[mythtv] [PATCH] Transcoder w/cutlist updates

[Geoffrey Hausheer]

On Tue, 17 Jun 2003 06:30:37 +0100, "Mark Levitt
mark.levitt-at-ntlworld.com |mythtv/1.0-Allow|"
<kpwrhgm2ab0t at sneakemail.com> said:
> Actually, I think it would be usefull to have a way to configure another 
> machine for transcoding.
As I said, I'm open to other's opinions

> 
> I've got an older PIII-800 that's small and quiet. I'm using this as the 
> master backend. It sits in the corner with no mouse, keyboard or 
> monitor. I've also got a new P4 2.4Ghz machine that I use for games, 
> day-to-day work, and transcoding DVDs.

Dealing with a 'sometimes availiable' system will add quite a bit of
complexity to the transcoder.

> Here's an idea for implimenting this as well:
> 1) Have a "transcoder priority" setting for each backend.
> 2) By default, the master backend has higher priority than other 
> backends, but the user can change this
> 3) When transcoding starts, check the priority for each backend. 
> Whichever has the highest priority does the actual transcoding.

I think this is way overkill for whatthe transcoder does.  It's only real
purpose is to save disk-space.  It does this by converting from one
format to a more compressed format (because the backend wasn't capable of
encoding to that format real-time), or by removing commercial sections. 
It should not affect the user at all (the user shouldn't notice whether
he is using a transcoded file or not).  As such, transcoding has a very
low priority.  It doesn't matter when it happens, and running it on a
slow machine is not going to make much difference (unless you are
recording 24/7 and don't have a big enough hard-drive)

Adding the ability to assign all transcodings to a given machine may make
sense.  That machine could be a backend with no recording capability for
instance.  I'm not sure what we gain by load-balancing the transcodings
though.  Most people will use a single backend, and will transcode rarely
(they will use manual mode to transcode the files they want to save). 
Everything you said could be done, and if there is significant interest,
I'll be happy to implement it (or someone else can do so), but seeing as
how noone is actually using the transcoder yet (at least not in a manner
which makes any of this relevant), I think I'll hold out until (a) it
works well enough to have it enabled, and (b) people start using it.

.Geoff
-- 
  Geoffrey Hausheer
  XXXXXXXXXXXXXXXXXXXXX