RE: WG Last Call on SUA Implementors Guide completes

This document is in its 5th iteration. 
It has some changes on some of the points you mentioned(and probably not
to your liking)  
That means that it has been read and reread... I think that this means
that it has been reviewed...

This document is ready for AD review...


>-----Original Message-----
>From: Brian F. G. Bidulock [mailto:[email protected]] 
>Sent: dinsdag 12 december 2006 8:57
>To: Ong, Lyndon
>Cc: [email protected]
>Subject: Re: [Sigtran] WG Last Call on SUA Implementors Guide completes
>
>Lyndon,
>
>Do you think that this document received any serious review?  I doubt
>that even the editor read through it.

You have been one of "greatest" commenters on it...
Before the last call, you suggesting to put it off untill the next
SIGTRAN bakeoff...
And the date for that has moved backwards since then...
There is no reason to delay any further...


>Quoted below is Section 2.7 from the document:
>
>" 2.7  Parameters should only occur once in a message
>" 
>" 2.7.1  Description of the problem
>" 
>"    Parameters in SUA messages can be repeated as many times as
possible.
>"    This would lead to inconsistencies, such as 2 or more destination
>"    addresses.
>" 
>" 2.7.2  Text changes to the document
>" 
>"    ---------
>"    Old text: (Section x.x)
>"    ---------
>" 
>" 
>"    ---------
>"    New text: (Section x.x)
>"    ---------
>" 
>" 2.7.3  Solution description
>" 
>"    Unless explicitly stated or shown in a message format diagram,
only
>"    one parameter of the same type is allowed in a message.
>

This is completely new text. There was no old text to replace. 




>--brian
>
>Lyndon Ong wrote:                             (Mon, 11 Dec 2006
19:55:10)
>> Hi Brian,
>> 
>> I take it you must have missed the WG Last Call announcement!  In any
>> case, we
>> should take a look at your comments.  More detail would probably be
>> helpful.
>> 
>> Cheers,
>> 
>> Lyndon 
>> 
>> -----Original Message-----
>> From: Brian F. G. Bidulock [mailto:[email protected]] 
>> Sent: Monday, December 11, 2006 11:44 AM
>> To: Ong, Lyndon
>> Cc: [email protected]
>> Subject: Re: [Sigtran] WG Last Call on SUA Implementors Guide
completes
>> 
>> Lyndon,
>> 
>> I do not agree that the document has had sufficient review for it to
>> advance.
>> 
>> Scanning backwards through the document:
>> 
>> Section 2.14 needs more work.  Section 2.12 is very inconsistent with

The Stream use of those SSNM msg is now consistent...

>> M3UA.  Sections 2.8, 2.9 and 2.10 do not address the comments on the
Yet another change for section 2.12. please send text...

The sections 2.8, 2.9 and 2.10 have always been problem childs. 
They were put in at the last minute and nobody offered text on what
should be 
done if some SG was to receive it... We have been iterating over it for
multiple times...
It was put in on demand of the ASP vendors... Unfortunaly it is the SG
which has 
to do the dirty work and most SG vendors are not in the least bit
interested 
in doing this because it makes the implementation a lot harder....
And most of the comments have been over these sections...(proving the
point that it
should have never been included in the original SUA spec in the first
place: 
it should been put in a extra draft on its own..)


>> mailing list.  Section 2.7 is incomplete (the old and new text
sections
>> are completely incomplete).  Section 2.6 should remove the useless
text

Section 2.6 looks fine to me....

>> (nowhere is a "response" defined).  Section 2.4 needs more work to be
>> consistent with the necessary order arrived at for M3UA.  Section 2.3

Section 2.3 is a addition. The problem of the original sequence
parameter is solved  by it.
So it might pose problems for implementations not wanting to implement
this...

>> introduces an incompatiblity with existing implementations.  Section
2.2
>> needs more work: ANSI does not use
>> 4 congestion levels for SCCP related SCON messages.

ANSI does use 4 congestion levels... So SUA has to provide for this from

wherever it can find the information...

>> 
>> --brian
>> 



Yours sincerely,
Lode Coene