ubuntu-tv team mailing list archive

[Bruno Girin <brunogirin@xxxxxxxxx>]

On 02/01/12 00:00, Thomas Mashos wrote:

<snip>
> I try not to weigh in on the UbuntuTV discussion until we get at least 
> a little bit of direction from Canonical, or at least an idea of what 
> they are thinking in terms of UbuntuTV. However, I disagree with the 
> idea of Live TV in general, and I disagree with the idea that UbuntuTV 
> (or MythTV) should have any direct involvement in Live TV. (I do agree 
> we should use MythTV for DVR purposes though)

Live TV is still alive and well, at least outside of the US. But if as 
you say below, it's easier to just let the TV itself handle that, I 
agree with you that we should not go there.

> 1) You don't want to run the MythTV backend on low powered systems 
> (ARM, ATOM, etc). Things like the scheduler and the commercial flagger 
> (and even the mysql server) require more CPU than those can 
> comfortably handle. Don't forget you also need decent hard disk space 
> for recordings. A faster second box is better for this. Some people 
> aren't going to want Live TV/recording functionality at all and will 
> use this to watch internet content exclusively. This is why the second 
> box is a better idea.

Absolutely agree, which is why I asked the question. If UbuntuTV is 
built as a MythTV front-end, then you inherit the MythTV design where 
you can have some very lightweight front-end boxes and a single back-end 
box. This provides maximum flexibility and can also make it an 
interesting proposition for hotel entertainment systems.

> 2) While wanting to have the backend be quicker at changing channels 
> is a noble idea, it requires an upstream code change that the MythTV 
> developers are unwilling to write. Partially because there are not 
> that many MythTV developers, but also because Live TV is an antiquated 
> way to watch TV. IMO a better solution would be to completely forget 
> the concept of Live TV and force 'Live TV like' functionality though 
> using the TV guide first and selecting a channel/show to watch from 
> that. In my testing, that worked great and allowed my wife to 
> transition from Live TV to only recorded rather easily. I can't 
> remember the last time we watched something at home that was truly 
> live. As an added benefit, she now laughs at all the people that are 
> forced to watch commercials and can't fast forward/rewind. The faster 
> we get rid of the concept of Live TV, the faster the consumer wins 
> against the cable companies/networks. If true Live TV is a deal 
> breaker, then have the TV do the regular channel tuning (Having 
> UbuntuTV as an overlay) so that it could be fast. Then hitting the 
> record button at that point would cause MythTV backend to start 
> recording that channel and having UbuntuTV switch to the recording 
> from LiveTV. 

I agree that upstream code changes are not a good idea. So as you say, 
if you can have the TV handle Live TV with UbuntuTV as an overlay, that 
would be the best solution.

Bear in mind that live TV is still the main way a lot of people watch TV 
outside of the US (well, certainly in France and the UK at least). A 
typical usage scenario here: the whole nation sits in front of the TV at 
the same time for shows like X-Factor, Strictly Come Dancing, East 
Enders, etc and will make live comments on Facebook and debate Jedward's 
hairstyle or Ann Widecombe's dancefloor moves while the show is going 
on. Then everybody puts the kettle on to brew tea during the commercials 
(seriously, it actually creates surges on the power grid). With the most 
popular shows, live TV becomes a nation-wide social event, hence why I 
think it's important to at least consider it.

My £0.02

Bruno