Re: [Patch] gst-plugins-base

[Efraim Flashner]

On Thu, 31 Mar 2016 13:17:42 -0400
Mark H Weaver <address@hidden> wrote:

> Efraim Flashner <address@hidden> writes:
> 
>  [...]  
>  [...]  
>  [...]  
> >
> > 64-bit Arm is Aarch64 and not arm64, so this probably won't capture
> > 64-bit Arm if/when we start to support it.  
> 
> We shouldn't assume that the same problem will occur on Aarch64.
> I'd prefer to wait and see.

Agreed. I normally try to look at supported-arch and the like while keeping
in mind that hopefully we'll have more architectures in the future.
 
> > It turns out I didn't search gstreamer.scm well enough, so if this patch
> > looks good then I'll copy the syntax over to the other packages in
> > gstreamer and make patches for them too.
> >
> > address@hidden:~/workspace/guix$ grep \(\"orc\"\ \,orc\)
> > gnu/packages/*scm
> > gnu/packages/gstreamer.scm:           `(("orc" ,orc)))
> > gnu/packages/gstreamer.scm:       ("orc" ,orc)
> > gnu/packages/gstreamer.scm:       ("orc" ,orc)
> > gnu/packages/gstreamer.scm:       ("orc" ,orc)))
> > gnu/packages/gstreamer.scm:       ("orc" ,orc)  
> 
> To be honest, I'm having second thoughts about this entire approach.
> 
> My concern is that without orc, the codec implementations (especially
> the video codecs) might be so slow as to be practically unusable.  I'm
> reluctant to apply this workaround too broadly until we know whether it
> results in anything good.  Really, we should be trying to investigate
> and fix the underlying problems instead of sweeping them under the rug.
> 
>      Thoughts?
>        Mark

If its unusable without orc then definately we shouldn't substitute it out.
On the other hand, `guix refresh -l orc` shows the gnome metapackage,
enlightenment, libreoffice, iceweasel, etc. Other than rhythmbox and guitarix
nothing seems like it should rely too heavily on gstreamer or on its usage of
orc's "just-in-time tools for compiling and executing simple programs that
operate on arrays of data." Giving the other programs a chance to build on
arm and mips I'm sure would be nice for their users. The best would, of
course, be to test it, but otherwise it should be okay to go without it.

-- 
Efraim Flashner   <address@hidden>   אפרים פלשנר
GPG key = A28B F40C 3E55 1372 662D  14F7 41AA E7DC CA3D 8351
Confidentiality cannot be guaranteed on emails sent or received unencrypted

pgpxqY6yboSuF.pgp
Description: OpenPGP digital signature