Migrating from 0.10.x to 1.x

[Edward Hervey]

Hi,


On Tue, 2014-06-17 at 20:37 +0400, Kirill Kirichenko wrote:
> Hi there.
> 
> In my company's product we use gstreamer 0.10.35 and plugins based on
> it.
> 
> We compile the gstreamer code ourselves and deploy the bits with the
> rest of the product's code. We use very limited set of standard
> plugins - base, good, some ugly and not all of them.
> 
> Now our management decided that it's a good idea to switch to 1.x
> 
> What are the good reasons to do so ?
> 
> I can think of:
> 
> 1) more bugfixes in the code

  Essentially 3 years of bugfixes, improvements and speedups. There has
been *thousands* of commits since.

> 
> 2) refined architecture. I kinda understand that but what exactly has
> become better?

  At the application API level there hasn't been that many changes (i.e.
it won't take you that long to port your application). The
design/philosophy/usage stays the same. There might have been some
slight API changes, but they are well documented (see links below).

  Where the most happened is the underlying (core/element level)
improvements. Essentially you will end up with more optimized/reliable
configuration and data passing.
  * Better memory usage (less copies, usage of target memory/layout
whenever possible, ...)
  * Better negotiation, and reliable/faster re-negotiation
  * From the above, you will end up more often with the most optimal
data/processing path between the various elements in your pipeline
  * Usage of HW-accelerated components/systems (without having to
custom-craft your pipelines) when possible
  * And obviously the past 3 years of improvements/speedups

> 
> 3) better compatibility with the latest plugins available in vanilla
> GStreamer.

  You'll indeed get more support from the gst community if you use
latest stable series :)

> 
> 
> Any more ideas ?

  Depends on your application/usage, but there are also new
features/elements available. At the top of my head: DASH support, GL
support, ...

> 
> I'm not trying to convince them, I just to know why and how difficult.
> It's gonna be a real huge task of upgrading.

  If you just have an application and don't have custom plugins, it's
not that hard. The concept and design at the application level have
stayed the same.
  If you have custom plugins it's a bit more work. Check if there isn't
already a new element that does the same thing upstream. If it's a
custom plugin that could benefit the community, maybe it would be
worthwile proposing it upstream (in which case the community could help
you port it).

  More info:
http://gstreamer.freedesktop.org/data/doc/gstreamer/head/manual/html/chapter-porting-1.0.html
http://cgit.freedesktop.org/gstreamer/gstreamer/plain/docs/random/porting-to-1.0.txt

  Don't hesitate to ask if you encounter issues,

    Edward
> 
> 
> Thanks,
> 
> K
> 
> _______________________________________________
> gstreamer-devel mailing list
> gstreamer-devel at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: This is a digitally signed message part
URL: <http://lists.freedesktop.org/archives/gstreamer-devel/attachments/20140618/b1ac74fb/attachment.sig>