Interesting results. Go ahead... :-)
Two rulesets are due to their separation in XCAP/authorization related
drafts, but you are right, in one file it could be better so I will
modify the parser to accept more rulesets in one file...
Vaclav
On Tue, May 16, 2006 at 09:44:56AM +0200, samuel wrote:
Inline...
2006/5/16, Vaclav Kubart <vaclav.kubart(a)iptel.org>rg>:
First of
all, I have to thank you for the time you spent writing the
handbook, it's really really helpfull....I wish all SER related parts
had this docs..
Thanks. :-) It is nice to hear something like that.
I'll try to get familiar with the code of the notifications and I'll
try to find something....which I don't thing so :P. I'll also merge
the two functionalities (proxy + presence) in a unique config file to
see if it works.
I hope I can provide more info these following days.
Thanks.
I've merged the configs without much success...I still have the
problem of sending the NOTIFY but in another function:
May 15 20:39:07 localhost /usr/local/sbin/ser[19354]: ERROR: uri2sock:
no corresponding socket for af 2
May 15 20:39:07 localhost /usr/local/sbin/ser[19354]: ERROR:
euac_funcs.c:219: BUG: can't send SUBSCRIBE without contact
Internally, though I thing everything goes to the same point, when
trying to set the dest_info structure in tm/ut.h, there's some problem
(which I have no clue yet why) when selecting the socket and I think
that the condition that always raise the errors when sending notifies
is:
dst->send_sock==0
This is just the first impression I have...let's see if I can find
something more usefull today or I'll run out of time :(
There's the "CVS log"
2006-04-13 added uri2dst(), simplified uri2sock() (andrei)
So we can ask Andrei wether something has changed this last 2 months
that can affect sending the requests from the uac-presence connection.
>>
>> About the missing things in the presence handbook, probably the most
>> important is the new xcap module because in the sample config files
>> it's missing.
>You are right, but in the
"compiled" version of presence handbook
>(published on iptel's ftp) is described current presence snapshot
>which doesn't have xcap module.
>In the source version of the handbook
(in directory doc/presence of SER
>source tree) is the description still missing too, but will be added
>soon. :-)
>> Another thing is that in the XCAP
structure description, the im-rules
>> directory is missing, which might lead to misunderstandings. I
>> downloaded the structure from the iptel's ftp and inside the im-rules
>> there were several files corresponding to presence-rules which should
>> be either removed or updated with the im-rules namespaces and removing
>> the whitelist.
>Thanks! I will correct it.
>By the way, "im-rules" are
NOT standardized in any way - we (at iptel)
>only needed something like that, so it is there...
I don't know wethere it's required to have two different rulesets but
if you have required it I just haven't faced yet the use case so I
guess it's the way to go...
> Vaclav
>>
>> Thanks,
>>
>> Samuel.
>>
>>
>>
>>
>> 2006/5/15, Vaclav Kubart <vaclav.kubart(a)iptel.org>rg>:
>> >Hi,
>> >this problem I'm trying to solve with Ilker Aktuna. I try to simulate
it
>> >on my machine and let you know. Or if you solve it, please let me know.
>> >:-)
>>
>> >Please, could you
tell me, what things you were missing in presence
>> >handbook? I'm trying to do it as useful as possible and whatever ideas
>> >are welcome...
>>
>> > Vaclav
>>
>> >On Mon, May 15, 2006
at 01:38:02PM +0200, samuel wrote:
>> >> Hi all,
>> >>
>> >> I recently had a few hours and start installing the presence staff and
>> >> I have to say that I have it amost workign thanks to the presence
>> >> handbook, the mailing list and, obviously, a little bit of code
>> >> review..:P
>> >>
>> >> I have two SER instances, the "proxy" and the "presence
server" (both
>> >> with last CVS code) co-located in the same host and I have an issue
>> >> when the "presence server" tries to send the NOTIFY requests.
Below
>> >> there's an attched log showing the problem (on IP a.b.c.d I've
got the
>> >> two instances):
>> >>
>> >> 3(30682) DEBUG notify.c:378: sending winfo notify
>> >> 3(30682) DEBUG notify.c:383: winfo document created
>> >> 3(30682) DEBUG notify.c:391: creating headers
>> >> 3(30682) DEBUG notify.c:398: headers created
>> >> 3(30682) DEBUG:tm:t_uac:
>> >> next_hop=<sip:a.b.c.d;transport=tcp;ftag=c77b3f33;lr=on>
>> >> 3(30682) t_uac: no socket found
>> >> 3(30682) DEBUG notify.c:402: request sent with result -7
>> >> 3(30682) ERROR: notify.c:404: Can't send watcherinfo notification
(-7)
>> >>
>> >> This problem appears in other places, not only in the notifications
>> >> for winfo so probably there's somthing in the selection of the
>> >> outgoing socket directing to the local IP.
>> >>
>> >> >From the proxy part I just ust t_forward_nonack for the
"SIMPLE"
>> >> messages with record route....maybe adding the port in the record
>> >> route should help?
>> >>
>> >>
>> >> Samuel.
>> >>
>> >> _______________________________________________
>> >> Serusers mailing list
>> >> serusers(a)lists.iptel.org
>> >>
http://lists.iptel.org/mailman/listinfo/serusers
>>
>