[reSIProcate] Invite state machine errors or errorsinClientInviteSession?

[Scott Godin]

Thanks Micky,

 

I've reviewed all your comments on the UAS and UAC states.  I've fixed
either the .dot or sources appropriately (I hope).   Changes are in SVN.

 

A lot of the UPDATE in early dialog code and PRACK code is not completed
- so there are some unimplemented states and transitions.  I elected to
leave out most of the transitions to Assert state for simplicity.

 

Let me know if I missed something.  Thanks,

 

Scott

 

________________________________

From: resiprocate-devel-bounces at list.sipfoundry.org
[mailto:resiprocate-devel-bounces at list.sipfoundry.org] On Behalf Of
Micky Kaufmann
Sent: Sunday, September 18, 2005 10:52 AM
To: resiprocate-devel at list.sipfoundry.org
Subject: RE: [reSIProcate] Invite state machine errors or
errorsinClientInviteSession?

 

Attached another file for 'dum-serverinvite-state.dot'.

 

By the way the comparison between the state machines and the source was
done on svn version 5412 checkout.

 

________________________________

From: resiprocate-devel-bounces at list.sipfoundry.org
[mailto:resiprocate-devel-bounces at list.sipfoundry.org] On Behalf Of
Micky Kaufmann
Sent: Sunday, September 18, 2005 10:53 AM
To: Scott Godin; resiprocate-devel at list.sipfoundry.org
Subject: [reSIProcate] Invite state machine errors or errors
inClientInviteSession?

 

Hi All,

 

I've attached an RTF file containing differences between the source code
and the file 'dum-clientinvite-state.dot'.

 

You can see in the attached file that in some transitions the
description of what to do is missing a few methods or a method was added
to it.

For example: some transitions from the start event are missing the
'app::onNewSession' method, but there's a transition from Early to
EarlyWithOffer that contains 'app::onNewSession' even though it seems an
error.

 

In some cases the event causes a transition from one event to another
when the second state seems wrong!

Some transitions are missing and I couldn't find some of the transitions
in the source code.

 

Since I'm not always sure if the fix for an error should be in the
source code or in the state machine (like in the 'app::onNewSession'
example)  I'm posting it to this newsgroup...

 

For now I only reviewed the client state machine and since I'm pretty
sure the document I'm sending will solve some bugs I'll continue to
review the other invite state machines.

 

 

By the way I also attached the draft drawings of the state machines -
all in one visio file.

 

 

Regards,

 

Micky

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.resiprocate.org/pipermail/resiprocate-devel/attachments/20050919/b5ea4593/attachment.htm>