stuart wrote:
> This raises the age old question of how to develop code for Mythtv.  Are 
> you willing to run the SVN version of mythtv so as to be able to 
> contribute?

True, but if we're talking about Mythweb specifically, and you're not 
too concerned with contributing code upstream (which ideally you should 
be), then you've already got the PHP source and can hack away on the 
installed version.

I suppose the same could be said for the compiled MythTV code, where you 
can always check out the source for whatever release you happen to be 
running. So I guess it comes down to how much effort you're willing to 
go through to be able to contribute code upstream.


> My guess, people who develop mythtv code maintain two boxes.

Yes, that seems like the most practical approach. Even if you wrote some 
scripts to automate switching between production and development 
versions, you'd constantly be risking interrupting scheduled recordings. 
Unless you can figure out how to make two MythTV installations run in 
parallel on one box. Perhaps a dev mode where it binds to an alternate 
IP address, uses an alternate database name, and treats the recording 
directory as read-only. Still, given the way MythTV clients have 
tight-coupling to the database, that'll break some clients, and at 
minimum require inconveniently reconfiguring clients (as someone asked 
about on this list recently).


> The web route would be faster, but it will always leave you wanting
> or adapting to different clients.  (Example, look at all the web page
> styles SqueezeCenter can support. I know, it looks great, but we
> don't have the room.)

An mvpmc command line option specifying the location of the web page 
that should be served up for the UI would address that. (Or more simply, 
the "doc root" of the mvpmc web server.) That way the end-user could 
have it point to a network mounted file system, which could provide a 
rich UI with graphics, CSS, etc. without concerns about storage space 
limits. It would also facilitate easier end-user hacking without 
necessitating rebuilding a dongle. Perhaps even lowering the effort 
barrier enough that the web admin UI will get that long awaited face lift.

That's not to say that a CLI or other interface shouldn't be implemented 
as well. Just that a web UI is likely to be more immediately useful. 
Another option to look at would be hacking the IR receiver driver so 
that it can accept keystrokes from an alternate source, like a named 
pipe, or TCP socket. Still not a deterministic interface, but should be 
far easier to implement that a proper CLI.

  -Tom


------------------------------------------------------------------------------
_______________________________________________
Mvpmc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mvpmc-users
mvpmc wiki: http://mvpmc.wikispaces.com/

Reply via email to