rant...

[S Boucher]

Or why trying to do something with gstreamer is a threat to your sanity...

There are some mailing lists, but no one seems to read them, so that
leaves you with the documentation.

How's the documentation?  Vague at best.

Example 1:

The manual entry for gst_event_new_new_segment() says of the 'update'
argument "is this segment an update to a previous one".  The obvious
question for the reader is then: when should I set this to true and
when should I set this to false?  Or put another way, why do we even
need that flag?  This may be obvious to someone who spends his/her
time reading/writing gstreamer code.  Less so for someone trying to
understand how to use gstreamer.

Example 2:

gst_event_new_flush_start()/gst_event_new_flush_stop().  The documentation
seems straight forward enough, except its not clear whether there are
counter indications, or situations when this should or should not be
used.  You just try to use it, and if it seems to work, then you
assume everything's fine.  If it doesn't work - as is my case - you
scratch your head.  You sure can't refer to any worthwhile
documentation.  There's also no point in going to the mailing list for
the reason stated above.  There are always the GST_DEBUG traces, but
they are for the most part useless unless you are intimate with the
code emiting the traces.  You are left with digging deep into the guts
of gstreamer, not knowing exactly what to look for.  All of this
because the documentation is, hmmm, vague.


When I started to work with gstreamer, I tought it was cool what you
could do with it.  But after trying to use it beyond trivial cases,
gstreamer has most definitely succeeded in cooling any enthousiasm I
have towards it.

Now back to banging my head against the wall as to why I hear nothing
after flushing/sending a new segment, despite pumping data into the
pipeline :-)



      __________________________________________________________________
Yahoo! Canada Toolbar: Search from anywhere on the web, and bookmark your favourite sites. Download it now
http://ca.toolbar.yahoo.com.

------------------------------------------------------------------------------
SOLARIS 10 is the OS for Data Centers - provides features such as DTrace,
Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW
http://p.sf.net/sfu/solaris-dev2dev
_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gstreamer-devel

[Sean McNamara-4]

Hi,


On Thu, Feb 11, 2010 at 7:03 PM, S Boucher <[hidden email]> wrote:
> Or why trying to do something with gstreamer is a threat to your sanity...
>
> There are some mailing lists, but no one seems to read them, so that
> leaves you with the documentation.

This mailing list is read and there's quite a bit activity on it,
based on what I read in my gmail inbox every day.

>
> How's the documentation?  Vague at best.

In some places, I have to agree. You're always welcome to improve it
and submit patches based on what you learn about the behavior of the
functions.

>
> Example 1:
>
> The manual entry for gst_event_new_new_segment() says of the 'update'
> argument "is this segment an update to a previous one".  The obvious
> question for the reader is then: when should I set this to true and
> when should I set this to false?  Or put another way, why do we even
> need that flag?  This may be obvious to someone who spends his/her
> time reading/writing gstreamer code.  Less so for someone trying to
> understand how to use gstreamer.
>
> Example 2:
>
> gst_event_new_flush_start()/gst_event_new_flush_stop().  The documentation
> seems straight forward enough, except its not clear whether there are
> counter indications, or situations when this should or should not be
> used.  You just try to use it, and if it seems to work, then you
> assume everything's fine.  If it doesn't work - as is my case - you
> scratch your head.  You sure can't refer to any worthwhile
> documentation.  There's also no point in going to the mailing list for
> the reason stated above.  There are always the GST_DEBUG traces, but
> they are for the most part useless unless you are intimate with the
> code emiting the traces.  You are left with digging deep into the guts
> of gstreamer, not knowing exactly what to look for.  All of this
> because the documentation is, hmmm, vague.
>
>
> When I started to work with gstreamer, I tought it was cool what you
> could do with it.  But after trying to use it beyond trivial cases,
> gstreamer has most definitely succeeded in cooling any enthousiasm I
> have towards it.
>
> Now back to banging my head against the wall as to why I hear nothing
> after flushing/sending a new segment, despite pumping data into the
> pipeline :-)

Instead of just vaguely describing your problem in passing, we would
be more likely to be able to help you with your problem if you would
do the very things you complain about (such as enable GST_DEBUG) and
give us some program output and/or code. The more we understand about
your situation, the better equipped we would be to point out problems
in your code or spot a bug. I'm not saying you need to have these
capabilities on your own, but at least use some common sense and,
instead of ranting, come to us with a detailed description of your
problem, what you tried, what exactly failed, and what you expected it
to do -- and then post the relevant code sections. See
http://catb.org/~esr/faqs/smart-questions.html for more advice.

-Sean

------------------------------------------------------------------------------
SOLARIS 10 is the OS for Data Centers - provides features such as DTrace,
Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW
http://p.sf.net/sfu/solaris-dev2dev
_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gstreamer-devel

[Tim-Philipp Müller-2]

On Thu, 2010-02-11 at 16:03 -0800, S Boucher wrote:

> The manual entry for gst_event_new_new_segment() says of the 'update'
> argument "is this segment an update to a previous one".  (...)
>  (...)
> gst_event_new_flush_start()/gst_event_new_flush_stop(). The documentation
> seems straight forward enough, except its not clear whether there are
> counter indications, or situations when this should or should not be
> used.  You just try to use it, and if it seems to work, then (...)

The API reference is not (meant to be) comprehensive. It is not quite
clear from your mail and your previous mail if you have seen the design
docs at

http://cgit.freedesktop.org/gstreamer/gstreamer/tree/docs/design/

If there are specific things that you think are not explained, badly
explained or could use some more detail, please file a bug for each
issue so we can improve the docs.

If no one has answered to your mail after some time, a quick follow-up
message is likely to be more effective than a "rant".

The #gstreamer IRC channel is also a good place to get help.

 Cheers
  -Tim



------------------------------------------------------------------------------
SOLARIS 10 is the OS for Data Centers - provides features such as DTrace,
Predictive Self Healing and Award Winning ZFS. Get Solaris 10 NOW
http://p.sf.net/sfu/solaris-dev2dev
_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/gstreamer-devel