Art S R wrote:
> On Thu, Mar 19, 2009 at 9:21 AM, Duncan Webb <dun...@freevo.org> wrote:
>> Art S R wrote:
>>> On Wed, Mar 18, 2009 at 12:32 PM, Duncan Webb <dun...@freevo.org> wrote:
>>>
>>>> I've been working on the recordserver to see if I can improve the
>>>> timings of the recording. If you want to try out this version you can
>>>> get it from svn with the export command, i.e.:
>>>> svn export
>>>> svn://svn.freevo.org/freevo/branches/rel-1/freevo/src/helpers/recordserver.py
>>>> and you will need freevo_config.py from
>>>> svn://svn.freevo.org/freevo/branches/rel-1/freevo/freevo_config.py
>>>>
>>>> You will need to change the RECORDSERVER_ATTIMER = 0 in your
>>>> local_conf.py to start the recordings at the top of the minute.
>>>>
>>>> Attached is a log of four consecutive without any gaps and including a
>>>> small overlay of two programs.
>>>>
>>>> Actually there is an odd delay after a kaa.rpc time-out error, I'm not
>>>> sure if this has been caused by this error.
>>>>
>>>> Duncan
>>> Thanks for the quick fix, Duncan.  I tried the new recordserver.py and
>>> freevo_config.py, but got the following errors when I tried to launch
>>> recordserver:
>>>
>>> Traceback (most recent call last):
>>>   File "/usr/lib/python2.5/site-packages/freevo/helpers/recordserver.py",
>>> line 1481, in <module>
>>>     main()
>>>   File "/usr/lib/python2.5/site-packages/freevo/helpers/recordserver.py",
>>> line 1454, in main
>>>     rpc.register(recordserver)
>>> AttributeError: 'Server' object has no attribute 'register'
>>> WARNING: 'Server' object has no attribute 'register'
>>>
>>> Any idea what may be the problem?
>> Ah yes, this is part of the kaa that is only in svn. You will need to
>> install these kaa packages and freevo-1 from svn.
>>
>> Do you know how to do this? If not then use the wiki page
>> http://doc.freevo.org/SourceSVNInstallation.
>>
>> Duncan
>>
> 
> Last night, I was looking at an earlier version of recordserver.py
> that did not have this error and saw that it was using
> "rpc.connect(recordserver)" instead of "rpc.register(recordserver)".
> On a whim, I changed that statement in you lastest svn version and, lo
> and behold, recordserver was able to launch successfully and record
> programs.  However, when I recorded back-to-back programs, the second
> one was still starting 1 minute late:
> 
> 2009-03-18 22:29:50,810 DEBUG    generic_record.py (157):
> Record_Thread::run: mode=idle
> 2009-03-18 22:29:51,643 DEBUG    recordserver.py (1438):
> RecordPostProcess.run() finish
> 2009-03-18 22:30:00,004 DEBUG    recordserver.py (921): overlap_duration=0
> 2009-03-18 22:30:00,006 INFO     recordserver.py (934): CALLED RECORD
> STOP 1: Wed Mar 18 22:00->22:30 (22:00)  I33.28460028.microsoft.com
> ALF
> 2009-03-18 22:30:00,007 INFO     recordserver.py (954): delaying: Wed
> Mar 18 22:30->23:00 (22:30)  I47.28458302.microsoft.com The
> Exterminators
> 2009-03-18 22:30:00,512 DEBUG    generic_record.py (157):
> Record_Thread::run: mode=stop
> 2009-03-18 22:30:01,016 DEBUG    generic_record.py (157):
> Record_Thread::run: mode=idle
> 2009-03-18 22:31:00,003 INFO     recordserver.py (956): going to
> record: Wed Mar 18 22:30->23:00 (22:30)  I47.28458302.microsoft.com
> The Exterminators
> 2009-03-18 22:31:00,005 INFO     recordserver.py (974): start
> recording: Wed Mar 18 22:30->23:00 (22:30)  I47.28458302.microsoft.com
> The Exterminators
> 
> Program 1 completed post-processing at 22:29:51,643 with CALLED RECORD
> STOP at 22:30:00,006 and, even though overlap_duration=0, program 2
> did not start recording until a full minute later at 22:31:00,005.
> Seeing that generic_record.py was involved, I checked svn and saw that
> you had updated this too for the recording time issue, so I grabbed
> that file as well to see if it would make any difference.
> Unfortunately, the 1 minute delay between program recordings was still
> there.

Trouble is that this happened sometimes and sometimes not. Anyway it's a
pain fixing the recordserver as it takes lots of time testing...

In svn there is a new version which should fix this and here is the log:
21:13:54,756 INFO     RECORD_START Thu Mar 19 20:15->21:15 (20:15)  Mein
neues Leben
21:14:45,000 INFO     currently_recording=Thu Mar 19 20:15->21:15
(20:15)  Mein neues Leben
21:14:45,005 INFO     CALLED RECORD STOP 1: Thu Mar 19 20:15->21:15
(20:15)  Mein neues Leben
21:14:45,007 INFO     delaying: Thu Mar 19 21:15->22:15 (21:15)  Hagen
hilft!
21:14:45,939 INFO     RECORD_STOP Thu Mar 19 20:15->21:15 (20:15)  Mein
neues Leben
21:14:45,941 INFO     post-processing started for Thu Mar 19
20:15->21:15 (20:15)  Mein neues Leben
21:14:46,438 INFO     RECORD_START Thu Mar 19 21:15->22:15 (21:15)
Hagen hilft!
21:14:53,224 INFO     post-processing finished for Thu Mar 19
20:15->21:15 (20:15)  Mein neues Leben
21:15:45,006 INFO     currently_recording=Thu Mar 19 21:15->22:15
(21:15)  Hagen hilft!

The above shows that:
 - the recording was started less than a second after the last one was
finished.
 - the post-processing was done in its own thread

> When I get some more time and am feeling more ambitious, I'll install
> the svn versions of kaa and freevo-1 and see if that makes any
> difference.

Good idea... :-)

Duncan

------------------------------------------------------------------------------
Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are
powering Web 2.0 with engaging, cross-platform capabilities. Quickly and
easily build your RIAs with Flex Builder, the Eclipse(TM)based development
software that enables intelligent coding and step-through debugging.
Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com
_______________________________________________
Freevo-users mailing list
Freevo-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freevo-users

Reply via email to