" MU`8`9V_B.+YV[(P\'2;8OKMO.2EB_YIH9>]OUK#ABZAMR"YUM74:`KE\<"'7 M9L-BXQ&SPS9GS8;CQ7$[THMLO#@1V#F&&-:8I>LS[KC!B<`=-^P5TTVKXY[# M?<L66WMCSWW;:F_E\:FM]CJ+3VVUU_GB55OM=1>?VFJO^\6KMMH;7WQJJ[U5 M&I\ZZSC9Z65-"\=J6G>(3A:=K+MMC<Z7;A3"+>PB+]BE*>6[)G%U;6.E8[&Z M)3;<M[G:O3F08S<C?82P&0S!D&?K\-ZF4B_5I;X=O1#!UN%MB#B\#2X='CO. M+L2:'8?#.YH<9C^:')_#"U8+^\Q@M?@='C=\NBLM,8>WKV;8=^ZKF;C#*\L( MBRO+<!=5VL&];O)\/&(:HJ[0<,#(/$ELC!*YW/2)W1A[40ZNSZS,%B[`,=2X M5RQ7\;S`F7F[(_82_#/Q=D?XG2]4:'+A*+3+^28<*K53+QPJM5*OA->01VMF MO%M(>BV2A.&%!5/L*,;EJX%VE9&LL>RD_FFUSK+3YV='A6MLE>UNF$"S8L'% MP9U(a),-=V0;17208KME?DPFFO$#QB&UH3JL.?Z3K<Q*&A1;$8AA::V,6>#O9T M$FQ78B.^ACA>Q+5<(8&&,79L8C,VF,"32&#E=")P_Q#C#MKBGS`B-MP)A*:1 MQV@SE.R:*;Z:3,ANHQ.K+.*.C03,A=BPVFA%Y:/O-B\8Q,44N^VMRI6:X9$5 M9VS8]F+;BW",L^*1K3CT39&BSEFR7)QP!F*`2IQH%;<A5;F@*';V8N[ZJ-\8 M5\MM![O<.(]"['&.'$$"R?K#`B!LM8V$!V(<LISQN;>H-NFLMXI1;=`Y;Q6C MFK$+_(5XL'.UG0>*/,\0J_X\G3H%^T7WEB+;&W&>#1/.Q3BV'YR+B<]^7.X* M<=%H;-2"'7<:R[V6>#0]/HW=73LL8G?M^#66O7(?\JO'ZB>FL7L;A47M;92X MQIYHEE&Q]:0F[D0SIQKKSJ>?G9$O_;AGQRK/;HQS<='YVNV/8C]GW7[W"^#/ M5K<__HU0+%3K'+PT&!OND@5#>JEA`1!Q'!UO/GC=Q='Q/M*%-D-9H7S@X?%` MD-G6W.\)`"1QIA:U)\`Y.CR>E=CJH=OL"4"KS$NZD9F:5(;H;2-2S.H;YNS? M@28_2"M&)(:)<LSVGCG"&G'L48XYFMWH>D+VB'./<LSVGCE:!([KGI/5H_9M M0R_ZX%_5'WPOCS_L(BV\@G[_..HB_ZI[A7%>?V7Q=10/?:1.H@W6Y]#NIY;E M-,3@>=S$%3R?6UWC>>-1,)X_7JWB923RJJ>3N3K[T6#+PW`3T-H+XK`HG\&* M<V3DPO*E]@ZQZB8:W'$0UQQVUG6.IP_.F-'CH)!CP\=!H<2#J*HZ8"4!5@LN MQ[!;,FYC2V=HUR*O!6MI%4Q-G]%*=1GEZ#=4A:^/]P6%>)<X:I)RH#SK_7.U M2,?VC=]N-,[X67=XW5QL<3`UPC;5Y1N_&R.L&^UG6FW?^#T<"8P.QXZM;-_X M)9$EGJ3RF5;;-WZW^L.>>JO?D8NU?>-W;T347K]C%VO[QF]>(.RQ\P*)QE:V M7H>$<V'-$GL2_SS<Y,5YV'ZF-WEQ'[:?K4U>X@K;0XL]B!F5J99M8&@[?6*P MX?2)G6I=A-,G9S\0N>!8+;@X7!"O[AD3JCAB<R&-V"!XQ%XHB:,DCHT-1TGB M-;98RA?INI;&O_(]]CRKU60_FV<MU`K&(a)/.LV-+$T](DL.M$PAM6Q$85F9SU MII9UR.AA231'2<8ZF=B1&?:<XK:7BZSG%!MF)<13MW2."^?4K=@PY]F`KE`= M(DZF`[<E>6PP8>=;SS8!P37#`B`B)F>S38"[<(FXYO#:<.@9:TP?W>?ZBIW& M7N>=B&^2PFS?N.A$/!<(LSWGLA/QV]0PZV_C>&.$&6+H`:RL&5^[4$5OC&RH M%_:^&^HEUBY4\1LCFYN&XZC-39VV"_K#LE.OJAP,2F#Y#@X&7;B#06Y\76S_ M'?.MWS/6ICA!U96QTHS:3WGOB_0O]H&5`N[S9^\0;$XO2:!0.%N$?9[S<[;( M7BMMWA=.LV$Z^R^^="&&(>.++T"06>\OOA=?[#37\2$I"5J-RQ>,W:8'WNHG M-JQZ>E6]TCT!UCBF7(V8UR%+%31RX0`=HVQR3AJYL_W"4>B!1['&;.3.F@&Y MV%71)BIEB&N)*,()8C\3-JC8OZH&%:OB38E$H`=FT44E%JXQGFD=BV98V[?4 M=K4E\SQK3W&>Q_#_WA";M?^/8G.^VC**U=UJRRHXJRQ*1'RK+:OPK+(H48FM MMCP79Y55T1)ZFQB!W1C[J)[KA/"6^-=Q0:&AQM9,>%+?^:#]Y_-QJ!%]#&F_ M><J]JL<%&D2S7DWI`DJ'TXT]UGZ4,F?JS$?#&^>%TK.Q@RB6*AAK-U+$7L"% M-]9NAX2TEK'%U:FW2M':Q%P=TQI[?/<\?-'$^?CNF7[1Q/WX[MEZT>0,K6DY M^Y-)+@\FK)+I$^>L49N@Q9FJRY7-[$9<J&86@9,V;B*Z>%^=C3W(91,A(JH* M";@VNQA"UX3=0G@F;+?@8#^D'/JD]^D&F^-QN!\2-'(GOD7G*_!F5.S0QTH* MO,[W1=)X3U#>=)WWA-?=_D@:/SN=>8/N#C;XW.^3I,E(IX9[0C?<]*3X]DNR M<`#Q.A-=%CLZ^I"HR<KV.]H_*2[+3T!?[29YXEUK427S918#I&YR&'NM/+0/ MTUKYV.9;Q4MC+HAW&G$7J@NU:^UJE-0^LHY4,YA8JY+S..QP1@6[=EYX=D05 MV&F459QS.ST?SHZ(K8\VL]EQ]DK/U_%5[)6>?TL)+;A^Q7!7D6`-R$7R.CV> MG7)^GIT2N_V(=:QD#F2*;,S&VC%C8[;0>CYBABX^ZCTN=F/LQLGFU:_*-#>* MS;GF1K%B#^6\60)HKUHVD;B=/L8[SAH;";C]N.S2Y1$NV%9@6T%<G[,52[U< M3RK8J7E5#9=%G;J<&*J\[/$O^TP-RX#0,X0+S`BP=%Q(a)%A9WCMXA.R,6[;XY MCYU87.LCSO4R[5\;8H=&<6_K8272\;IUA!DNEUCK[CU6^QJUMB9D_+BV)BP" M1S'=-&9N_+;+P(_=;K]Z-,9Z.*L7E^MK3$':CD_55P($4T*O_?;2=@<</2Q` M,QQ@&?:14:24E)-B"S&(^'`>QT)5M7PS=HMJ\YXA]B7,(K`O83=J$-?[5G:( MO?@WY@G("=B(!1<O0/O?SH+NIX*GZ8+O#\_PQPY?HEC.]_`%Q[G/R4LNL9&` MOC-]<W^4;*%6%`;<XAC5_<RHNWUC8K&2*I[-P(AKCC,6ML4<D++@XI!@GNR[ M+Z<B\>"I<#S8OI+T$BA((HA=$IM)P@1KSDIU&MAZS;X!(7=8*^L-B:S"@P9A MWQ'Q8H8DDCLLDY=46<R@RUKI#V\HM=+/RVK"RW(22.CRMB>'WP/8GJS+$S1Y M#7AY3J.+9II,;C;Q<(I^Z!R321L.VF,W2783>.C2V4Y8(;^W*RUTI)TA73!) M=^ND]138G&0HA?0:IOR;4XC'?S<(IQ)J45;6C*1BDA^O4]?3R*X9?E/C:'JD MKDRCHE4_R1"U!=:9@9T!V\0>.(5N%]W@%+J=SKL++>Q>K76UO:*FEYXH[72^ MO:*7]"%&FX2;*P)!9KV]^#97C`WL*&)'\>([=<V^DUD1F?*I\!B=S-AP%^E4 MY:C^1<3*_&#!][K.U:C*KG`5;HJ1R,I]=J/QKJ1#%IS2=NK[?M4OYEATG=Q$ MH`@>"?B2B`"(B+'8+"6(TY^>L2/?$A8<1QW8,YVA4;P<F%[D_1:=U4DB,9NL M!/*GWWBV;#B!C"8P5:O#I:*X?`$B<B,VN1$1V.2Z[V[$]E`Q%GR=D=4("0A. M#@N`B&%!>C<>WR:/B$`+2FCL\Y!I._'H<4N-@QNVI)GP$2,'=JY9Y^+&*,L" M?C_A\VVG5U%#DSFI_@`QE]A.K71N;I`RO8:_6A2W#TJ`PTF:!&YL\FBZ/\U2 M0I))2LR!R=VU*QF8M-.IJ(')#?4<#$S:J534P.3>1E4R,)F`%ZS<G5=EE'IV M_*V=6:"_17^+_A;];272C%.VXAP,C7N'8$<:[+"?FT#^]4Q$3UK:.'<O85.= M[<](">RK$5?JAY67$X$K]1TV&/J#ME/6Z*UG[0(1_7[<>M9)Y*'+N#BVGG7G M7L_.RTLQP^ZXW+E5-IV]2FCWCFLS"]T/+PXUECBZ6"*JR[-9(FK(<[505)=H MLU#4D.AZN:@NM5T@;'_M`M%2$U@T>C@EW%`<3HDL1*VB1:-Y*6&;S4LQ+1HU M2X]OT2A;%A'*?UF@DD6O\?B,9F&?$4HED!Q=_W6J8NFHGA+K"H0"6.J=HIX' M2"EQ?U/IM'Y5QF5.X":U,^`/]?58,$2DSP"C1(P2P^+.UNOT"">([:3P*(B+ M,1[G5M)&1Z<Q5B5&A:R5KTI$.$>"[7&<:[7T5>O1#J#20;,XT_LKE7"@YH,E M/%L_3Z>6//WHE)MJ\71N\B.7\70W`F'.O=G1^$VTV9^8K:":B:YCHNN:Z(<G M3L_D:455-1NK*-HO"!Q]-$2_WV;9`/[^W<E_G,S3[U\S=#A/FZNZ@8EN:*(; MF>C&)KJ)B6YJHIN9Z.8F.M-$7V*B6YOHR"L9.CJ8Z$XFVNQ5S)&Q^14-DXLG MT&,3DF:B>1_*4,]$MS#1;4QT6Q/=WD1W--'9)KJ+B>YJH@U]Z5<,]>>>$+WG MDEF7\O?OJ"&#*O(,W=F7IT]?>150^?S+RT$1OIKT41E/[YSTZE">/EFH`!,Z ML.HS8"*9A=.!B:R;NC6?IXN*?PL>V?;9*Z[AZ:8E#P.5SK@L:S9/[[YBVD*> M;N5OTI.G_T!>Z\73<]HM6<W339=O`"KZ3^\0D-[\Y7\NY.F.91MJ\/3JJ7^] MG*>'-^\'3$QJ6]J'IV]9N?XJGDZ;?B.P[WK#OAS&T[-33@!_*+;X)[!_?X,7 MBWCZ>/HG@+YC:;VY//U1RV;@^3\UDX#G4:OYUV"QT>KRW_?@Z9>G/WJE%7W* MI(\C9NNTX2^V4SJ%HS\+T;<5-@7E,7[/G@/O7QNB#?D'YT#YK>;"^\M,]/,F M.C!/IV]8D-6/<##DCY^GR_^DQ5&@_RV7IRSE:8-_=^C^R[H/`"[CCJ5YP*7\ M?O#'H'X-_N=#_.*")-"^]9PV!]C/F$6+QO&T.=J60_DW<&F'SN#^EM7'`7O; MT[35<IX>/OT[D/\#@PN6\?2?:G\"[&=!\C5`7ND5&<">WEU3!LK3N?Y_07M] M;8TU4WCZSLF3`?^;LY\>R],[FUP/FMCVI3N!_#X=TH%_:KYF/_!'=Q7\!30I MN36;`!<[9\W]@,Z>^RZPG]FESZ_CZ=^L&SR=IV^K^05P^=F-[@'Y&S]@.]#W MQV<M6,G3[7J?!DUV:OY!X-\&]6TW@Z<?G94#[N_1X''P/#[H=#.0?W/GR6MX M^KGNQ<#?/K[PH5*>_KRH#VBR.O9Z93Y/A^TQI+]#9XP'];.L70U0/\G]=];F MZ1M,^FK(:S4?^I,-)OI#$]U[@4ZO:_S/<E[>DY=>`OR5<?_VT/VW]OO-$O[W M+4FM0$AEW/]3Z/Y;IC\`0A#C]_P2_?>)"ZX$_MKX_9'0[UOR4T?QO_\W[Q,0 MWRY?]@YH+W9>^@7(7X]9JT"(>Z#69N!?<M?MF\#31OJ!A7KZ#=MXUEK]7A3Z MO=_`ZL!>UJV_"H1()5W^.HBG-XZZ#Z2_IMJ8*WCZR7H;@?YN(=-`^^GW;K?, M[[Z%\/DV603IN2':\'^OA&CS2ME^>?O!5X.N.`3\S?.^+XIYVI#?H%27U_?* MMT'\M/B:5T`(/#9M-7#!=3L6@^=_>.V[!3Q]I/L($"]]W,T+[&5RO_=`_1IT M>2@_;_0N!_5KT.^&?G^I;RX(*;]=E@_:#X/N7J;?OZG/=07\[_LZ_7X!3X^< M>3OH`JUMM@V$\'TG/PGT]:766T$72YKR#O"G62.?`5V,=5=L&\_34Y=U`?[V MM@[WS>/I-_M[KN;I3YL7@?:XW[1ZH/ZG-*@'XN%BSQS(a)[Z[._8_`TQT+;P,A M?5)&-]"%>+7_-M`EFEKKEVD\7=;_&Y#^6\D"-[-)]3BU*VC//ECX`XC7%I8N M`^4]T347Q.<9`2^0?T_R$R#>G51],F@??UCV'M#O]5<V`5V^M[Q?Y_"T-_^J M$3S=[JKE(%Z\?E4)*/_2[AM!_35;^Q<0+XSI.+"0I_<MOA[H5W;C\:`+F=I[ M6W^>+NKU*>@/EQ0%P>_"_.)6X/[\Y4!^U_))0%_*VV7-XNEK>\\'^OFOE-M! M_7O2;@'Z7G[9(>"_F_4]D,G3C\WH"?2M5<Z;H+V5IQP`[6'OHB&@"SRWU2'0 M7I\H&`KR_W&]^X`]=V^PLH"GOUZ_`:3_2J-#H#^S8'5UX,_;=LH$[?M[_>H# M_W]#40;H;ZE%/X'^X2<#+RW@Z?^;6'LT3_]O^4^`?T#C,N#_ZZU=!N+K+@O: M@GA\\9+NP%YWS>X-].7@PL[@^4SLU1$\W[;E<T![?N<UW4?R=,K0,:#]V3[^ M\C$\/7/1;/"\/RM.!?*7=LH!7?;:0S-7\'3GC!]!_A8$O@3YET8>`O'B=1/> M!_U!@ZXH@^W?NR%ZV[@*T)\T?L]>K/^^<O(+P%Z,W[<LAO*.F>B!2R"](T2# MP6H*\#"Y^T^%[D\W_0Z<(04(CBE`<$@!@B,*8.P4_QN\`L07*XN>!D,\SW?: M"IY_K:O_`.RQ^94]07RXV]\3Z._#,]^?Q-/;EP\'[=4ELUX'[471U9M`^W:X M?0YHGZXK&07BA>R&7J!_'>OWR>'IU,&/`O]_;=^>P!]>-B4=I+<JY0M@3X=+ MQX$AJ_:+[@7MQ;TUE@/_<O/BKT'_HD?!!!#O++A\"O`G';N5@WAR8NM:(%[Z MDV\]J&^C_Y"W5.]_WUM[),AOC=[W`GNX?\`BX!\_G?HEH+^>VQ@\GW[EOP/Q M^1\ZG<KDZ0[COP;ZT2-%`O%*AXEW@/&;9_(^!2I^WZ)"X"^'+;D/^(-B_TA0 MGSU[?(a)WZ8\498\!XRKZBU\#O+PS9-).GIPQ40'GK+_@&M,=+Z^T")C4^LSMX M?D<;+@?^YE7_###>D]'S>5#_@]?_#;1/7_D>`?JWMDXIB`=7=6T!VI,I7=-! M_'%\X"T@O579FX'^/7O)%X-Y>M?(WX#Z'-5R%*B/SOZ[`?\C:?Y+K.@[ENK^ MY\;+;P+UFS]R(W`Q?5IV`/'V/V?<#?J_/ZPOF\/3T\=^!NI_Y()'%_/TP+43 M@?Y=4F<WB)?>F_0"H.M-&Y[+TRN]-8`]K6L!R]OKTJ]`?H7^<T'_Z\[.:<!E M_KEV#=">/C3D!5`?ZT?\`NSGC],(\*?;JCT&]']G]F'0__@P^\@0GBZ\[!;0 MWNP+7`7:SVDC'@7C+Y_5(;UY^G=9*T!\88P_[0CYBYT-_@3L=TV].2#>*^C] M!1B_^D/U=.!_[V@]`\1[O7+30'R\<XX*Y%4K^@[TMX^5]@/Y:]YU,!@_6YVR M<SU/KUO[`!A_/K[P;>!?&C0507UF+V@/QE>>G-(1Z-<UUZ2"^'KTN&G`?J^^ M[#'0'WNHO`3$$\J8AA-Y>D+M6\'SJCN_1QY/KZSV"$BO^?I<H"]/^,9UY^GG M5VX#_._YR4">OGK9KJD\_=WE=X#^>5JS9T"\=+04QLN7I=\*_,&1OM-`?7]? M>(@G29IO!XB7^R[;`.+;C[+^#/S5K6,#H'\VK\8;H#\P>.@`,+[8?<+]8#SS MV9IUP?TO-R\$_:TC_@4@'NDSX@K0_JRJ?AKHW_A1WP']6KML)]#?I6,?`_7S MY/QO`/_>)1.!_@_*W@)"KMN*.H#V_;Y1`T"(=:+[+#"_D98U!>AKZ1`5V/^5 M27M`/*H6OP+TL^/T5X&^'>^X9A%/O]:^#/1?JF>V!_*>'_@N\&\MVOP"["O8 M^AW0'OP\>!"84FSA[0/ZHVD%)<`?_:O54R#^^W#@GX$^7]Z_,:B_WQ0$0'_A M:+UA.3Q]V8AVP'[FECS`D[3__!08#WYDP8N@/F]8.@Z4_[^='@3/Z_:V#X/V M;&9#+Z";-GD:U/=[E[U4P-/_:[(4M,_?-?\+B%^>&G</T+=7QTX!^G"P]Q[8 MW[]D(NA?G9IZ)VCO#DQ;NXJGFX^_!G01:HR^#CR_NLW+0/LW.W,?T,]J0]J` M^GN@\]U@O*;7B*=`O/QQ7AG0Q_YKK@/^H4O2[>!Y?]&R!\C/<X.G`O]4.[`: M^%?O'!_P+Y_V:`/B[R=2/P3Y:4*&0_ZF!T![NVGR(R"_2Q>N`O'@B]XLT)Y? MN;83B!\F]-T/QI<>:747R,_OZG\`TNN9+P!_4ZTA[)$%?$>!_T^>_`;PY^^- MJ0?TM;#Q6)#?7>T_`_'BYJ9U@+]=F?H^L,\//9^"]KG;LN;`7QO^ZSB-#UC' M2FK=$=AG=LE\T/XL[_)[\'S6+'X!R'NB^\V@?Y,T^1CHWV<WNAWXTU-K'@;Z M,[UX"-#W35<W`OPMEJ>#"LT>^0"(SWSSUP!]KMWN&*CO.7E'`'^7?!^(GS]> M?P+XK\=:_17HUVU=ZH+^QZ1&74#_=TC61'#_1[TK@'U.R'T:].<"[>L`?1': M_@OT-^Y*:@CZRU<,/0[:[_M;#@?Q5/]V?P7RZEWU%6@?[ZQS`NC+'P;Y@?Y] M'?@]:`]KS?L!]"<'%ST%XL,W!R>!]N&EZ:/`_&;RNL:P/5^V"[3WK2_;`NK_ M'PT>!_5S4X->(%Y?.68;B*\_2GD0U._<!C>`_L/\:05`_Z>,/`+Z)YTFU0// M<_3TIT'[\'[CCJ#\=?M^`.8??E?_.>`?7DD:`_J3G=JT`./Q_UF?#^K_6,$; MH+\YO<9!H)\?+7T=E.\_Y<-`_/"G%2N`/ZTU*AW43[U^+<'\=>KLCL`^7BY[ M#+3?.=ECP//TMIX/GO?\A<^!(:S>ZU\!_8F!9!F(CPTZ4*[W/PQZ?(B>/&@H M>+X#^F2"^<<M5\'R?'(-[-^UGMP(Z,OB-5M!>[=U[2#0__RE^[W`WR]-OQ., M%S]#4D!][%J0`I[_]3/7`?W*J-XYDZ>-\NTPE?>XB>Z]3*>_'34)Q,.MQOT" M_&>+B2^"^GYSZ&D0W]V\;C70YYU7'P/CO3<U2P+C#2UJ?`[&S_[>X2LP?[;I MDJ]!>8_U>@3XH^]:I0#_WW#P+V!\VBA?1:A\*T;>#N(-<5`-T'\S[C^T#-9/ MD^4ZW:GK6C`?]YO1QX!];C3-?__&1%]KHC>8Z,TF>KV)OL9$KS/1:TWT-U-R M0/LU>T(_8-_;1K\&_.NF]4?`\]DVX3;0_B\:,A#T=PS_.C=4/Z.JWP7:QZ&+ M/:"^M(a)2F`O_^ZF(8;SS7%-I#]7FC0+Q1/OD.$&\UNNI]$.^.6_%;$+]_U7H# MB&^+IRT"Y>_=>`+0E^+EZ6`^[IUEKX'^L)H`MA;M%S)):/5O#@%H3+],U?:# M^\:;4?&Y8EKUR]B/QV"?&V;OK>V"9L%^OB"O.'8Q:FK%8)O5;U=RAZ7!-SXZ M:T79%4-$D["(5E3$9W8BSF<$9NO%TPPQ!TS'VNP;R+@FA[@8S)72R6=PG?!G M5+PKL]7\C,>8T6%26)C`:"/<-II](QPVPFJC^37<`&L>M%P3*(N%)?'*8F#> MA,EC>4I4'FLBJDK6NW.<6'%CP<:*S5E)I,KM9+'/1&6QQYBH+-:R\]4=CRR& M@GG[A0)R6R6>3W<9,6R=92>=$VO`Z)"8OS_?\>X\IHEIMFY0KY8F`G.#H%J. M6KK`[/FLEN>Y5^R/-7%;*'OF+47N<Z.S']-2+ZZ$O6DT^PR-?>`">ZM\*-PD M/$S9)]$FH19L$JX']7%_#%$Q<P+;65+B1(Q%?43$3"MA=5)J6RA=Q$-:H6Q$ M['64"[&RPEBASD)[]8O.'7>`2.>PI94MC-.7SG#4B+^AB>_F_DGJ"M%NT7YA MAW3#`+\%N[D=-DI53GF.JGK+:N^J=!Z+^CY$V0L4K_[:2XX0Q>[EV'L$M!E1 MJX*W*J4Y[[FI$H>I;SP71ZBXH31V@`(+:=IFPA!RM"J$,.24193(>;B$.#O8 M6A9YS(+CIW.TS-XQ1"LP>.O02+C?8N8\?=[8OF6%QR9.8]B^V(DQ6_2X=/?T MDR-V"U_006//7V+/WL1CL+,=&#X[;9'Z(Y2]8&*R915FAE/?3,M_-)9O#2RE MN2`^[8`ELYA#[,OPD\@^S;OV<PUS]X%]\C$H^ZP3^JP;^JP7^JP?^N1CT/,- M1"N?<$[RQ^HF7;AXZQ9Q=N$C<.C`&]*GRCE_G1!"]954!?;G(>?63QIEJ8KT MC0Y^5<@R=/)<U4O2&4C?7X4RC>&^<U4_9J2>\_R4@XD3!`*!N#BAYN3E9XXH M7%J\9%AQ>6;WD9GY)<N6SAA8N&+1L-*YBY:6+B+3IHU;439K6MGR\LY%\]EW M/4A>?G[NB!XCQQ$+7JOO@#S:!?7=07*(=P7Q&_N,,:^=(DSV,'IRU&F^[L&V M/>.=N$#3&SENPL"`URK'/#X+9:B?]NDC1=I&-!=CF<@&(Y^T?+OF>B+YCL!C M+!&:UY*0'6R(($??N6=(O:TL.QZ_-\F7Y/'ZS+M=&*+&DWED(2DF2TDF&44_ M5]#/L:24?E=(V/+B'E2.AR0E"1XAV>\);QK'KTS:P/X;1U91GIF4DTWJ=VNC MI9[J]WD8;%._@BRAZ1=J/'UJ;07EL^/))4-H'B/YIFDS3FYM5N5EK;S<0UI0 MF=QZ%SN9>92WA"RG_Y=3B;,H-RS5+)8W;IVHTU)EDG9D$NFB[8)465[[^K5G M9)(<_8PF43F+R!RJ7OHGJX<!/;0GE>Q-\7B2/#[;)Y5+TUNFE:N8_F_D@Y`6 M`C4<7Z#!Q#0V1*4-,Q+YX;PV[,JJR+U.-X@M]*9O&S)I;$F3CVPD_-)S7;&- ME"[Q3DPC%001%Q[P[TO>0_;0!UXKFZE`0\(6*;*-3MX'J[GB@_XT5=681N]. M]3^+="-=Z95)1E(=F4DUI(1>Y=0?9)(RJBNE].]2:AG,,QC?E-%KJ68AF92C MD/Y=3/.92;]=!.044VHHR0_O_=6,_EU(]9#Q#:<IC]7XBFA:G32=G$<6D+D: M9R89J%G,7)H6LY(B2E4W<6=R_!<T%)4:5;7H3A$SJ:.;_O"#.'IN^E]N#9". M;9_\-ULCECTLTE/;0?3'NHOHSND8T;>"^XGHHQYL]H[YUG:"OL5?=T'?*B1/ MT!N:`D'O_94(^A9_FP5]J[@[!'V+N!V"_NKS(X+^RO->^LG>N]LGZ![JH*#G MY9"@]R8/"WI^/O/I6T[U8&EE=>O:+7-DZ<QY)?/*5V66+2DM+YU56L+^*"M= M6EB2.9.VKD69I8OT>XHSA^93C<DO;!8J(Z'\[&]F`?F%RTHRAV>-+2XJ6M4I M<]2\!7/GE60.+%XZMZQP41'UY6$>>I_Y;U:N4:5+%A:69!654@UGW]6(R,S4 MA))1\]CWK`Q=B<['RC=RWJPEI4M+9Y=G3BI=4I39.RN;E.20"+2_2UH=W'G+ M"X+V]Y)U1W:F[@-_>T/Y89_,I;)/YE8-I\W>?\G/BXS0!$BUL:>7WC=)8$_* M6[\5RXFVX:S6/V_CKTEKF-WKTVB6KQJF?#(-2-44B_V5YIG=N+`V*XULH6RY MI<N6S"M>DCFJF#4,3(<Z"_JG1Q"TD0#9LR:T9/)YC=^70]+'SUM8O)3Q9(XM M75BX*,PI:"FFA-.NXUGBR:G)?LNBO[%QA*[5V#>!T,:%A;5G-[XB/[=SES[9 M70O(&%)`74)NZ)_.9<B91]+]>F[N">5FJS:]&RLWGG`Y6+TVTRRAGH_5!@G= MTU;01[(B98U()Y5(]UI*9=;!GE$][1G1W-.Z:$UO;$F9QB>1$`]?MFKAVHK4 MF^!AS$SR*8^@9RB4JD^S7\FC!P[LKA:A7Y*T3Z\0T/AJT1QU]A36;N+1M6%T ML/[/NC9\Y;W3:_ZK_JD[O8S;3^M!3U67;:3JMTTUF1A\>CJ;D^O_S*RK%95\ M0_"HD'F*72=/WQ!DUU%!ET/",.0$PBD9Z>O0OTV)2FV4ED9KLEO<IK#K;?%Z MY7MQE5);6J1TDXJ5<=)T9;$T6=DJC5<>E<8JK]/K&/W;(T]6&LM7*]WD8F6H M7*I<+:]1RN3-RGQYJS)9WJ@,DE<IE]+O&\IS%2+/5+Z0KJ:\TZB,:536U<I2 M::8R7IJC])!*E;K22N5G<:/RGGBKLD?<H5TL+^82V95AF%:&EF1[4%:W!S]7 M[PJ^J>X(/JT^&+Q7?2*X27TA6*+^O^`$]:/@Y>H/P39J0*RA9HI!I8_XJ3)> M_'_*8O$YY19QEY8F2_MWXFYEA?BR<I7XOC)`_$9I(Q(U7:RG2L%.ZI?!7/7] MX%3U0'")^GSP!O7QX$[UX>!>]?[@6^H?:=I_U/(@JT[SGJ7EO2Z1Y9DJNT[+ M<U157J`*I\I4[ZER->G4*M5_:IT:.+5133FUF?YVHZK(F]13\K7T_G6J**]4 M3\KEZL]RJ?HCY?M>GJU=3);3/!@ZX)7ZJNSJ*+57QTH-U=52=76GY%4/2K+R MG712J27_K/20?U$FR9*R7!;4.^1JZJ-R/76_W%K]0.ZI?BL/U])EZ7\NCU'_ M*?=7]\F=U8?E9NI6N8Y:+J>J$V2?VD,F:BU9I3(%]34I2?T336>M5%^=(+54 M+Y6ZJ<G2(.UB>7%;AI^"Q[3KZ^!;Y//(a)4^33X._)%\%UY'BPD/P2S"5>L2VI M*U8C[<1OU0'BV^H4\2EUA7B7NDW<H.X5YZH?BF.T=%GZGXA3U+^+9>H]XF9U MG?A'M5#\/W6(^"\U6_Q1;2JFD^HB;6C%(53N+'(BN)Y\2]/Z7_!)FOZ;1`E^ MIETL+T[+T$$K0VU24ZPAI(G5A(#H%WRB(`CB*7(Z>))(P1_(R>"WFDPF6PU^ M23SB<9(D?D]2Q%](FBB3=)$(=42?4%\,"$R.6SW\5M:O'^3FY&>Y(PG*O8@D M#Z37&'I=3>D2^OTJ\HV\@GPGSR<GZ'<GZ&\GZ#W?R7WH]YW),;DE^51NJ%U, MEMMG6"`%57:ME[Y2[Y..J'^7_JD>D5Y7)>D?:D/Y1:H_^]1\>BV07U8WR@?5 MN^6WU,?D?ZLORY^K[\H_J!_3!IJER]+_MQP(a);\B2^G_RM^I?Y(_5.^E]U\KO MJ//D-ZF,UZBLUZC,U]53TEOJ1])[ZCZ:WA^E+]6-TH]J$76P[&)Y<5N&WXI# M!'9M%+L)Z\7FPC5BBK!!_)%L$H^0F\1_D.WB8^0!<3MY2KR6'!`7D`_$">1' M\0J2*G4F;:3&Y$HI14N7I9\G99!.4@N2(?4@HCB4?"I.)V^)B\G_B=>3OXKW MD1WBD^1.\55RB_@1N5'\F6RF:=T@-A5N$KL(OQ-SM8OEQ:T>_EN\47A3W"2\ M*%XC/"6N$/XBE@H[Q=G"[\5IPFWB.$TFDWV7.))^/U%X2"P0'A?G",_1^_XA M+A?>%M<(1\0-`I/C5@_GR/NUJT3^%UD@?TUFRZ?)5#E=&"FW$/K*/80V\B"A MNCQ>2)&'"9GRY4(/N;,P2&XBC).K"=-DD13*Q\A,^=]DNOR:=C%9;O/PDO2L M=GT@/4I^EOY,JLM_)*WHU4=^@(R0'R'3Y&<TN4S^>/D%DB<_37K)C]%['J'W M/D).2KO)1])>LE]Z6;N8++=YF"\-$MA5+'6GUR7T[QK",DDFOY&^(K^3_D4> MD@YJ<IG\QZ1#Y&Z)ZH#T/[).$H1RJ990*K6D5U?Z]Q7:Q62YS0.K9W;]),T0 M_B/-%EZ4Y@E_H=<V^G>%=#65.UZ3R^2OE48(FZ6)PG9I&KUG!KUWAG"8_GU" MFJ`]*W8Q66[S<%)J)+`K66XJU)<SA794#RZ36PG#Y#;"5?2:1Z_EE%XI9]&_ ML^EWV?2W;*$WI=O(G82Z<@<A26XK_"RUTBXFRVD>#)MN+YY4V-5.;*MV$B>H M/<2-ZI7B8^HX\0-UKN@COZ'MP0YQ#'E1+">?B=M(FK2']*3/I%#ZDMPJG2(O M2]6U=%GZ+TKUA9LDOS!=^IYD2X>)7WJ>_)?:\K/B6NH;)I'58A=20-NIH>)' M:F_Q2;6CN$EM(4Y2F]&TF],\-`_EQ6D9#)N^7+Q=Z2WN5+J+?U6ZB']3LL0# M2B?Q7:6C^)'2@<8Z[4.R,^G?E]#O6HCO**W$?RAMQ;WT][_0^^]5NHFW*$R. MVV<H2RT$=OTL90O?20.$SZ51PH=4?_XI+11>DU8+^Z3KA&>D6X7'I!N$O=(& MX26I7'A=FB.\*TT6CDC#A,^DOL(W4B?A>ZF9=C%93O-PLY:'062"-%5AUSQI MKK)66J'<+%VG[)"V*8]+NY1]TM^4MZ0WE`^ECY2OI!^5GZ445:"Q2ZK<3:TM M#U4;RP5J"QIOM9=O4;/D!]6N\@NT_?BWVHNV.;WE:M3V6Y*><C_279Y`NLAE MI)-\`VDK_XEDRL^3AM2/I<LGB%].T?+-\G]*2A+\\G%24_XGJ2?_C325=Y(6 M\O649Q'I0&5THK(Z4YE95'9GFD9'FE8[VOZUDA]2F\NWJ@WDU6J&7*2FR"-I M/GNI)Z46ZK<TEOI8$I5WI<^5`]([REYIG_*0]*BR7;I/V2S=K"R7UBNSI#(E M7YJE7:PNW.K0#V)G19"Z*#6EGDH3J9_27KJ2QO%Y2HXT6ADN3=!D,ME#:-^A MGS1)Z4+[#*VE$4H#:;"2*N4HI\6^R@FQE\+DN+7!GZ2F`KOJR&E"-SE(\N5/ M:1OQ)JF@OO<>^3[RN+R)O"(O(._)^>03&KM\)S<F)VD\H,A?JIY3;]"8^@D: M6V]74T]M4ZO3O]/H=X%37]'O?#2V:4Z.RWW)$7DBC1D6DV?EWY+[Y?O)%OGO M9(G\'IDH?T-]/Q$:R76$H-16NUA>G);!Z*?4HW%=>[&!D".V$R:)/83%XA7" M%G&8\+`X1C@H3A*^%*<(`>DJH3W5^V'4MRZ@=G(S]:^/29<)[TB=M319VN]+ MO80GI8'"5FFTL)#>.X+:4GMI)N4MHC)F"?O%0F&7.%6X41PO+!2'"^/$`<)E M8K;04LP4:M#8D.7#;?T?"OY)NSX/ODKCSV])$[$ZE=E!F"8.I''.5<(#8@F- M&=8(/XG7"\VDFX4ATFW"8FK3=TM;A'](&ZG=+Z/^>Y;06%XH_""M%PY(-PGW M2MOH/7?3>^\1FDC;A1_$6X57:=RQD\80:\29PD1QA-!=[$KCVKK"R>"/Y`B- M[?<'']0NEA>W9;A5O$IAUP*Q5!DBWD!]VY^5:N(KRO^"7RC_#J;0OMVEZM[@ M)/61X'KU@>!#ZL[@.[1?=YKV)]N0W<%\\O?@:BU=EOZ>X#7DS\&QY-Y@.]K_ M/*W>2>^]._@@O7^=^FAPG/I_P8[JV\$DVF\\JGC$OU/?>I]RA7BM4BS.46X2 M\[6+Y<5M&0[+C55VO4S[9`_);RM;Y,>51?(=R@C:[VXOSU*\\ACJQZY4]E`; MO5&Z5"F4.BJ]I$Y*#6JWGXJ7*7\3\[1T6?I/B_V5C\5L)4UJ2^]I0>^]1-DB MM5:>IO<?E7HKR?(0I;,\51DK+U;*Y2W*'?)#RN/RZ\KK\O^43^3Z*KM87MR6 M@>D!NRZ71PF%<G?A6KF^\`<:NSU+?>5;\A[RL7PS[5?,)J?D_L1[JAY)/G6< M]H5?TNPW*"]4OY8':NFR]'^2<ZA]+]#L.IW>DT'OK7ZJ`?&?RJ']EOFT7W(; M^9>\E[PH'R$/RBJY06XNS)'[";GR1$T7V<7R8BZ#,8JR?CW;+TK_MEI4B=BX M8BOOM627N(7<)_Z6W$&O&^EU+;U6TFL1O8KH;U>)-Y%Y8BEMW<?1'D(O>F\3 M\@2-'/:+WZ@?B.^HWXO/T9[G_2JU'+6[M$H=+A73GM`8=8747[U)ZD1[)HW4 M;5)M]3HI35TJ^>AOIY51M&=^N?2#<JGTG=)&.JY<*;VO7"V]I*RA7O_WTCWT M*?Z6M(a)9KE1^DA4J&/$.Y5!ZKC)0'*25R'^5&.8MJ3TOE@-Q`^5RNKGA/>943 M<IIR1*ZGO"8W5YZ6VRM_DKLKM\C]E37R4&6N/%X93V5<*2^@<E8H3>6-2II\ MBR+3=([3%O2(])3R/^E9A<A_5]+E5Y1,^35ZW]O*`/G?RFCY$V6&?%P9*(M* M.SF@5J=/[0>IG?HOJ8_ZO#1"W2G-4*^7EJ@+I>O5\=+=:E]IM]I2VJ^F2A^H M/XG?J_\5D\E!L3&-DIJ0[\6FQ",U([6D3'()[1EUE5J3P5)[,E7*HCVG`:2] M-(;VEHII;VDU^42\E;PN/D1[6B^1/X@?TN=RDI2+M849M'<T5!Q-/<LBH9EX M"^UM[Q:^#[XC_!(\*)P*/D][WH_07OA=U$M>)UPBE@F=J5>^C'KJ(6)[8;R8 M+LRD<LIHCVN]>)#VZ)X@=]'HC3U_*RLX$=89?Y3N--"LH1I9)N:H<\11],I5 M-VK7*'69.$W[W$@_EVE7CHV51<MMILFM3L:(4U1V31"'JS.HW/E4QC+MFJ8N MHM<"<;JZ4)Q*KRG:Q>YUFD8M+8UDFL;PT#4E)&<X61CZSJFL5IJL=&H?$\D* ML8CR%Y%"<2:9*$[1Y#!Y);3WNT#,(Z7T6D&O3=J][%/_F_$Z3<^H]Q6TYSU/ MG*7Q;M*N6?2[Q=KG)OJY0KLFN):;1/E/!B=HO$S&R>!B\E5P5NA3_YO=XU;N MR6`1Y9NH\3(97P4GTN_RM,^OZ.=)[2IR+-?0DZ:T?MG5@-9W=5KO`JU_)H?) M4^GEI77LH?4OT'N$T+UN]:0IU0W]&AZ2,T450M^YU9-D:A=BD-H#U>6:5*\; M4OUN&I(G4'WV4KTF5+_%X#3UF^"HT*?^-^-U6^]B<`SER]5XF8QO@KE49@_M M\QOZ*6K7&-=RF;TO$\=HO$S&,K&'9O_ZI_XWN\=:;G3KVE.3VY#Z$R9CE%I" MKV(Q7RT4QU'[GT*OX=2NQI#95"\7:?;#+J:7[)I("*4#]+?J])Z:]-Z:E*<Z MY4VA,CQ:G;/K_[-W+O!15/<>_^^2%WE`0"")&W"C8$0V(=DD$)(BA`1((A)" M$@($`GDL(1(2V&QXB`J^%5$K*FJ+M47[D%9LJT6K^"JHU8I6BK2VUOJXVENQ MEFNOU9W13G__<\[L;!*6[.+M_7SN_63R^6;._N><_YPY<\[_/&;F'-8=*DZA MKW6Y5F)4:U6@S.@05!DKM.5BWZ$M$V[V$ZY>,^\6:0T&,UU;8,S6RHTJK53H M87V+4*_4:"M1IS0:Z)D(V&^D<2_72G&>`G$NJ:<`LDPEE\?8':Y>LTRL01BF M7)`)MTGXNO*%KE3:KCGH+BV)[M32:!?JQ5LU%^W0\NE:K8@NTV:23YNESI=) MZ[1S:3WP:>-I`_QN1GUZ"<)OT4;3I="Q73,,WE^*_7:!(^SXF&EV/<+O0KCM M`(a)-M%\.X7\AX+]WL)U*]N[0ZA&L18>\7M$"V7NSOQWZ7('*;_1C*P5[D?0[+ M.O;";N\6,MY+]V.G8;/WHMWRF#])A-TM2(+,,'B_&_N]@LC3]P<(OQ_A]@H, M8S\X)&2\EV[V$ZY><Z3FF+]4<-@_BP[Z9](3_B*TG_+I(;\+YSF;'O"G";VL M_SG_:+1?'?2"?QR]B&.'_>/I5?^Y]!HXYL]4A)^7S6N+%:/=I2HNF7!GTG%_ M@9#'(@_S,7:'J]<L;[$HNY("H>\XZ@HF-@*;,)YD'73"7V6\YR\Q/H$M_,)? M;D3!_K`>UO>1O]'XV+_2^-"_PG@'==%1^)7[9<+-82--DW?]"Q&N3(0]*B@S MWO5/$?NCV+\K6!BQ7K;'C=I"$99U-&I3A'V6>^EF/R?7&[J'URAT5,'^5AOS MM%JC0JLWRI$^Y4C[-<`+V[11RX-MFDQ7:)/H&K3CKX<=ND%+@;U@.\%P^6,X M7R?1H_X4E*-QR),3Z&G_).3!R<AW><AS^<@GG&>8!N-S?[WQJ;\6]Z`:UU0% M."ZAKB%T69BAYQE,C9YOK-8+C2WZ#.,6O<+XMEYO[-/;C$?U+<83^D[C7G"/ M?I6Q2]]H[-#7&%OU)F.]OL18J=<:5?H"Z*A2Y$4<AV>TG09S2-MB'-96&6]J MBXR/M0K#CKB,09PF(6XRGE7&9)QK',Z9B'-_H37!W\7&^]H&XRWM*H3;*6!= MX<;!+#?/:)6*G4I/);VI9.'J&D^RW.S3%M,!$;9"(?6POM_AV*^U)GH>_;<# MVBIZ";_W"2I%&'9'FK\/P);O0S[BL"\)6B"K4W)YC-V1ZIVJM]!$?;T(R_HF MZG7D@&RJ8+TXQNY(]4[4*Q%NL0CK$"R&;)62RV/L#E>OV4Z:@#`FX\6^0NAA MW6?I'AJEKZ0$?1$-A6RH\A=I/IF@WVU(*I6>NXVA2A:N+C/?S].W&LQL?9MQ M(a)7Z%,46_VLC1KT-^O]$X3[\%.F\7>EE_@K[#2(:?%)2_L_4NPZ5?#/]M1HG> M"ATMBJT1Q\$+W9+;C2Z<LT/?;K0A#LV(RU*<KP9QD_%L,>IQKF:<LP/GWH@X M;-,W&=?!SP[$[6;HN%GIBC1-O4A'R=U*3R7=K&3AZAI/LNPU(.^T(]RZ@$X) MZ]N.8]OT9O(A/[2#C?C=(*@48=@=[OG,O-R.,M"`/,QA-PK0UT8YD7)YC-V1 MZCV"<(<0KEU0!W<=[8?LB&"].';D-,K>(5SG$<23P^X7+(9LE9++8X=.H^P] M@3`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`J56?GNK-R\R?G3G$N M;%JWKLOK\_1XNYT7>3R^]LXVEW,!)!W(0"[G?)4&&4$I(=.I$'^<6AF@4*1: M&C5BSY)T<8Q3=+KRDZY"L*Q_2IKQCDGL?7=9$GQWJ[R>MJY.E[.NIL3E=.=F MN0N<)3UM/=T^ISLG)V>^.+M;G$G>S4*0UNL*I)1]9$*2IG)`J+L;R1UC>\NV M9&Q,!IV`K8"UL!]3SYPLBQR9CWOLAOW-`75LCCJ5CZHDP[[UE#Y8QRLCOHJ/ M^?88BAYRJFN1/DX54^FC]UGZ6\X#]H'>:$]4]_#6Z#R[F9OX,R%W/$NX3CQF MJXG>'#4LIJRKI6>MI]-7!)O(5JX$EC!?V+V^%BU4:1\6LSFJ)EI>=W#)E:'F MQ%0.,6,P2L2`)3$B!G$(>3"FMMW7X2F:*ZQJAK"SX9WW($+'Q?0_[\G\;HK= M'/5D"+]6:MT9VQ!(+1E7ELBX>J%A5%Q-5=W"TMDE2*M:FB_^^L>UK\4:Z%XE MJ;![8K,#%BM)G)TEMJ"S%R%]PDN;47&;H[RQH:[7O#=W1%LV4GYSP9*X(%L3 MWC<7EJ6([$L*&<:LY9?&6+&Y6,2&)3."8M/G2Z$:CW=#>PO2M*<;!MM9,;MV MSLYNZR,A)VRWLQ/)[=GD\W2V>EI%K5^L:N[5JKZ7?V9=GAEH`V0+O]EDUO3< M:I@+/YGBB!4Z7QPKP?]\Y9-+4:WZ;;4>BL.\<Z'22H;Z("B-9'EFR="@-#(_ MG^H6;9)B$?]\U7(YO1CPERJZ_<F^K39GB;>]J8/,7#PL<.=[W]4K8ZT8GR%B MS)+HH!A7S*^=/0]QK14I/#=DWAIVTEA*7]?&^J/-LPP59V$)1S67)O316*SN MA55G)82=:_O7":;E"UUK#.3C'GM-]$#U&]OJ4_FH2JJ)'JA^JXD>J'X;R,=] M]E%Q`UTM6YY3QW147.^8GJQE:.[E'8L?^H;-O+N\RH`[GB7Q07FH9%9-[<*2 MTMHB9SFL4R%:7*7$KNEH<_5M:UV6HU]HZNT*Y,U+E-ZNJ,H@O;6KV[N=+5V= M/F][<X]H0LNO$F%T?%W.IM969XBO%[NMKQ:E79HN6H[</G2I5F2&:DD&M[,S M5$LL4[4ETX7<%6BUF6VU-%H2")NN?,G0['8%SB5;\BZRVO<92)/I2*'^]7MZ MQ^E8I"N#[DVI2$.63`A*P]`&V4JM[*#K<8E8%@9=-U]%N?J=KJ3F-?6_YJ]V M5?]NO[U3[[G15NI-%*G'DE%!J;?B))O97A_H[_120K9?^MG[D.V7X7U*\?!3 MV.F.,=;U2CO-$FFGS^X37[>ZEY;M-7O276/T@!;9;F%)<+O%68HRV]3B2T.> M=*M<X\+_KYXWK%BDI5C7,DS$@B7VH%CD9KM/FA]EC=KO"](^-6JR"G%&ROA` MJU2V"UEB"[Y:<H:XLN#8E@6T-*K8EMFG!FGI]9VSRUG1Z?-T(!V]Z!F+X1.7 MLW).7E;>U*G<Z\S-=<[/GIWM=!>(?J];V%G9SRQ5)53V?3-4ZJ>3O!?2PF4( M27K`5Z;HCYK6S^RM2C=KG1Z0];].F5^?#SN_CE!A$U+R$WK?/Y8$WS_?ZF2T MXP9.V::`GIDJ99L2S@F^/R4;/)T]'I>SO+VCH[NYR]OE<E8M=.YR3IN:Z\X7 M/7DRQUVD;9,]=VG;,U6*F:,SI2(EK!0RY=S/_Y_,W[>F=/7),2P)SC&UGHXB MYZ1<YWD%.7D3G>XI^5G3"J=-<3GG-&WJ(R_,+<!USE[;U(X0N)OR[J>I$1+S MOEMC$:9$7IN5P\+S/YWFJS/TK_VM'!-4_FRGSC$C^Z5,<^".NU3*-">,"4J9 M=2A+V6NRO5R49K9S2<INZ5KK#ER'O`JSG"P3)<$L)2SMVZ^T<MU(,0^#/",/ MACF=/QJ2%HC?Z=_MXM2^UHPEP:7!'<*:23VFM2I([6NM6!*IM2I.M:S5&!6; M,GMLD!8UMX,H)6YRA[S7IK[E@5&(Q4K?\B%Y0?H"DT2$L'SN+#</I?:R?.6B M=25+K%5N__T63Z:5:<>R4I_J8\=8$KD=*TX]'M!3I-+H>,)9X=LQ^FHV[.3Q MBZPU^G"J9;-N%M?`DOJ@:X"M=?:V6].FYF05%N0/;+<Z19;+7IO=:N85JV@' MQEG_=VV;.1HM?\EQZKYY,GSK$EX/V`S%1]AUAGV5PW(UG2%=G-JFRPS!AE;* M/`[+M5+YLY'E,N^GG>S*7ZO#<JU0_NQDN<RG(SS;AI2U."Q7H_(WA"R7F>^C M<'U2UNRP7,N5ORBR7%$J1#1%*W]-#LNU3/F+)LL5K4)PSUC*5CHL5X/R%T.6 M*T:%B,4]D+(5#LNU5/F+)<L5JT+$49SRU^BP7$N4OSBR7'$J!/?,I6RYPW(M M5OZ&DN4:JD+$4[SRM\QAN>J5OWBR7.;X3P(E*'\-#LNU2/E+(,ME/KE)I$3E M;ZG#<M4I?XEDN<QQVR1*4OZ6."Q7K?*71);+'&L=!DLI98L=EJM&^1M&EFN8 M"C&<ABM_]0[+M5#Y&TZ6:[@*D4S)RM\BA^6J5OZ2R7*9=><(&J'\U3DLUP+E M;P19+M/VCZ21RE^MPW)5*7\CR7*-#,PC-+C]W]I"S?_%$OOKAU_?G>U(ONW. M.)KD\N\K(a)RRZCXPGFN2Y)[DUQIVC=23GZ-I$TK+?3M(^[P8\%]C]).?J^A') M&9T?(3DWU@&2LQ8=`EPMO$QRYN>C)'6_27).L+$VZC7'%[^[T:])1=DR/^Z6 M,QWQA(%#E![>\W1VO,\='4?F.K2A]N.2Y76=]EQBB<DRPAQ93@#Q0$J<7O0E M<VRDV@]$YRGA0I)!>$U/]OA3]9OA]OF*!15E*^;6592)HXFFI'Q>Q?P+:P*I M4X(]OTFZA>:@5IZ*5E$>34,[8!9E"3<_A\NB7/R5X5<6CI31%$BS*`=_A:`4 MH6;#_S1HF`+DFJ^LUZ<NAL]?31V!TA^\Y[O.LZ;QA?-TDSRU'D_G5R2FU>.I M]+)I#?`&)M*;*2:7](EI,;/%M']KE0X;#6Z#V^`VN`UN@]O@-K@-;H/;X#:X M#6Y?;>.>%?<8N<?&_5#NM7'/CL=H^:U\?H[.?4[N#W-/D_O0_!R`^_C<7^>^ M//?!N<_.;S7RTR_NMW/?GK\)X??(^*T:!T@'8TGV"+E7YP09Q-\#$)T#QH,) MQ&,2AI%)LC_,[RWPS%D\C3L_#<L"_&W19)"C_+JQSR,2WW%S/YV_X><^/[_A M/`WP\X9B\#4P'5P`9I"<-YM[L[.(Q-LE/+XQ&\PA.1]V.:@`_(;WA6`>X"\$ MN%==!180]WYE?[V&>+$%HCJPB'C!!!+/@I:0[,<W@&5@.>#GG2O`2M`$FD$+ M:`4>L`JT@=6@'?#[BFL`/[WG'C$_.^PB.=:R'GA!-\E>>0_8`'@>;1Z#X>GI M^>VC+>!2P'WXRP'/Q[,-7`&N!%>!J\$UX%IP';@>W`"V@QO!#G`3X&<?MX"O M@UO!3G`;R;&>.\`N<">X"]P-O@&^27(,Z![P+7`O^#;X#M@#[B,Y-O1=\#WP M?9+SRC\`]H(?DAPS>A#L`P^!'X.?D!P?>9CD6-+/P'[P*'@,_!P\#IX@.<;T M)'@*/`V>`<^"7X"#),>>G@//@Q?`+\&+X"7P*Y)C4H?!*^!5\&OP&C@"?D-R MK.IU<`S\%OP.O`%^#_Y`<@SKC^`M\"?P-G@'O`O>`_\!W@<?@#^3G$__+^!# M<!Q\!/X*/@9_`R?`?X%/P-])SKO_*?@'^`Q\#OQ``SKX`GP)_@G$XVD4?AN/ MI]EXQFZ4?W[Z8N/OF%#^01P8"N)!`DCD<3XP#`RWR=5%1X"1X`PP"HP&8T`* M2`5IX$R;7,R7UP/@\;MQX"S@!!DV7@@$Y1^,!Q/`N39^(B37#9@(S@>3@,LF MGQ#QLZ7)-CEVE@O<(,\FUQ<H`%/`5%`(IH$B4&R3XY33P05@ADTN0UT"9H%2 M4`9F@SE@KDVN4U`!*L&%8!ZX",P'56`!J`8+00VH!75@$:@'B\$2L!0T@&5@ M.6@$*P"O?]`$FD$+:`4>L`JTV>22*.W@8K#&)M=)6`LZ01=8!]8#+^@&/M`# M-H"-8!/8#"X!6\"EX#)P.=@*>.SV"G`EN`I<#:X!UX+K;'(]AAO`=G`CV`%N M`C>#6VSR;9A;P4YPFTVNVW`'V`7N!'>!N\$WP#?!;G`/^!:X%_#Z#M\!>\!] MX'[P7?`]\'WP`YM<HWXO^*%-K@/Q(-@''@(_!C\!/P4/@T?`S\!^\*A-KA?Q M<_`X>`(<`$^"I\#3@->1>!;\`AP$A\!SX'GP@DVN+_$B>`G\"KP,#H-7P*LV MN>[$:^`(^`TX"EX'Q\!O;7(]BC?`[\$?P)O@C^`M\"?P-G@'O`O>L\FUG-X' M'X`_@_\$?P$?@N/@(_!7\#'XFXV_L4/Y!Y^`OX/_!I^"?X#/P.?`#S2@@R_` ME^"?0"SI:><Y_K$#0T`4X)4"8D`LB+/+U0#X\0"_#I((DL`P,!PD@Q%@I)V? M`LFGMJ/!&)`"4D$:.!,X['(][+%@'#(a)+.$$&.!N<`\:#">!<D`G.`Q/!^6`2 M<($LD`TF@QR0"]P@#^2#`C`%3`6%8!HH`L7@:V`ZN`#,`#-!"9@%2D$9F`WF M@+F@'%2`2G`AF`<N`O-!%5@`JL%"4`-J09U=MDGJL5\,EH"EH`$L`\M!(U@! M5H(FT`Q:0"OP@%6@#:P&[>!BL`9T@+6@$W2!=6`]$)NP&M;]H`-8"/8!#:# M2\`6NXS79=A?#K:";>`*<"6X2AW_,@0WJN,F@]O_CXT7`^-EEIQH^?*S%Z]< MT"B>6^0))[C%OXVX9<_V0JPF>X)[">>R'QH!]QBT&$Q=W(>(D<N'B+85R:8T M,N!GM"?A&=O;'[;;>'^376IE?65H`?M$*UCJDST.V)D3\JE8"GH:42=FP;4G M,?X$AW#Q@7CN<22<B('#'C@_;RDA-24(3=QVY]@_A<Y*5A2_&5R+LS>+!:-D M#XAK1INH'Y)AW=E,&K;D?PW<UX'X71G)!M":*"9P'PIF/[!T`DHX<(%Z!-P? M0#T=%T9T4PV!_'!/%@:0J>'`T"\"MOQ=P!==E8(O:\L#M^0A;H7Y6A_J:V.@ MJ:#^5@,#J-_5!>Q;,3&"^FNL4#=`>E4P]T'FXV`A(`0.`0\PVP34J6$P8&$% MNA-D*VB&#A3WGD#;T\!NR@7'3";X8BV8>2PHYHD!S=-`"W^(*YV!K@3U_C#] M#@K+,&#$X#MEP6Z&]WGQ+D'9IX%T'X6!N3P1\0-`]R_3HR04)0&AV(-*(F" MK[AS860$7T^7"^Q+^0-30198`A'7,@SXS9-E$0+:#S*.G0%Q(S6Z?[&!+##) M#[4U%1@")<#^9#[T$D3\X8=PGR0P#AA1;D/&9A<LW\%H4%N&`9JC0<S1\GSD M`D9@[#-S0=(N>MD-2H%H=X_!-O2"QX1\@T%B0"%PB@2Q]6#R>A8,7RPW%6)+ MC:-@,`$`4$L!`A0`%`````@``&`5*=-$ORDYL0```!X%``P````````````@ I`+:!`````$%00RTQ.3`R+F1O8U!+!08``````0`!`#H```!CL0`````= ` end ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com

2 1

Re: [H.323 Mobility:] Comments on APC-1932
by Roy, Radhika R, ALCOO 18 Aug '00

18 Aug '00

Hi, Jaakko and all: Let me explain some fundamental aspects of H.323 protocol. The way the H.323 protocol works is as follows: 1. Communications among the GKs (or BEs) for resolving addresses If a terminal wants to place a call and it does not know the address of the called party, it simply sends the query to the GK. If the said GK does not have the answer, it simply contacts the other GK to resolve the address. When a GK knows the answer, it sends the answer back to the originating GK, and then the originating GK sends the reply back to the terminal. Please note that H.323 does NOT answer the question how a GK will contact another GK for resolving the address of the called party. There is no standard mechanism how the GK-to-GK communications are made for resolving the problem. The same is also true for the BE-to-BE communication for the inter-domain communications. In the same token, we also defining the same process in mobile environment. 2. Efficiency of the LRQ messages We know that LRQ messages are not very efficient. That is why we created H.225.0 Annex G (inter-domain) message sets that are more efficient than the LRQ. One of the proposal has been to use those H.225.0 Annex G message sets for the intra-domain communications instead of LRQs. However, this proposal is yet to be accepted. So, we can discuss what are the probable solutions can be: 1. Augment LRQs for intra-domain 2. Augment Annex G for inter domain (if needed this can also be used for intra-domain), and/or 3. Create new messages for intra/inter-domain. 3. VLF and HLF communications across the domain (i.e., inter-domain) Yes, this violates the fundamental protocol architecture of H.225.0 Annex G. Please note that Annex G message sets basically provide the address resolution across the administrative domains if the address of the called party is not known in any given domain. The way this works as follows: If a terminal dos not know an address of the called party, it contacts its GK. The GK then contacts the BE (or another GK) if it thinks that answer is not known in this domain. The BE then contacts the another BE of another domain for resolving the address, and then the reply comes back following the reverse path. Again, the H.323 standard does NOT specify how a GK or BE knows which GK or BE needs to be contacted for resolving the information. It simply defines the protocol, nothing else. If the VLFs and HLFs are allowed to by-pass the BEs, then we do NOT need the H.225.0 Annex G BE-to-BE communications protocol at all. This VLF <-> HLF communication becomes another parallel protocol set that can do the same job what Annex G protocol is doing. (By the way, an interesting thing is that if there are many VLFs and HLFs, we may have same problems: How does a VLF knows which HLF needs to contact unless one "hard-wired" in the protocol that there is only "one" HLF, or the address of the HLF is "priori" known?) We have proposals how we can address the mobility problem using the H.225.0 Annex G protocol that uses only the BEs for the inter-domain communications. 3. New terminologies We can talk about this. We may strike the balance in the context of the IP world. For example, 3GPP is using the mobile IP. In this context, home and foreign (visiting [serving]/visited) are used. "Visited" can be currently visited, previous visited, etc. (The terms like "old" and "new" in the classical world of cellular mobility and we may avoid this.) In addition to the above, please also see my reply embedded below [RRR]. I appreciate your response. We will also talk more in the Portland conference. We will miss you. Best regards, Radhika R. Roy AT&T -----Original Message----- From: Jaakko Sundquist [mailto:jaakko.sundquist@nokia.com] Sent: Friday, August 18, 2000 8:53 AM To: ITU-SG16(a)MAILBAG.INTEL.COM Subject: [H.323 Mobility:] Comments on APC-1932 Hi Radhika & the rest of you mob people, Since I'm not coming to Portland, I'll send you my comments on APC-1932 here. You can discuss these issues further with Mr. Rissanen in Portland. ** Comments on chapter 2 of APC-1932 ** The LRQ mechanism will, of course still work, but I think that for a true global mobile system, the mechanism is VERY inefficient and thus a better way of locating the called user must be standardized. I.e. generally a GK has no way of knowing under which GK (or set of GKs) the called user is located and thus in order for the LRQ mechanism to work, the LRQ must be sent to all GKs in a domain. The domains may include many GKs, thus the LRQs would create very much traffic also adding one GK to the domain would necessitate all the other GKs to be informed in order for the LRQ mechanism to work. Furthermore the originating GK does not even know whether the user is located in the same domain, thus the GK would have to wait an arbitrarily long time for an LCF and if the LCF does not arrive, inter-domain communications would be needed -> how long should this time be? [RRR] Please see my reply provided above in items 1 and 2. We will only define the protocol, and will NOT address the implementation issues. I really do not see, how the VLF <-> HLF communications across domain boundaries BREAKS the H.323 model. The H.225 Annex G states that "all inter-domain communications are executed via border elements" (not an exact quotation), but we can either change that statement into "all inter-domain communications must be executed by using the H.225 Annex G protocol" and use the protocol between VLFs and HLFs or we can say that the VLF and the HLF must always be co-located with a BE (to create elements HLF/BE and VLF/BE). [RRR] Please see my reply provided above items 1 and 2. No, we do not need to perform inter-domain VLF and HLF communications. Proposals are there how we can use only the BEs for the inter-domain communications for solving the mobility problems as specified in H.225.0 Annex G. (By the way, those who know why Annex G has been developed using the BEs as the ONLY legitimate entities for the inter-domain communications, they will not raise this issue at all.) As for the question about the "distributed" vs. "hierarchical" model. Again, the VLF, HLF and AuF are just FUNCTIONALITIES, i.e. you can co-locate them with each other or with a GK or a BE. Thus, if you want your GK to be able to directly communicate with a HLF, you can add the VLF functionality to that network element thus creating a GK/VLF (or perhaps a GK/VLF/BE), which can communicate with the HLF. Thus the current H.323 Annex H architectural model does not stop you from utilizing either the "distributed" or the "hierarchical" model. [RRR] No, picture clearly shows that VLF and HLF and other communications scenarios. As I told before, we are NOT considering the VLF and HLF communications per se because there are NOT considered as a part of H.323 protocol. This is called the backend server to the backend server protocol. We will only concentrate for the GK/BE <-> VLF, GK/BE <-> HLF, and GK/BE <-> AuF protocol from the H.323 point of view. (By the way, there are many candidate protocols for the VLF <-> HLF and HLF <-> HLF protocol, and other questions and SGs may also be involved for this.) These comments apply on some parts of the chapter 3 also. ** Section 2.1 ** I do not understand this at all. We are producing an enhanchement to the H.323 protocols and in doing this, we have identified new functional elements that are not present in the H.323 standards beforehand. So, the interfaces may not have been a part of the H.323 standards before, but after Annex H is done they will be. As for the specific interfaces, I don't think we need to specify the HLF-HLF interface (in fact, the interface is not even present in the architecture diagram) and my opinion is that we can leave the AuF-HLF and especially the AuF-VLF interface unspecified in the first version of the annex (thus assuming that the AuF is co-located with the HLF). [RRR] Please see my answers for Section 2 as stated above. If there are more than one HLFs, as agreed, then HLF <-> HLF also needs to be defined because we NOT considering that our protocol only the centralized architecture. So, We will only concentrate for the GK/BE <-> VLF, GK/BE <-> HLF, and GK/BE <-> AuF protocol from the H.323 point of view. I have proposed to use the H.225 Annex G protocol in these interfaces, but there can, of course, be other options as well and I would really love to see contributions utilizing other mechanisms, such as the IETF's AAA, etc. My opinion is, however, that we must select and standardize the use of some specific protocol to really enable interworking between different manufacturer's products. [RRR] No, we do not need to do so. Please see my replies above. These comments apply on some parts of the chapter 3 also. ** Section 2.2 ** I think we have a linguistic misunderstanding here. I would like, if some native English speaker would comment this, but I understand the terms in the following way: - Visited Domain, etc. should be read: "the domain currently being visited by the user". I think that this kind of interpretation is commonly used in technical language (or am I totally wrong?). - Visiting Domain would sound to me like: "the domain is visiting somewhere". - Serving Domain is OK, and is already used in Annex H. - Target Domain is OK. In the present draft I have used "new Domain" instead of "target Domain". Target and source are words that are commonly used for the new and old domain, etc. (E.g. in GSM handovers, the terms "source BTS" and "target BTS" are used. If using the words "new" and "old" is confusing or misleading, I ahve nothing against using target and source. [RRR] We may need to work together. Please see my suggestions provided in item 3. These comments apply on some parts of the chapter 3 also. ** Chapter 3 ** Most of my comments on this chapter are already captured in the above comments for chapter 2. A few short notes on the other issues: - You're right, the security check for unregistration is definitely needed. - Also Connection Termination needs to be handled, although I imagine that there is nothing really special about it compared to the "normal" H.323 system (that is why it isn't included in the present draft), unless supplementary services (and perhaps some billing issues) are taken into account. Those are my comments, I will discuss the issues with Mr. Rissanen so that you can have a meaningful discussion about them in Portland. Have a nice meeting everyone! *** Finally, just to annoy the editor of H.323 Annex K: Ice-hockey isn't enough for us Finns anymore, now we are beating you in football (socker) also... Finland - Norway 3-1!!!! ------------------------------------------------ Jaakko Sundquist * +358 50 3598281 * Audere est Facere! jaakko.sundquist(a)nokia.com * ------------------------------------------------ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com

1 0

[H.323 Mobility:] Comments on APC-1932
by Jaakko Sundquist 18 Aug '00

18 Aug '00

Hi Radhika & the rest of you mob people, Since I'm not coming to Portland, I'll send you my comments on APC-1932 here. You can discuss these issues further with Mr. Rissanen in Portland. ** Comments on chapter 2 of APC-1932 ** The LRQ mechanism will, of course still work, but I think that for a true global mobile system, the mechanism is VERY inefficient and thus a better way of locating the called user must be standardized. I.e. generally a GK has no way of knowing under which GK (or set of GKs) the called user is located and thus in order for the LRQ mechanism to work, the LRQ must be sent to all GKs in a domain. The domains may include many GKs, thus the LRQs would create very much traffic also adding one GK to the domain would necessitate all the other GKs to be informed in order for the LRQ mechanism to work. Furthermore the originating GK does not even know whether the user is located in the same domain, thus the GK would have to wait an arbitrarily long time for an LCF and if the LCF does not arrive, inter-domain communications would be needed -> how long should this time be? I really do not see, how the VLF <-> HLF communications across domain boundaries BREAKS the H.323 model. The H.225 Annex G states that "all inter-domain communications are executed via border elements" (not an exact quotation), but we can either change that statement into "all inter-domain communications must be executed by using the H.225 Annex G protocol" and use the protocol between VLFs and HLFs or we can say that the VLF and the HLF must always be co-located with a BE (to create elements HLF/BE and VLF/BE). As for the question about the "distributed" vs. "hierarchical" model. Again, the VLF, HLF and AuF are just FUNCTIONALITIES, i.e. you can co-locate them with each other or with a GK or a BE. Thus, if you want your GK to be able to directly communicate with a HLF, you can add the VLF functionality to that network element thus creating a GK/VLF (or perhaps a GK/VLF/BE), which can communicate with the HLF. Thus the current H.323 Annex H architectural model does not stop you from utilizing either the "distributed" or the "hierarchical" model. These comments apply on some parts of the chapter 3 also. ** Section 2.1 ** I do not understand this at all. We are producing an enhanchement to the H.323 protocols and in doing this, we have identified new functional elements that are not present in the H.323 standards beforehand. So, the interfaces may not have been a part of the H.323 standards before, but after Annex H is done they will be. As for the specific interfaces, I don't think we need to specify the HLF-HLF interface (in fact, the interface is not even present in the architecture diagram) and my opinion is that we can leave the AuF-HLF and especially the AuF-VLF interface unspecified in the first version of the annex (thus assuming that the AuF is co-located with the HLF). I have proposed to use the H.225 Annex G protocol in these interfaces, but there can, of course, be other options as well and I would really love to see contributions utilizing other mechanisms, such as the IETF's AAA, etc. My opinion is, however, that we must select and standardize the use of some specific protocol to really enable interworking between different manufacturer's products. These comments apply on some parts of the chapter 3 also. ** Section 2.2 ** I think we have a linguistic misunderstanding here. I would like, if some native English speaker would comment this, but I understand the terms in the following way: - Visited Domain, etc. should be read: "the domain currently being visited by the user". I think that this kind of interpretation is commonly used in technical language (or am I totally wrong?). - Visiting Domain would sound to me like: "the domain is visiting somewhere". - Serving Domain is OK, and is already used in Annex H. - Target Domain is OK. In the present draft I have used "new Domain" instead of "target Domain". Target and source are words that are commonly used for the new and old domain, etc. (E.g. in GSM handovers, the terms "source BTS" and "target BTS" are used. If using the words "new" and "old" is confusing or misleading, I ahve nothing against using target and source. These comments apply on some parts of the chapter 3 also. ** Chapter 3 ** Most of my comments on this chapter are already captured in the above comments for chapter 2. A few short notes on the other issues: - You're right, the security check for unregistration is definitely needed. - Also Connection Termination needs to be handled, although I imagine that there is nothing really special about it compared to the "normal" H.323 system (that is why it isn't included in the present draft), unless supplementary services (and perhaps some billing issues) are taken into account. Those are my comments, I will discuss the issues with Mr. Rissanen so that you can have a meaningful discussion about them in Portland. Have a nice meeting everyone! *** Finally, just to annoy the editor of H.323 Annex K: Ice-hockey isn't enough for us Finns anymore, now we are beating you in football (socker) also... Finland - Norway 3-1!!!! ------------------------------------------------ Jaakko Sundquist * +358 50 3598281 * Audere est Facere! jaakko.sundquist(a)nokia.com * ------------------------------------------------ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com

1 0

Re: H.323 Annex K and the Portland meeting
by Sakae OKUBO 18 Aug '00

18 Aug '00

Dear Mr. Skjaeran, At 19:45 +0200 00/08/17, Espen Skj $B%; (Bran (ETO) wrote: >Dear Mr Okubo and Mr. Skran, > >Unfortunately I can not attend to the upcoming meeting in Portland. >I am sorry fopr this confict, as the editor of H.323 Annex K I really >should have attended, especially as there now is an contribution with some >rightful comments to it. > >I have made an updated version of H.323 Annex K with the obvious errors >that APC-1934 points out corrected. Can you provide me a TD number for >this? TD-13 has been allocated to your document and it is now available at /0008_Por directory of the ftp site. >Mr. Christian Groves will present this (and comment further the APC-1934) >for me at the meeting. >As he is mostly covering the H.248 work, we would be glad if this could be >discussed in the joint session. Best regards, Sakae OKUBO *********************************************************** Waseda Research Center Telecommunications Advancement Organization of Japan (TAO) 5th Floor, Nishi-Waseda Bldg. 1-21-1 Nishi-Waseda, Shinjuku-ku, Tokyo 169-0051 Japan Tel: +81 3 5286 3830 (to be transferred) +81 3 3204 8194 (direct) Fax: +81 3 5287 7287 e-mail: okubo(a)giti.or.jp *********************************************************** ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com

1 0

Re: H.323 QOS
by Chip Sharp 17 Aug '00

17 Aug '00

Could I ask a big favor? Please don't copy all the emails on this thread in your responses. Please snip off most of the thread before sending. The copied material is exceeding new content by an order of magnitude. The last email on the thread was 39,000 bytes. With a best case packet size of 1500 bytes, that is 26 packets injected into the network (67 packets with a packet size of 576 - Windows default). With 419 people subscribed to the list, that is about 10,000 packets injected into the Internet for each of these emails sent. Thanks, :-) Chip ------------------------------------------------------------------- Chip Sharp Consulting Engineering Cisco Systems Reality - Love it or Change it. http://www.netaid.org ------------------------------------------------------------------- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com

2 1

Document TD-12 uploaded
by Pandel Juergen 17 Aug '00

17 Aug '00

I have uploaded the document TD-12.zip (Liaison statement to IETF AVT Group for Unequal Error Protection) to the Incoming directory of the AVC-site. Juergen Pandel ---------------------------------------------------------------------------- -- Dr.-Ing. Juergen Pandel Siemens-AG Corporate Technologies ZT IK 2 D-81730 Munich Phone: ++49 89 636 42071 Fax: ++49 89 636 52393 e-mail: juergen.pandel(a)mchp.siemens.de ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com

1 0

Re: H.323 QOS
by Roy, Radhika R, ALCOO 17 Aug '00

17 Aug '00

Hi, James: It is good that you are interested about QOS. Now there are couple of things: Application vs. Network. People know about the monolithic network layer QOS like RSVP, DiffServ, MPLS, ATM QOS, etc. However, with the advent of the network independent multimedia (audio, video, data) application like H.323 where the requirements of the QOS need to be satisfied at the application layer for its audio codecs, video codecs, and data (T.120) from end-to-end basis no matter whether there is a single network, multiple networks, or a single network that implements different network layer QOS schemes in different domains, the situation is a little bit different. H.323 QOS Annex N has been opened to deal with this. The issues like echo and others are also the function of some performance parameters and when the needs of those performance parameters are satisfied, the echo and other criteria are also satisfied. That is, the application level performance parameters are the bottom line, and these parameters are sent end-to-end basis. The end-to-end performance parameters remain the same. In turn, these parameters may also be translated to the lower network layer QOS parameters to have them deal with those in the network layer as appropriate. In H.323, there are H.225.0 (RAS, Q.931) and H.245 where QOS signaling messages are needed to deal with on end-to-end basis. Similar is the case with SIP, although SIP much thinner compared to that of H.323. However, SIP also uses SDP. With some indirect hooks in SIP along with extension of SDP to support QOS, many of those QOS issues can be addressed. However, in Q.13 of SG16 where H.323 is addressed, the interworking issues between the PSTN and packet-switching networks (e.g., IP, ATM) is not addressed. Any contributions will be welcome. In the meantime, please look into Appendix of H.323 QOS Annex N where end-to-end H.323 QOS has been addressed. If there are any questions, please let us know, and we would be glad to answer your questions. Best regards, Radhika R. Roy AT&T -----Original Message----- From: James Dam [mailto:James.Dam@CWO.COM.AU] Sent: Wednesday, August 16, 2000 10:05 PM To: ITU-SG16(a)MAILBAG.INTEL.COM Subject: Re: H.323 QOS Hi Mike, I am new to this forum, i may be wrong, the following is my offering: >From a carrier/operator point of view , we need standard body addressing the end-to-end QoS problems, Echo issues.... We need specs/standards to discuss with other carrier/operator....right now. So much debate between SIP and H.323 but no real progress on this subject end-to-end QoS. As far as i know to date , H.323 is probably in front in many aspects. Why not going beyond the call. Best Regards James Dam +61 29342 8195 James.Dam(a)cwo.com.au ---------- From: Mike Buckley Reply To: Mailing list for parties associated with ITU-T Study Group 16 Sent: Wednesday, 16 August 2000 11:55 PM To: ITU-SG16(a)MAILBAG.INTEL.COM Subject: Re: H.323 QOS Hi Radhika, The problem is I don't see a framework in SG16 for how H.323 will be used in the real world to solve the general end to end QoS problem. It's no good putting fields in protocols without some idea of how they are going to be used in practice. This doesn't mean you want to tie down the implementer to go down one particular route but you have to have an architecture in mind when you develop a protocol. The IETFs architecture is clear - it is the Internet. I would submit that this is not necessarily what we are doing in SG16. Let's define the framework of the problem we are going to solve. Much of it has been covered in this recent thread. But it needs documenting and general agreement. TIPHON is developing a framework for VoIP. SG16 is developing a suite of protocols for VoIP. Both need an agreed architecture. The early discussions on architecture in TIPHON lead to gateway decomposition and the birth of H.248. The logical conclusion of this work has lead to the present TIPHON architecture. Just as the gateway decomposition exercise lead to a set of requirements and architecture for H.248, the latest TIPHON architecture leads to a set of requirements for QoS control and a QoS architecture. See DTS 5003. I want to get general agreement on the problem we are solving and the roadmap of how we are going to get there. Then we can start on the ASN.1. We should not be considering SIP/SDP in SG16 - neither of these protocols are needed if we do our job right with H.323. Policy in the IETF is exclusively transport domain policy. Again it has nothing to do with the policies we may adopt on application level QoS. As far as I understand the MCI work this is all about the Internet and makes no recognition of H.323 or other applications. So lets not get lead astray by orthogonal activities to those we are engaged on. Our job is application level QoS and defining the services we need from the transport domain to deliver it. We are not solving problems in the transport domain. But we can draw up a set of requirements for an interface and feed this to the IETF. Let's get our problem statement clear and our TORs and architecture documented and agreed. Mike Mike Buckley +44-1457-877718 (T) +44-1457-877721 (F) mikebuckley(a)44comms.com ----- Original Message ----- From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> To: <ITU-SG16(a)MAILBAG.INTEL.COM> Sent: Wednesday, August 16, 2000 2:08 PM Subject: Re: H.323 QOS Hi, Mike: Our aim is to extend H.323 for supporting QOS. The framework is already there. Please see H.323 and H.225.0 Annex G specs how the general framework is done in the context of H.323. Again, we are dealing with Q.13/16 and TIPHON exceeds the charter of Q.13/16. It goes beyond H.323 and involves almost all SGs of ITU-T and IETF that are dealing with QOS. I understand what TIPHON is trying to do and TIPHON is just NOT limited to Q.13/16. The same problem has been in the IETF when TIPHON presented its QOS model for extending SIP. It appears that it exceeds the charter of SIP. Please note that people are NOT saying that TIPHON's work is NOT needed. All they are saying: Let us divide the work in different items and each item will have its place in the corresponding WGs. For example, policy will be addressed in the Policy WG, if any particular messages in SIP needs to be extended to support QOS following the SIP charter that work may be addressed in the SIP WG, if any particular item needs to extend in SDP to support QOS that work needs to be done in the MMUSIC. From end-to-end call point of view, people will need SIP, SDP, Policy, Security, Billing, firewalls, NATs, and others. when the standards in all areas are done, people just use them. This is the second step. I have also been working with Joon Maeng of VTEL in this area because Joon submitted a contribution in Q.13/16 last Feb'99 meeting. We are waiting to complete the H.323 QOS work first. You can also see an IETF contribution from MCI WorldCom how the interdomain QOS can be used by SIP using Policy, Security, etc. This work is dependent on the standard work of Policy WG, and others. Again, this contribution for supporting interdomain QOS by SIP using Policy, etc. does NOT need any additional standard work is SIP, it is, rather, an implementation that shows how an end-to-end call can use all those standards including QOS that have been developed in different WGs. Appendix of H.323 QOS Annex N has the general framework. Now you can go through those items and please let us know where you like to see changes and why. Best regards, Radhika R. Roy AT&T -----Original Message----- From: Mike Buckley [ mailto:mikebuckley@44COMMS.COM <mailto:mikebuckley@44COMMS.COM> ] Sent: Tuesday, August 15, 2000 7:51 PM To: ITU-SG16(a)MAILBAG.INTEL.COM Subject: Re: H.323 QOS Radhika, My aim here is to come up with a general QoS framework in H.323 that will handle all these situatioins not just special cases. I think we agree on this. I would be reluctant to solve a special case without an idea of how we are going to solve the general case. This can only lead to backward compatibility issues. I believe the TIPHON approach is totally general. However I agree it need some new interfaces defining. Mike Mike Buckley +44-1457-877718 (T) +44-1457-877721 (F) mikebuckley(a)44comms.com ----- Original Message ----- From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> To: <ITU-SG16(a)MAILBAG.INTEL.COM> Sent: Tuesday, August 15, 2000 8:21 PM Subject: Re: H.323 QOS Hi, Mike: Yes, you are right that NATs and firewalls problems are yet to be solved with complete satisfaction from address translation point of view. This may be a separate work item all by itself. I do not think that we will bring H.245 to address those problems. In fact, I have not seen any proposal in that direction. We have to separate our works step by step. I would not like to include those general problems as a part of H.323 OQS for now. Let a separate group or annexes deal with those special problems so that we can use those solutions for QOS as well, if needed. Best regards, Radhika R. Roy AT&T -----Original Message----- From: Mike Buckley [ mailto:mikebuckley@44COMMS.COM <mailto:mikebuckley@44COMMS.COM> ] Sent: Tuesday, August 15, 2000 1:13 PM To: ITU-SG16(a)MAILBAG.INTEL.COM Subject: Re: H.323 QOS Hi Radhika, This model is only valid I think when you have end to end address transparency as in the Internet. When different transport network domains are present this doesn't necessarily work. NATs and firewalls will mean that IP addresses and port numbers in different domains will be different for the same media stream. Terminating IP and port numbers may well be different at each end of the call. This is one reason why H.323 will not work through firewalls. To fix this gatekeepers will have to perform a mediating function with the H.323 domain and the transport domains. This mediation will not be via H.245. SIP has the same problems by the way. Mike Mike Buckley +44-1457-877718 (T) +44-1457-877721 (F) mikebuckley(a)44comms.com ----- Original Message ----- From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> To: <ITU-SG16(a)MAILBAG.INTEL.COM> Sent: Tuesday, August 15, 2000 2:15 PM Subject: Re: H.323 QOS Hi, Mike: Just for information, H.245 does many things. As I indicated earlier, one of the most important things of H.245 is the opening and closing of the logical channels that binds the application layer's logical connection abstractions to all transport network connections (e.g., ATM, etc). So, this is the fundamental mission of H.245 to provide the continuity between the application layer and the lower layer in a transport independent way using the universal abstractions. These abstractions can be for anything that we can think of. Hope this helps. Best regards, Radhika R. Roy AT&T -----Original Message----- From: Mike Buckley [ mailto:mikebuckley@44COMMS.COM <mailto:mikebuckley@44COMMS.COM> ] Sent: Tuesday, August 15, 2000 8:22 AM To: ITU-SG16(a)MAILBAG.INTEL.COM Subject: Re: H.323 QOS Hi Radhika, This use of H.245 as a bridging mechanism is where I have difficulties. The original intenmt behind H.245 was media control and negociation. Transport mechanisms should not be signalled in H.245 or elsewhere in the application. This is entirely the business of the transport network provider. Unfortunately I will not be present in Portland, however, I am happy to continue this productive dialogue on the list with a view to providing a new draft in the next few weeks. Regards, Mike Mike Buckley +44-1457-877718 (T) +44-1457-877721 (F) mikebuckley(a)44comms.com ----- Original Message ----- From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> To: <ITU-SG16(a)MAILBAG.INTEL.COM> Sent: Monday, August 14, 2000 7:29 PM Subject: Re: H.323 QOS Hi, Mike: I guess that we have agreement on the fundamental basis. However, I have problems when you translate this in signaling terms. As I have explained earlier that we are following the exiting framework of H.323 which is the application layer, and H.323 is transported over one or different transport networks. H.323 uses H.245 as a common mechanism that helps to transports the abstraction of H.323 application signaling messages over any networks (e.g., IP, ATM, etc). You may look to the H.245 spec very carefully instead of making any generalized statement. H.245 provides the bridging between the application layer and transport layer mechanism. For example, H.245's OLC abstraction is the link for ATM network layer logical connection as well as for other transport networks. In the same token, H.323 QOS abstraction may also provide link to the lower transport layer QOS signaling mechanisms (e.g., RSVP, DiffServ, MPLS, ATM QOS, etc), if needed. H.323 DOES support RSVP and ATM QOS now. Please also see my reply embedded in you email [RRR]. Again, I would suggest to bring contributions explaining the solution and we will go from there. In absence of your contributions, I would suggest to provide comments on Appendix of H.323 QOS Annex N. We can then answer your questions step by step. Best regards, Radhika R. Roy AT&T -----Original Message----- From: Mike Buckley [ mailto:mikebuckley@44COMMS.COM <mailto:mikebuckley@44COMMS.COM> ] Sent: Monday, August 14, 2000 12:56 PM To: ITU-SG16(a)MAILBAG.INTEL.COM Subject: Re: H.323 QOS Radhika, >However, this is only good for the single network. For example, >end-to-end RSVP or end-to-end ATM QOS. >If there are multiple networks or if a single IP network implements RSVP in >one domain, DiffServ in another domain, and MPLS in another domain, there is >no transparent H.323 QOS signaling mechanism that is universal on end-to-end >basis so that it can be mapped over the RSVP, DiffServ, MPLS, ATM QOS, etc >transparently This is the nub of the problem and I am afraid why this approach is flawed. [RRR] I am NOT afraid of anything. Rather we are solving all problems including the above as well. No one should be afraid of anything. The only thing that I can suggest is to bring contributions. We will go from there. Please remember that H.323 QOS supports RSVP and ATM QOS and we MUST provide backward compatibility. We will be waiting to see contributions from you to prove your points whether things are flawed or not. Please read Appnedix of H.323 QOS Annex N very carefully along with exiting H.323 and H.245 standard how they support H.323 QOS for RSVP and ATM QOS. >We have shown how the H.323 QOS can >be mapped over the RSVP, DiffServ, ATM QOS, etc. if needed. This mapping must be done within the transport plane not by the application. All the application needs to specify are the entries in your Table 1. i.e delay, jitter. packet loss and peak bit rate (the latter parameter for policy enforcement and resource reservation). [RRR] Please see my generalized clarification how H.245 provides translation between H.323 signaling schemes over the lower layer transport networks (e.g., IP, ATM, etc). The mechanism already exists in H.323 and this is the beauty of the transport independent H.323 application. Now there are two separate things: one is what the calling side of the application demands and what the called side of the application agrees on. This is called negotiation capability (e.g., choosing a particular type codec that has a certain bandwidth, jitter, and other requirements). How the choosing criteria is determined, the H.323 application is NOT aware of those things. It may depend on many criteria (e.g., policy, bandwidth, type of network [e.g., IP, ATM], pricing, security, personal preferences, etc.), but H.323 is not involved. These are separate issues and we are NOT addressing at the H.323 layer. I am still not clkear how you propose to signal these parameters to the transport plane. [RRR] It is NOT new in H.323, and this mechanism already exits. Please see the support of RSVP and ATM QOS in H.323/H.245 spec. We are not inventing anything NEW here. Then, if you read Appendix of H.323 QOS Annex N very carefully, you will find how it happening (whether you call it as transport plane or something else, it does not matter). I think we are getting loser to a common understanding. Mike Mike Buckley +44-1457-877718 (T) +44-1457-877721 (F) mikebuckley(a)44comms.com ----- Original Message ----- From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> To: <ITU-SG16(a)MAILBAG.INTEL.COM> Sent: Monday, August 14, 2000 2:37 PM Subject: Re: H.323 QOS Hi, Bob: I agree completely agree with you. In fact, we are all trying to achieve the same goal. For example, H.323 does supports QOS today via H.245. You can easily see that RSVP and ATM QOS are supported using H.245. The beauty of this approach is that H.245 is still transport independent. All we have done here is: the abstraction of H.245 has been used to support the RSVP and ATM QOS to implement the network layer QOS. However, this is only good for the single network. For example, end-to-end RSVP or end-to-end ATM QOS. If there are multiple networks or if a single IP network implements RSVP in one domain, DiffServ in another domain, and MPLS in another domain, there is no transparent H.323 QOS signaling mechanism that is universal on end-to-end basis so that it can be mapped over the RSVP, DiffServ, MPLS, ATM QOS, etc transparently. In Appendix of H.323 Annex N, we have done the same. We have the abstraction of H.323 QOS in the application layer. We have shown how the H.323 QOS can be mapped over the RSVP, DiffServ, ATM QOS, etc. if needed. It also provides the backward compatibility with the existing H.323 standard. However, we have done only for the pre-call setup signaling part. We have not done the call setup part yet. In the call setup part, we will include H.245 in a similar way what H.323 is supporting RSVP and ATM QOS today in a monolithic network. I appreciate your email. Best regards, Radhika R. Roy AT&T > -----Original Message----- > From: Callaghan, Robert [SMTP:Robert.Callaghan@icn.siemens.com] > Sent: Monday, August 14, 2000 9:08 AM > To: Roy, Radhika R, ALCOO > Cc: 'Mailing list for parties associated with ITU-T Study Group 16' > Subject: RE: H.323 QOS > > Roy, > > In my view, the network, as a minimum, needs only provide transport. > H.323 > is an application using this transport and is not different from any other > application being transported. Any application may request a desired > quality of service that the network may grant based on policies, service > level agreements, and availability. Each data connection used by an > application may request a different QOS. Most likely H.323 is such an > application as needing an enhanced QOS. This simple case should work. > This > is all that is required. > > In addition to transport, the network may optionally provide other > additional services. These services may include application layer > routing. > These optional services should be configured and indicated independently > from the basic transport QOS. > > It is correct that H.323, as an application, is (mostly) transport > independent. However, the interface between the application and the > transport layer is not transport independent. The interface specification > used to request a given QOS is totally dependant on the standards body > that > specified the transport layer; and for this there has been little or no > coordination. > > Because transport QOS over IP is based on IETF specifications, it is > necessary that the interface used to request a given QOS conform to the > appropriate IETF specification. If such an interface specification is not > available, that input might be provided to the appropriate IETF body as to > the requirements. > > I can see the endpoints negotiating the desired QOS base on need, price, > and > other considerations. For me, this is the limit to the application layer > involvement in QOS negotiation. At this point, the application > negotiations > with the transport provider as to the desired and available QOS. It is up > to the transport provider to arrange for the end-to-end QOS. (Again, it > is > not necessary for the transport network to know that H.323 is involved in > the transported data. In fact it may be encrypted in order to mask the > presence of voice transport.) > > For me, this is simple. > > Bob > > ------------------------------------------------------------------ > Robert Callaghan > Siemens Enterprise Networks > Tel: +1.561.923.1756 Fax: +1.561.923.1403 > Email: Robert.Callaghan(a)ICN.Siemens.com > ------------------------------------------------------------------ > > -----Original Message----- > From: Roy, Radhika R, ALCOO [ mailto:rrroy@ATT.COM <mailto:rrroy@ATT.COM> ] > Sent: Friday, August 11, 2000 7:59 AM > To: ITU-SG16(a)mailbag.cps.intel.com > Subject: Re: H.323 QOS > > Hi, Mike: > > Let me try again. > > What is the reference point of H.323 QOS? Is it not H.323? If it is so, > what > do we mean by H.323? > > The answer is: Audio (different codecs), Video (different codecs), and > Data > (T.120 applications) that are used by H.323. > > What are the QOS/performance characteristics of audio, video, and data > from > the application point of view that is generated by audio codecs, video > codecs, and data (T.120) applications? > > These QOS/performance characteristics come from the SOURCE codecs and data > applications. Per transport independent H.323 specifications, an enduser > express their QOS/performance requirements on end-to-end basis purely from > application point of view irrespective of the transport network (e.g., IP, > ATM, etc.). > > Moreover, H.323 is meant for the packet network, not for any > circuit-switched network like PSTN or ISDN. > > Let us NOT go beyond this before we start debating transport layer QOS or > service provider requirements. These are NOT the concern of H.323. H.323 > is > the transport independent application. > > H.323v2/v3/v4 has also provided mechanisms how RSVP and ATM QOS can be > used > for H.323 audio, video, and data. So, H.323 QOS that will be defined in > H.323 Annex N MUST provide mapping for the backward compatibility. It is a > requirement that MUST be met per the norm of ITU-T. > > So, what is left for mapping? Mapping is simply a by-product of the above > requirement. Mapping is simply a table, nothing else. > > Did I miss anything? > > Best regards, > Radhika R. Roy > AT&T > > -----Original Message----- > From: Mike Buckley [ mailto:mikebuckley@44COMMS.COM <mailto:mikebuckley@44COMMS.COM> ] > Sent: Thursday, August 10, 2000 10:19 PM > To: ITU-SG16(a)MAILBAG.INTEL.COM > Subject: Re: H.323 QOS > > > Radhika, > > Thanks for the input which I welcome as I will unfortunately not be > present > at Portland. > > Let me ask a few questions and make a few comments hopefully with the > intent > of opening up the debate. > > 1. I am not sure I understand your concept of a mapping table between the > H.323 QOS and the transport layer QoS. My understanding is that QoS is on > three levels: > > a) that specified from a service point of view between the user and > service > provider (e.g PSTN quality, conference quality etc) This is the domain of > the speech experts and can be characterised by Listener Speech Quaklity > (MOS), end to end delay, and absolute category rating, R. > > b) application specific parameters, (e.g. equipment delays, codec choice > and performance, codec frame size, packetisation arrangements, jitter > buffer > design, overall packet loss etc.) Optimisation of all these will > determine > what can be delivered in a). > > c) transport parameters for a given choice of application parameters. > This > boils down only to three parameters as far as I cna see: tranport network > delay, packet delay variation in the transport network and packet loss in > the transport network. Again these parameters will determine the results > in > a) for a given choice of the parameters in b). These parameters are > generic > from the perspective of the transport network. i.e the transport network > does not need to know the details of the application. > > So the sequence of cause and effect and control is: > > a) User requests QoS class from service provider, > b) Service provider determines application specific parameters in > conjunction with users equipment and other service providers, > c) Service provider requests required delay, delay variation and packet > loss from network provider. > > I see no need for mapping here. The only QoS info flows within the > application are specific to the application and those between the > application (service provider) and the transport network are generic. i.e. > delay, jitter and packet loss. Have I missed something? > > 2. The issue of bit rate and media stream statistics I think need to be > decoupled from QoS. These are specified to enable optimisation of > resources > within the transport network. They have no QoS significance from an > application point of view. i.e the apllication does not care about the > media stream bit rate and statistics but the transport network provider > does > as it eats up his resource. They may be used for policy enforcement > however > in the transport network so they do need to be agreed between service > provider and network operator. i.e the network operator agrees to provide > a > given QoS level (delay, jitter, packet loss) provided the media properties > are within an agreed profile (bit rate, flow statistics). > > 3. The next point is how can the service provider know the statistics of > a > particular VBR stream? These can only be specified over a large number of > similar calls and will depend, for instance, on who is speaking, the > nature > of the speech interaction etc etc. They can only be measured not > calculated. The service provider is in no better position to measure > these > than the transport network operator and, in fact, where no gateways are > involved, may not be able to. On the other hand the class of signal would > have to be signalled to the network operator for him to be able to > distinguish which class a particular measurement belonged to. e.g > voice/speech/data, codec type, conference, multicast etc. So I see no > purpose in trying to exchange statistics between the service provider > (application) and transport operator. I think peak bit rate is all that > can > be meaningfully excanged. The specification of media class is however > perhaps worth exploring. > > 4. The controlled category has always puzzled me. I only see two > possibilities. Either the requested QoS level is guaranteed (on a > statistical basis e.g 95% of all connections over a specified period) or > not > guaranteed. Is your controlled category a way of saying guaranteed, not > to > 95% but to some lower figure? If you can't put a percentage on it then it > seems it is plain and simple not guaranteed. Anything that is not > guaranteed to some specified statistical level is best effort and you > can't > say anything more about it. So I only see two categories here. > > In summary, I think we need to do three things in Annex N. > > a) Figure out the QoS information to be exchanged within the Application > between service providers and end users. This will go in H.225.0 and > H.245. > > b) Figure out how we are going to signal QoS and media information between > the application (service providers) and transport domains (IP or ATM > networks etc). The info is basically delay, jitter, packet loss > requirements and peak bit rate. We need a protocol for this. > > c) we need to work out the interactions between the application QoS > signal > flows and the application/transport signal flows. I don't think we need > worry about how the transport network mechanisms assure the requested QoS > paramerters. RSVP/Intserv, Diffserv, MPLS, ATM, over provisioning are all > possibilities. > > Would welcome comments and views on the above. > > Mike > > > Mike Buckley > +44-1457-877718 (T) > +44-1457-877721 (F) > mikebuckley(a)44comms.com > > > ----- Original Message ----- > From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> > To: <ITU-SG16(a)MAILBAG.INTEL.COM> > Sent: Thursday, August 10, 2000 10:15 PM > Subject: H.323 QOS > > > Hi, Mike and All: > > It is time to discuss about H.323 QOS. > > I believe that we have an agreement as follows: > > · H.323 QOS MUST be backward compatible to support RSVP and ATM QOS as it > exists for H.323v2/v3/v4 > · Like H.323 spec, the application level H.323 QOS MUST be independent of > the transport layer QOS and should support all transport networks (e.g., > IP, > ATM) > · A mapping table between the H.323 QOS and the transport layer QOS (e.g., > IP QOS [DiffServ, RSVP, etc.], ATM QOS [CBR, rt-VBR, nrt-VBR, ABR, etc.]) > should be provided. > > From the H.323 multimedia application point of view, there are following > performance parameters can be used to characterize the traffic > characteristics: > > · Bitrate characteristics: Peak bit rate (PBR) or peak rate (PR), > Sustained > bit rate (SBR) or average rate (AR), minimum bit rate (MBR) or minimum > rate > (MR), and mean bust size (MBS) > · Delay and loss characteristics: end-to-end delay (EED) or delay, > end-to-end delay variation (EEDV) or delay variation (DV), and > bit-error-rate (BER) or (packet) loss rate (LR) > > We can now form a table with all parameters as follows: > > Table 1: H.323 Multimedia Application Performance Matrix > Audio (codecs)--- Video (codecs)--- Data > (T.120) > PBR/PR Yes/No/Value Yes/No/Value Yes/No/value > SBR/AR Yes/No/Value Yes/No/Value Yes/No/value > MBR/MR Yes/No/Value Yes/No/Value Yes/No/value > MBS Yes/No/Value Yes/No/Value Yes/No/value > EED/Delay Yes/No/Value Yes/No/Value Yes/No/value > EEDV/DV Yes/No/Value Yes/No/Value Yes/No/value > BER/LR Yes/No/Value Yes/No/Value Yes/No/value > > From the above table we will have the opportunity to choose each parameter > for each medium (audio, video, data) that makes sense from the > application's > and enduser's point of view. Again, these parameters can be specified as > follows: > > · Guaranteed: The value specified for each parameter MUST be guaranteed. > · Controlled: The value specified for each parameter MAY be satisfied as > far > as practicable (possibly with certain range), but definitely NOT > guaranteed. > · Best effort: No commitment will be made. > > Now each medium (e.g., audio, video, or data) will have different > categories > of performance matrix depending on its selection criteria and this can > also > be mapped to RSVP, ATM QOS, and others, if needed. > > Once we agree on this format, the next step is to create H.323 QOS > signaling > messages. > > This is my input for discussion in the upcoming Portland Q.13 meeting for > H.323 QOS. > > I like to see the comments from other members as well. > > Best regards, > Radhika R. Roy > AT&T > +1 732 420 1580 > rrroy(a)att.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com

1 0

APC-1947 from ECMA
by Sakae OKUBO 17 Aug '00

17 Aug '00

Dear Q12-14/16 experts, The web page document referred to in APC-1947 "IP Telephony and CSTA" is now available as APC-1947-att at the /0008_Por directory. You can browse it locally by opening APC-1947-att.html. Best regards, Sakae OKUBO *********************************************************** Waseda Research Center Telecommunications Advancement Organization of Japan (TAO) 5th Floor, Nishi-Waseda Bldg. 1-21-1 Nishi-Waseda, Shinjuku-ku, Tokyo 169-0051 Japan Tel: +81 3 5286 3830 (to be transferred) +81 3 3204 8194 (direct) Fax: +81 3 5287 7287 e-mail: okubo(a)giti.or.jp *********************************************************** ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv(a)mailbag.intel.com

1 0

Re: H.323 QOS
by James Dam 17 Aug '00

17 Aug '00

Hi Mike, I am new to this forum, i may be wrong, the following is my offering: >From a carrier/operator point of view , we need standard body addressing the end-to-end QoS problems, Echo issues.... We need specs/standards to discuss with other carrier/operator....right now. So much debate between SIP and H.323 but no real progress on this subject end-to-end QoS. As far as i know to date , H.323 is probably in front in many aspects. Why not going beyond the call. Best Regards James Dam +61 29342 8195 James.Dam(a)cwo.com.au > ---------- > From: Mike Buckley > Reply To: Mailing list for parties associated with ITU-T Study Group > 16 > Sent: Wednesday, 16 August 2000 11:55 PM > To: ITU-SG16(a)MAILBAG.INTEL.COM > Subject: Re: H.323 QOS > > Hi Radhika, > > The problem is I don't see a framework in SG16 for how H.323 will be used > in > the real world to solve the general end to end QoS problem. > > It's no good putting fields in protocols without some idea of how they are > going to be used in practice. This doesn't mean you want to tie down the > implementer to go down one particular route but you have to have an > architecture in mind when you develop a protocol. The IETFs architecture > is > clear - it is the Internet. I would submit that this is not necessarily > what we are doing in SG16. > > Let's define the framework of the problem we are going to solve. Much of > it > has been covered in this recent thread. But it needs documenting and > general agreement. > > TIPHON is developing a framework for VoIP. SG16 is developing a suite of > protocols for VoIP. Both need an agreed architecture. The early > discussions on architecture in TIPHON lead to gateway decomposition and > the > birth of H.248. The logical conclusion of this work has lead to the > present > TIPHON architecture. Just as the gateway decomposition exercise lead to a > set of requirements and architecture for H.248, the latest TIPHON > architecture leads to a set of requirements for QoS control and a QoS > architecture. See DTS 5003. I want to get general agreement on the > problem > we are solving and the roadmap of how we are going to get there. Then we > can start on the ASN.1. > > We should not be considering SIP/SDP in SG16 - neither of these protocols > are needed if we do our job right with H.323. Policy in the IETF is > exclusively transport domain policy. Again it has nothing to do with the > policies we may adopt on application level QoS. As far as I understand > the > MCI work this is all about the Internet and makes no recognition of H.323 > or > other applications. So lets not get lead astray by orthogonal activities > to > those we are engaged on. Our job is application level QoS and defining > the > services we need from the transport domain to deliver it. We are not > solving problems in the transport domain. But we can draw up a set of > requirements for an interface and feed this to the IETF. > > Let's get our problem statement clear and our TORs and architecture > documented and agreed. > > Mike > > > > > > Mike Buckley > +44-1457-877718 (T) > +44-1457-877721 (F) > mikebuckley(a)44comms.com > > > ----- Original Message ----- > From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> > To: <ITU-SG16(a)MAILBAG.INTEL.COM> > Sent: Wednesday, August 16, 2000 2:08 PM > Subject: Re: H.323 QOS > > > Hi, Mike: > > Our aim is to extend H.323 for supporting QOS. The framework is already > there. > > Please see H.323 and H.225.0 Annex G specs how the general framework is > done > in the context of H.323. > > Again, we are dealing with Q.13/16 and TIPHON exceeds the charter of > Q.13/16. It goes beyond H.323 and involves almost all SGs of ITU-T and > IETF > that are dealing with QOS. > > I understand what TIPHON is trying to do and TIPHON is just NOT limited to > Q.13/16. The same problem has been in the IETF when TIPHON presented its > QOS > model for extending SIP. It appears that it exceeds the charter of SIP. > > Please note that people are NOT saying that TIPHON's work is NOT needed. > All > they are saying: Let us divide the work in different items and each item > will have its place in the corresponding WGs. For example, policy will be > addressed in the Policy WG, if any particular messages in SIP needs to be > extended to support QOS following the SIP charter that work may be > addressed > in the SIP WG, if any particular item needs to extend in SDP to support > QOS > that work needs to be done in the MMUSIC. > > From end-to-end call point of view, people will need SIP, SDP, Policy, > Security, Billing, firewalls, NATs, and others. when the standards in all > areas are done, people just use them. This is the second step. I have also > been working with Joon Maeng of VTEL in this area because Joon submitted a > contribution in Q.13/16 last Feb'99 meeting. We are waiting to complete > the > H.323 QOS work first. > > You can also see an IETF contribution from MCI WorldCom how the > interdomain > QOS can be used by SIP using Policy, Security, etc. This work is dependent > on the standard work of Policy WG, and others. Again, this contribution > for > supporting interdomain QOS by SIP using Policy, etc. does NOT need any > additional standard work is SIP, it is, rather, an implementation that > shows > how an end-to-end call can use all those standards including QOS that have > been developed in different WGs. > > Appendix of H.323 QOS Annex N has the general framework. Now you can go > through those items and please let us know where you like to see changes > and > why. > > Best regards, > Radhika R. Roy > AT&T > > -----Original Message----- > From: Mike Buckley [mailto:mikebuckley@44COMMS.COM] > Sent: Tuesday, August 15, 2000 7:51 PM > To: ITU-SG16(a)MAILBAG.INTEL.COM > Subject: Re: H.323 QOS > > > Radhika, > > My aim here is to come up with a general QoS framework in H.323 that will > handle all these situatioins not just special cases. I think we agree on > this. > > I would be reluctant to solve a special case without an idea of how we are > going to solve the general case. This can only lead to backward > compatibility issues. > > I believe the TIPHON approach is totally general. However I agree it need > some new interfaces defining. > > Mike > > > Mike Buckley > +44-1457-877718 (T) > +44-1457-877721 (F) > mikebuckley(a)44comms.com > > > ----- Original Message ----- > From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> > To: <ITU-SG16(a)MAILBAG.INTEL.COM> > Sent: Tuesday, August 15, 2000 8:21 PM > Subject: Re: H.323 QOS > > > Hi, Mike: > > Yes, you are right that NATs and firewalls problems are yet to be solved > with complete satisfaction from address translation point of view. This > may > be a separate work item all by itself. > > I do not think that we will bring H.245 to address those problems. In > fact, > I have not seen any proposal in that direction. > > We have to separate our works step by step. > > I would not like to include those general problems as a part of H.323 OQS > for now. Let a separate group or annexes deal with those special problems > so > that we can use those solutions for QOS as well, if needed. > > Best regards, > Radhika R. Roy > AT&T > > -----Original Message----- > From: Mike Buckley [mailto:mikebuckley@44COMMS.COM] > Sent: Tuesday, August 15, 2000 1:13 PM > To: ITU-SG16(a)MAILBAG.INTEL.COM > Subject: Re: H.323 QOS > > > Hi Radhika, > > This model is only valid I think when you have end to end address > transparency as in the Internet. When different transport network domains > are present this doesn't necessarily work. NATs and firewalls will mean > that IP addresses and port numbers in different domains will be different > for the same media stream. Terminating IP and port numbers may well be > different at each end of the call. This is one reason why H.323 will not > work through firewalls. To fix this gatekeepers will have to perform a > mediating function with the H.323 domain and the transport domains. This > mediation will not be via H.245. > > SIP has the same problems by the way. > > Mike > > > Mike Buckley > +44-1457-877718 (T) > +44-1457-877721 (F) > mikebuckley(a)44comms.com > > > ----- Original Message ----- > From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> > To: <ITU-SG16(a)MAILBAG.INTEL.COM> > Sent: Tuesday, August 15, 2000 2:15 PM > Subject: Re: H.323 QOS > > > Hi, Mike: > > Just for information, H.245 does many things. As I indicated earlier, one > of > the most important things of H.245 is the opening and closing of the > logical > channels that binds the application layer's logical connection > abstractions > to all transport network connections (e.g., ATM, etc). > > So, this is the fundamental mission of H.245 to provide the continuity > between the application layer and the lower layer in a transport > independent > way using the universal abstractions. These abstractions can be for > anything > that we can think of. > > Hope this helps. > > Best regards, > Radhika R. Roy > AT&T > > -----Original Message----- > From: Mike Buckley [mailto:mikebuckley@44COMMS.COM] > Sent: Tuesday, August 15, 2000 8:22 AM > To: ITU-SG16(a)MAILBAG.INTEL.COM > Subject: Re: H.323 QOS > > > Hi Radhika, > > This use of H.245 as a bridging mechanism is where I have difficulties. > The > original intenmt behind H.245 was media control and negociation. > Transport > mechanisms should not be signalled in H.245 or elsewhere in the > application. > This is entirely the business of the transport network provider. > > Unfortunately I will not be present in Portland, however, I am happy to > continue this productive dialogue on the list with a view to providing a > new > draft in the next few weeks. > > Regards, > > Mike > > Mike Buckley > +44-1457-877718 (T) > +44-1457-877721 (F) > mikebuckley(a)44comms.com > > > ----- Original Message ----- > From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> > To: <ITU-SG16(a)MAILBAG.INTEL.COM> > Sent: Monday, August 14, 2000 7:29 PM > Subject: Re: H.323 QOS > > > Hi, Mike: > > I guess that we have agreement on the fundamental basis. However, I have > problems when you translate this in signaling terms. As I have explained > earlier that we are following the exiting framework of H.323 which is the > application layer, and H.323 is transported over one or different > transport > networks. H.323 uses H.245 as a common mechanism that helps to transports > the abstraction of H.323 application signaling messages over any networks > (e.g., IP, ATM, etc). > > You may look to the H.245 spec very carefully instead of making any > generalized statement. > > H.245 provides the bridging between the application layer and transport > layer mechanism. For example, H.245's OLC abstraction is the link for ATM > network layer logical connection as well as for other transport networks. > > In the same token, H.323 QOS abstraction may also provide link to the > lower > transport layer QOS signaling mechanisms (e.g., RSVP, DiffServ, MPLS, ATM > QOS, etc), if needed. H.323 DOES support RSVP and ATM QOS now. > > Please also see my reply embedded in you email [RRR]. > > Again, I would suggest to bring contributions explaining the solution and > we > will go from there. In absence of your contributions, I would suggest to > provide comments on Appendix of H.323 QOS Annex N. We can then answer your > questions step by step. > > Best regards, > Radhika R. Roy > AT&T > > -----Original Message----- > From: Mike Buckley [mailto:mikebuckley@44COMMS.COM] > Sent: Monday, August 14, 2000 12:56 PM > To: ITU-SG16(a)MAILBAG.INTEL.COM > Subject: Re: H.323 QOS > > > Radhika, > > >However, this is only good for the single network. For example, > >end-to-end RSVP or end-to-end ATM QOS. > > >If there are multiple networks or if a single IP network implements RSVP > in > >one domain, DiffServ in another domain, and MPLS in another domain, there > is > >no transparent H.323 QOS signaling mechanism that is universal on > end-to-end > >basis so that it can be mapped over the RSVP, DiffServ, MPLS, ATM QOS, > etc > >transparently > > This is the nub of the problem and I am afraid why this approach is > flawed. > [RRR] I am NOT afraid of anything. Rather we are solving all problems > including the above as well. No one should be afraid of anything. The only > thing that I can suggest is to bring contributions. We will go from there. > Please remember that H.323 QOS supports RSVP and ATM QOS and we MUST > provide > backward compatibility. We will be waiting to see contributions from you > to > prove your points whether things are flawed or not. Please read Appnedix > of > H.323 QOS Annex N very carefully along with exiting H.323 and H.245 > standard how they support H.323 QOS for RSVP and ATM QOS. > > >We have shown how the H.323 QOS can > >be mapped over the RSVP, DiffServ, ATM QOS, etc. if needed. > > This mapping must be done within the transport plane not by the > application. > All the application needs to specify are the entries in your Table 1. i.e > delay, jitter. packet loss and peak bit rate (the latter parameter for > policy enforcement and resource reservation). > [RRR] Please see my generalized clarification how H.245 provides > translation > between H.323 signaling schemes over the lower layer transport networks > (e.g., IP, ATM, etc). The mechanism already exists in H.323 and this is > the > beauty of the transport independent H.323 application. Now there are two > separate things: one is what the calling side of the application demands > and > what the called side of the application agrees on. This is called > negotiation capability (e.g., choosing a particular type codec that has a > certain bandwidth, jitter, and other requirements). How the choosing > criteria is determined, the H.323 application is NOT aware of those > things. > It may depend on many criteria (e.g., policy, bandwidth, type of network > [e.g., IP, ATM], pricing, security, personal preferences, etc.), but H.323 > is not involved. These are separate issues and we are NOT addressing at > the > H.323 layer. > > I am still not clkear how you propose to signal these parameters to the > transport plane. > [RRR] It is NOT new in H.323, and this mechanism already exits. Please see > the support of RSVP and ATM QOS in H.323/H.245 spec. We are not inventing > anything NEW here. Then, if you read Appendix of H.323 QOS Annex N very > carefully, you will find how it happening (whether you call it as > transport > plane or something else, it does not matter). > > I think we are getting loser to a common understanding. > > Mike > > Mike Buckley > +44-1457-877718 (T) > +44-1457-877721 (F) > mikebuckley(a)44comms.com > > > ----- Original Message ----- > From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> > To: <ITU-SG16(a)MAILBAG.INTEL.COM> > Sent: Monday, August 14, 2000 2:37 PM > Subject: Re: H.323 QOS > > > Hi, Bob: > > I agree completely agree with you. > > In fact, we are all trying to achieve the same goal. For example, H.323 > does > supports QOS today via H.245. You can easily see that RSVP and ATM QOS are > supported using H.245. The beauty of this approach is that H.245 is still > transport independent. All we have done here is: the abstraction of H.245 > has been used to support the RSVP and ATM QOS to implement the network > layer > QOS. However, this is only good for the single network. For example, > end-to-end RSVP or end-to-end ATM QOS. > > If there are multiple networks or if a single IP network implements RSVP > in > one domain, DiffServ in another domain, and MPLS in another domain, there > is > no transparent H.323 QOS signaling mechanism that is universal on > end-to-end > basis so that it can be mapped over the RSVP, DiffServ, MPLS, ATM QOS, etc > transparently. > > In Appendix of H.323 Annex N, we have done the same. We have the > abstraction > of H.323 QOS in the application layer. We have shown how the H.323 QOS can > be mapped over the RSVP, DiffServ, ATM QOS, etc. if needed. It also > provides > the backward compatibility with the existing H.323 standard. However, we > have done only for the pre-call setup signaling part. We have not done the > call setup part yet. In the call setup part, we will include H.245 in a > similar way what H.323 is supporting RSVP and ATM QOS today in a > monolithic > network. > > I appreciate your email. > > Best regards, > Radhika R. Roy > AT&T > > > -----Original Message----- > > From: Callaghan, Robert [SMTP:Robert.Callaghan@icn.siemens.com] > > Sent: Monday, August 14, 2000 9:08 AM > > To: Roy, Radhika R, ALCOO > > Cc: 'Mailing list for parties associated with ITU-T Study Group 16' > > Subject: RE: H.323 QOS > > > > Roy, > > > > In my view, the network, as a minimum, needs only provide transport. > > H.323 > > is an application using this transport and is not different from any > other > > application being transported. Any application may request a desired > > quality of service that the network may grant based on policies, service > > level agreements, and availability. Each data connection used by an > > application may request a different QOS. Most likely H.323 is such an > > application as needing an enhanced QOS. This simple case should work. > > This > > is all that is required. > > > > In addition to transport, the network may optionally provide other > > additional services. These services may include application layer > > routing. > > These optional services should be configured and indicated independently > > from the basic transport QOS. > > > > It is correct that H.323, as an application, is (mostly) transport > > independent. However, the interface between the application and the > > transport layer is not transport independent. The interface > specification > > used to request a given QOS is totally dependant on the standards body > > that > > specified the transport layer; and for this there has been little or no > > coordination. > > > > Because transport QOS over IP is based on IETF specifications, it is > > necessary that the interface used to request a given QOS conform to the > > appropriate IETF specification. If such an interface specification is > not > > available, that input might be provided to the appropriate IETF body as > to > > the requirements. > > > > I can see the endpoints negotiating the desired QOS base on need, price, > > and > > other considerations. For me, this is the limit to the application > layer > > involvement in QOS negotiation. At this point, the application > > negotiations > > with the transport provider as to the desired and available QOS. It is > up > > to the transport provider to arrange for the end-to-end QOS. (Again, it > > is > > not necessary for the transport network to know that H.323 is involved > in > > the transported data. In fact it may be encrypted in order to mask the > > presence of voice transport.) > > > > For me, this is simple. > > > > Bob > > > > ------------------------------------------------------------------ > > Robert Callaghan > > Siemens Enterprise Networks > > Tel: +1.561.923.1756 Fax: +1.561.923.1403 > > Email: Robert.Callaghan(a)ICN.Siemens.com > > ------------------------------------------------------------------ > > > > -----Original Message----- > > From: Roy, Radhika R, ALCOO [mailto:rrroy@ATT.COM] > > Sent: Friday, August 11, 2000 7:59 AM > > To: ITU-SG16(a)mailbag.cps.intel.com > > Subject: Re: H.323 QOS > > > > Hi, Mike: > > > > Let me try again. > > > > What is the reference point of H.323 QOS? Is it not H.323? If it is so, > > what > > do we mean by H.323? > > > > The answer is: Audio (different codecs), Video (different codecs), and > > Data > > (T.120 applications) that are used by H.323. > > > > What are the QOS/performance characteristics of audio, video, and data > > from > > the application point of view that is generated by audio codecs, video > > codecs, and data (T.120) applications? > > > > These QOS/performance characteristics come from the SOURCE codecs and > data > > applications. Per transport independent H.323 specifications, an enduser > > express their QOS/performance requirements on end-to-end basis purely > from > > application point of view irrespective of the transport network (e.g., > IP, > > ATM, etc.). > > > > Moreover, H.323 is meant for the packet network, not for any > > circuit-switched network like PSTN or ISDN. > > > > Let us NOT go beyond this before we start debating transport layer QOS > or > > service provider requirements. These are NOT the concern of H.323. H.323 > > is > > the transport independent application. > > > > H.323v2/v3/v4 has also provided mechanisms how RSVP and ATM QOS can be > > used > > for H.323 audio, video, and data. So, H.323 QOS that will be defined in > > H.323 Annex N MUST provide mapping for the backward compatibility. It is > a > > requirement that MUST be met per the norm of ITU-T. > > > > So, what is left for mapping? Mapping is simply a by-product of the > above > > requirement. Mapping is simply a table, nothing else. > > > > Did I miss anything? > > > > Best regards, > > Radhika R. Roy > > AT&T > > > > -----Original Message----- > > From: Mike Buckley [mailto:mikebuckley@44COMMS.COM] > > Sent: Thursday, August 10, 2000 10:19 PM > > To: ITU-SG16(a)MAILBAG.INTEL.COM > > Subject: Re: H.323 QOS > > > > > > Radhika, > > > > Thanks for the input which I welcome as I will unfortunately not be > > present > > at Portland. > > > > Let me ask a few questions and make a few comments hopefully with the > > intent > > of opening up the debate. > > > > 1. I am not sure I understand your concept of a mapping table between > the > > H.323 QOS and the transport layer QoS. My understanding is that QoS is > on > > three levels: > > > > a) that specified from a service point of view between the user and > > service > > provider (e.g PSTN quality, conference quality etc) This is the domain > of > > the speech experts and can be characterised by Listener Speech Quaklity > > (MOS), end to end delay, and absolute category rating, R. > > > > b) application specific parameters, (e.g. equipment delays, codec > choice > > and performance, codec frame size, packetisation arrangements, jitter > > buffer > > design, overall packet loss etc.) Optimisation of all these will > > determine > > what can be delivered in a). > > > > c) transport parameters for a given choice of application parameters. > > This > > boils down only to three parameters as far as I cna see: tranport > network > > delay, packet delay variation in the transport network and packet loss > in > > the transport network. Again these parameters will determine the > results > > in > > a) for a given choice of the parameters in b). These parameters are > > generic > > from the perspective of the transport network. i.e the transport > network > > does not need to know the details of the application. > > > > So the sequence of cause and effect and control is: > > > > a) User requests QoS class from service provider, > > b) Service provider determines application specific parameters in > > conjunction with users equipment and other service providers, > > c) Service provider requests required delay, delay variation and packet > > loss from network provider. > > > > I see no need for mapping here. The only QoS info flows within the > > application are specific to the application and those between the > > application (service provider) and the transport network are generic. > i.e. > > delay, jitter and packet loss. Have I missed something? > > > > 2. The issue of bit rate and media stream statistics I think need to be > > decoupled from QoS. These are specified to enable optimisation of > > resources > > within the transport network. They have no QoS significance from an > > application point of view. i.e the apllication does not care about the > > media stream bit rate and statistics but the transport network provider > > does > > as it eats up his resource. They may be used for policy enforcement > > however > > in the transport network so they do need to be agreed between service > > provider and network operator. i.e the network operator agrees to > provide > > a > > given QoS level (delay, jitter, packet loss) provided the media > properties > > are within an agreed profile (bit rate, flow statistics). > > > > 3. The next point is how can the service provider know the statistics > of > > a > > particular VBR stream? These can only be specified over a large number > of > > similar calls and will depend, for instance, on who is speaking, the > > nature > > of the speech interaction etc etc. They can only be measured not > > calculated. The service provider is in no better position to measure > > these > > than the transport network operator and, in fact, where no gateways are > > involved, may not be able to. On the other hand the class of signal > would > > have to be signalled to the network operator for him to be able to > > distinguish which class a particular measurement belonged to. e.g > > voice/speech/data, codec type, conference, multicast etc. So I see no > > purpose in trying to exchange statistics between the service provider > > (application) and transport operator. I think peak bit rate is all that > > can > > be meaningfully excanged. The specification of media class is however > > perhaps worth exploring. > > > > 4. The controlled category has always puzzled me. I only see two > > possibilities. Either the requested QoS level is guaranteed (on a > > statistical basis e.g 95% of all connections over a specified period) or > > not > > guaranteed. Is your controlled category a way of saying guaranteed, > not > > to > > 95% but to some lower figure? If you can't put a percentage on it then > it > > seems it is plain and simple not guaranteed. Anything that is not > > guaranteed to some specified statistical level is best effort and you > > can't > > say anything more about it. So I only see two categories here. > > > > In summary, I think we need to do three things in Annex N. > > > > a) Figure out the QoS information to be exchanged within the > Application > > between service providers and end users. This will go in H.225.0 and > > H.245. > > > > b) Figure out how we are going to signal QoS and media information > between > > the application (service providers) and transport domains (IP or ATM > > networks etc). The info is basically delay, jitter, packet loss > > requirements and peak bit rate. We need a protocol for this. > > > > c) we need to work out the interactions between the application QoS > > signal > > flows and the application/transport signal flows. I don't think we > need > > worry about how the transport network mechanisms assure the requested > QoS > > paramerters. RSVP/Intserv, Diffserv, MPLS, ATM, over provisioning are > all > > possibilities. > > > > Would welcome comments and views on the above. > > > > Mike > > > > > > Mike Buckley > > +44-1457-877718 (T) > > +44-1457-877721 (F) > > mikebuckley(a)44comms.com > > > > > > ----- Original Message ----- > > From: "Roy, Radhika R, ALCOO" <rrroy(a)ATT.COM> > > To: <ITU-SG16(a)MAILBAG.INTEL.COM> > > Sent: Thursday, August 10, 2000 10:15 PM > > Subject: H.323 QOS > > > > > > Hi, Mike and All: > > > > It is time to discuss about H.323 QOS. > > > > I believe that we have an agreement as follows: > > > > · H.323 QOS MUST be backward compatible to support RSVP and ATM QOS as > it > > exists for H.323v2/v3/v4 > > · Like H.323 spec, the application level H.323 QOS MUST be independent > of > > the transport layer QOS and should support all transport networks (e.g., > > IP, > > ATM) > > · A mapping table between the H.323 QOS and the transport layer QOS > (e.g., > > IP QOS [DiffServ, RSVP, etc.], ATM QOS [CBR, rt-VBR, nrt-VBR, ABR, > etc.]) > > should be provided. > > > > From the H.323 multimedia application point of view, there are following > > performance parameters can be used to characterize the traffic > > characteristics: > > > > · Bitrate characteristics: Peak bit rate (PBR) or peak rate (PR), > > Sustained > > bit rate (SBR) or average rate (AR), minimum bit rate (MBR) or minimum > > rate > > (MR), and mean bust size (MBS) > > · Delay and loss characteristics: end-to-end delay (EED) or delay, > > end-to-end delay variation (EEDV) or delay variation (DV), and > > bit-error-rate (BER) or (packet) loss rate (LR) > > > > We can now form a table with all parameters as follows: > > > > Table 1: H.323 Multimedia Application Performance Matrix > > Audio (codecs)--- Video (codecs)--- Data > > (T.120) > > PBR/PR Yes/No/Value Yes/No/Value Yes/No/value > > SBR/AR Yes/No/Value Yes/No/Value Yes/No/value > > MBR/MR Yes/No/Value Yes/No/Value Yes/No/value > > MBS Yes/No/Value Yes/No/Value Yes/No/value > > EED/Delay Yes/No/Value Yes/No/Value Yes/No/value > > EEDV/DV Yes/No/Value Yes/No/Value Yes/No/value > > BER/LR Yes/No/Value Yes/No/Value Yes/No/value > > > > From the above table we will have the opportunity to choose each > parameter > > for each medium (audio, video, data) that makes sense from the > > application's > > and enduser's point of view. Again, these parameters can be specified as > > follows: > > > > · Guaranteed: The value specified for each parameter MUST be guaranteed. > > · Controlled: The value specified for each parameter MAY be satisfied as > > far > > as practicable (possibly with certain range), but definitely NOT > > guaranteed. > > · Best effort: No commitment will be made. > > > > Now each medium (e.g., audio, video, or data) will have different > > categories > > of performance matrix depending on its selection criteria and this can > > also > > be mapped to RSVP, ATM QOS, and others, if needed. > > > > Once we agree on this format, the next step is to create H.323 QOS > > signaling > > messages. > > > > This is my input for discussion in the upcoming Portland Q.13 meeting > for > > H.323 QOS. > > > > I like to see the comments from other members as well. > > > > Best regards, > > Radhika R. Roy > > AT&T > > +1 732 420 1580 > > rrroy(a)att.com > > > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > For help on this mail list, send "HELP ITU-SG16" in a message to > > listserv(a)mailbag.intel.com > > > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > For help on this mail list, send "HELP ITU-SG16" in a message to > > listserv(a)mailbag.intel.com > > > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > For help on this mail list, send "HELP ITU-SG16" in a message to > > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > For help on this mail list, send "HELP ITU-SG16" in a message to > listserv(a)mailbag.intel.com > >

1 0

Re: [IPTEL] URL to support Number Portability [was: draft-yu-tel- url-00.txt]
by James Yu 16 Aug '00

16 Aug '00

Would appreciate if people on the iptel and SIP discussion lists can voice their opinions (agree or disagree) on draft-yu-tel-url-00.txt (see attachment). This enhancement to the tel URL is intended to allow SIP to carry number portability related information in the Request-URI field. The information (e.g., routing number and POTS number) in the tel URL may be used by the LSs for call routing. Thanks. James > -----Original Message----- > From: Orit Levin [mailto:orit@radvision.com] > Sent: Tuesday, August 15, 2000 1:06 PM > To: James Yu; IPTEL WG; sip-h323(a)egroups.com > Cc: h323implementors(a)imtc.org; ITU-SG16(a)MAILBAG.INTEL.COM; > SIP reflector > Subject: [IPTEL] URL to support Number Portability [was: > draft-yu-tel-url-00.txt] > > > James! > After reviewing your draft, below are my comments. > As in regards to H.323, H.323 doesn't currently have an > explicit notion of > two different fields carrying the exact meaning of "current > destination" and > "the final recipient", using SIP language. > On the other hand, the definition of H.323 URL potentially allows for > cascading of URLs > of different kinds. This concept appear in the last draft of > "Annex O" to be > discussed next week. Therefore, tel url can be placed in the > "user" part of > H.323 URL. > Since, in your proposal, the original "final recipient" > information (i.e. > the phone number itself) is preserved in the tel url after > the NPDB deep is > performed, I don't see a technical problem for using the > proposed algorithm > within H.323. > > I will bring your work to the attention of H.323 community next week, > including the H.323 mobility WG. My question (and that of the H.323 > community) would be: what level of consensus does the > approach, presented in > your draft, currently have within SIP and IPTEL working groups? > > > Orit Levin > RADVision Inc. > 575 Corporate Drive Suite 420 > Mahwah, NJ 07430 > Tel: 1 201 529 4300 (230) > Fax: 1 201 529 3516 > www.radvision.com > orit(a)radvision.com > -----Original Message----- > From: James Yu <james.yu(a)neustar.com> > To: 'Orit Levin' <orit(a)radvision.com>; IPTEL WG > <iptel(a)lists.bell-labs.com> > Date: Monday, August 14, 2000 1:24 PM > Subject: RE: [IPTEL] A draft of H323-URL to be presented > during the IPTEL > session > > > >Orit, > > > >Please review the attached draft-yu-tel-url-00.txt, titled > "Extensions to > >the tel and fax URL to support Number Portability." It is > intended for the > >SIP to carry number portability (NP) information in the > Request URI field. > >It may be carried by H.225.0 directly or may be mapped to > the proposed > H.323 > >URL to support NP in H.323. > > > >There has not been lots of discussion on this I-D. Comments > from you and > >others are welcome. > > > >James > > > > > > > >-----Original Message----- > >From: Orit Levin [mailto:orit@radvision.com] > >Sent: Monday, July 31, 2000 9:05 PM > >To: IPTEL WG > >Subject: [IPTEL] A draft of H323-URL to be presented during the IPTEL > >session > > > > > >Hello! > >I have submitted the attached draft 3 weeks ago. > >Unfortunately, the document doesn't appear on the IETF Drafts site. > >Since this topic is going to be presented during the IPTEL session on > >Thursday morning, I am attaching the draft for you review. > >Best Regards, > >Orit Levin > >RADVision Inc. > >575 Corporate Drive Suite 420 > >Mahwah, NJ 07430 > >Tel: 1 201 529 4300 (230) > >Fax: 1 201 529 3516 > >www.radvision.com <http://www.radvision.com> > >orit(a)radvision.com <mailto:orit@radvision.com> > >-----Original Message----- > >From: Orit Levin < orit(a)radvision.com <mailto:orit@radvision.com> > > >To: internet-drafts(a)ietf.org <mailto:internet-drafts@ietf.org> < > >internet-drafts(a)ietf.org <mailto:internet-drafts@ietf.org> > > >Date: Tuesday, July 11, 2000 1:59 AM > >Subject: Request for Draft Submission > > > > > >Hello! > >I would like to submit the attached paper as the first > drafts to become an > >Informational RFC. > >Thank you, > >Orit Levin > >RADVision Inc. > >575 Corporate Drive Suite 420 > >Mahwah, NJ 07430 > >Tel: 1 201 529 4300 (230) > >Fax: 1 201 529 3516 > >www.radvision.com <http://www.radvision.com> > >orit(a)radvision.com <mailto:orit@radvision.com> > > > > > > > > _______________________________________________ > IPTEL mailing list > IPTEL(a)lists.bell-labs.com > http://lists.bell-labs.com/mailman/listinfo/iptel >

1 0

← Newer
1
2
3
4
5
6
7
...
15
Older →

Jump to page: