[mythtv-users] MythTV vs XBMC

Gary Buhrmaster gary.buhrmaster at gmail.com
Thu May 8 18:41:44 UTC 2014


On Thu, May 8, 2014 at 5:03 PM, Thomas Mashos <thomas at mashos.com> wrote:
> On Thu, May 8, 2014 at 10:01 AM, Raymond Wagner <raymond at wagnerrp.com> wrote:
>> On 5/8/2014 8:14 AM, Brian J. Murrell wrote:
>>>
>>> On Thu, 2014-05-08 at 08:10 -0400, Greg Thompson wrote:
>>>>
>>>>
>>>> I believe XBMC looks the same everywhere since they decided to use All
>>>> OpenGL for their interface, rather than QT and a window manager...
>>>
>>> And so yes, MythFE could copy that and shed baggage, etc., but again
>>>
>>> with the shoulders of giants.
>>
>>
>> Every time that is suggested in earnest, it is met with all kinds of push
>> back from users wanting to run on low end machines that don't have the GPU
>> necessary to run OpenGL only.
>>
>> _______________________________________________
>
> It's 2014. At some point these "low end machines" should be able to
> run OpenGL right? I mean, are there lots of users that run MythTV on a
> PIII?

(question targeted mostly for Raymond)

A far more relevant question is how many users are running
MythTV today that would not upgrade to (say) 0.28 if only
OpenGL was supported?  It is not relevant as to how many
run MythTV without adequate GPUs who plan to stay on
0.24 for as long as they can do so.

The MythTV smolt data is known to be limited (i.e. broken),
but suggests that there are not a lot of users on especially
older systems(*).  I accept as unverifiable from the data
that there are also users on newer systems which cannot
be upgraded to a decent GPU.  But only those who also
wanted to upgrade to (say) 0.28 would be forced to upgrade
or replace those older systems, or purchase a $20 video
card from the dust bin at the local swap meet (which is
about the only place you will find agp cards these days).

There is precedent, even by Microsoft, when silverlight
required certain mmx/sse instructions and failed to
install on older systems that otherwise ran Windows
just fine (well, as fine as Windows ran on those systems).
Sometimes, you just have to break with the past, and
users on (very) old systems are now "functionally stabilized".

And, the followup questions would be what improvements
might we see if the project required OpenGL only?  Would
it make it easier to move forward (whatever that means)?
Would there be resources to move forward regardless of
the ease of movement?  If not, does any of this matter?

Gary


(*) That is a assertion based on the WAG based on
the memory limits of most PIII systems of 1GB, and
the fastest PIII system bogo mips rating was around
1600.  There is not enough detail in the smolt data to
more than infer (and, did I say it is known to be
broken?), especially in regards to the GPU capabilities
on their frontend systems.


More information about the mythtv-users mailing list