Hi Christian

Thanks, I understand now the definition of vwp/wxp and vwp/wyp.

Regarding to the editorial problem. In the version, which I have (which is TD33(PLEN) from Geneva Jan 2004 meeting) in the figure 8 the text says:

Property Group 1 vwp/vid=1,vwp/wseq=1,vwp/wxp=2500,vwp/wyp=0, vwp/wh=7500,vwp/wl=5000,vwpcvsp/ovs=2 Property Group 2 vwp/vid=2,vwp/wseq=2,vwp/wxp=5000,vwp/wyp=2500, vwp/wh=7500,vwp/wl=5000,vwpcvsp/ovs=1

Regards, Sasha

-----Original Message----- From: Christian Groves [mailto:christian.groves@ERICSSON.COM] Sent: Sunday, June 19, 2005 10:21 PM To: Sasha Ruditsky Cc: itu-sg16@external.cisco.com Subject: Re: The question about extension of an H.248 package.

Hello Sasha,

Please see my response below.

Regards, Christian

Sasha Ruditsky wrote:

Hi Christian

Thanks, again.

It looks like I will have some more question about H.248.19.

Meanwhile.

In H.248.19 defines vwp/wxp and vwp/wyp as being from 0 to 10000.

The description text states: This property is set by the MC to represent the horizontal "X" position of the bottom left hand corner of a window. 0 represents the left hand side of a screen, 10000 the right hand side

of a screen.

So it makes the size of the screen equal to 10001 units.

In the same time vwp/wh and vwp/wl are also defined to be 0 to 10000.

I think it means that it is impossible to create window which fills the entire screen. (Screen is 10001 units and the biggest window is only 10000).

[CHG] Given a display 10000 by 10000 pixels the way I saw the property working was that 0,0 represented the bottom left hand corner of the display/window. A window position 0,0 size 1,1 would draw a pixel 1,1 starting at this point. A window placed at 10000,10000 couldn't actually have a positive size as it would exceed the display.

Do I misunderstand the description text? Shouldn't the vwp/wxp and vwp/wyp be defined in range from 0 to 9999? Or probably the range is OK, however the 10000 is outside of screen?

[CHG] So technically 10000 is outside the meaningful range for displaying anything.

What do you think?

Another small editorial issue is that in H.248.19 on Figure 8 the text

references the vwpcvsp/ovs, where I believe it should be cvsp/ovs.

[CHG] Is the editorial issue in Figure 8 itself? It should be cvsp/ovs but I can't see where the error occurs.

Thanks, Sasha

-----Original Message----- From: Christian Groves [mailto:christian.groves@ericsson.com] Sent: Thursday, June 16, 2005 8:55 PM To: Sasha Ruditsky Cc: itu-sg16@external.cisco.com Subject: Re: The question about extension of an H.248 package.

Hello Sasha,

In thus case the text encoding would use H.248.15 Gateway control protocol: SDP H.248 package attribute.

Regards, Christian

Sasha Ruditsky wrote:

...

Hi Christian

Thanks.

One more question. In H.248.19 most of the properties inside packages are specified to be

...

"Defined in: Local/Remote".

...

From what I know about H.248 -- local/remote descriptors contain SDP

messages.

How do the H.248.19 properties coexist with SDP? Is it only possible in binary encoding mode?

Thanks, Sasha

-----Original Message----- From: Christian Groves [mailto:christian.groves@ERICSSON.COM] Sent: Thursday, June 16, 2005 12:35 AM To: Sasha Ruditsky Cc: itu-sg16@external.cisco.com Subject: Re: The question about extension of an H.248 package.

Hello Sasha,

There is no hard rule of what to propose. It really depends on the type of information you want added.

I think the only time the package extension should be used is if the added properties/signals/events provide a set of functions of their own right. e.g. would it makes sense that these properties could be a package by themselves but need another package to operate. For example: The tone gen package. The DTMF was added as a extension package as DTMF generation is its own subset functionality but is related to the generation of tones.

Regards, Christian

Sasha Ruditsky wrote:

...

Hi

I'm working on a proposal for July meeting to add some new properties

...

...

to one of the H.248.19 packages.

...

From H.248 document it appears that there are defined (at least) two

ways of doing this.

• Creating a new version of an existing package
• Creating a new package that extends the existing one.

Can somebody please comment on the factors helping to determine which

...

...

of these two mechanisms should be uses?

Thanks, Sasha .