[mythtv-users] MythTV qtwebkit dependency

[Mark Wedel]

On 10/29/16 09:49 AM, Paul Harrison wrote:
> On 29/10/16 11:01, Tom Dexter wrote:
>> On Wed, Oct 26, 2016 at 5:24 AM, Paul Harrison <mythtv at sky.com> wrote:
>>> On 24/10/16 14:58, Tom Dexter wrote:
>>>
>>>
>>> I think John forgot we still use Qt Webkit for the browser widget that is
>>> part of MythUI and is still being used by the plugins.
>>>
>>>
>>> Paul H.
>>
>> Is that mythbrowser you're referring to? It's too bad this is
>> required, given that most users will never technically need it. Qt
>> Webkit is a big requirement the way it is, but the odd compile time
>> requirement for ruby (which I don't really understand) makes things
>> even worse.
>>
>> Thanks for the reply.
>>
>> Tom
>
> Yeah MythBrowser uses it. Also if you removed it MythNews, MythNetVision and
> MythMusic would be affected since they all use MythBrowser.
>
> They've gone a little quite about it but some of my fellow devs want to remove
> the plugins algether so you will probably get your way at some point. I don't
> see the tiny amount of disk space required for the webkit stuff to be a big
> problem when you consider the multi GB space required for recordings.
>
> You could make it optional and only allow those plugins to be configured if
> QtWebkit is available but then you run into problems with future development not
> knowing if you can use the browser widget or not.
>
> Personally I'd be sad to see it. If you remove the one or two remaining
> differences between Kodi and MythFrontend, the web browser and MythZoneMinder
> etc then what's the incentive to continue developing the frontend? Might as well
> just kill it off :(

  As someone who just recently ported mythtv to Solaris (I really need to bundle 
up the changes - fairly trivial, as a patch and submit it to the tracker), the 
hardest part of it all was probably porting/compiling qtwebkit (Qt had some 
minor issues, but not that bad either).

  For QtWebkit, some of that porting was 'this change is wrong, but at least it 
compiles'.  But the other problem is that since qtwebkit itself appears to be 
dead as far as further development goes, there may not be much point even 
bothering to bundle up the changes to make that compile and submitting it back 
to the devs.

  So if QtWebKit requirement disappeared, that would certainly make things 
easier going forward.

  I don't know if people porting to other (non linux) OS experience similar 
headaches on that, but I'm all in favor of a modular approach.