Comments inline...
On 2 Jul 2009, at 22:02, Kevin Fox wrote:
On Thu, 2009-07-02 at 10:41 -0700, Michał Sawicz wrote:
Dnia 2009-07-02, czw o godzinie 09:23 -0700, Kevin Fox pisze:
This whole thread is one of the reasons I've been arguing that
Moovida
should not have "a" "PVR" plugin, but be a client of PVR's. Why?
Has anyone argued that there should be one?
There has been a lot of discussion on how a PVR (backend) should work.
If we're just doing PVR plugins that interface to PVR Backends, then
we
don't need to discuss that part at all. Its out of scope of Moovida.
All
we need is coming up with the generic UI components needed by the
various plugins, and let the plugins or the backends specify
recording/user interaction behavior. For example, the EPG, Dialogs
that
the PVR might want to pop up asking questions, etc. The plugin decides
when and if to pop up a dialog related to scheduling questions.
Didn't we agree that we
should have it as flexible as possible to support multiple -
different
-
backends?
The two of us have. I'm wanting to make sure Fluendo is on board with
the ideas we've had. In particular, the comment in the other thread
about having Movies/TV Shows/Pictures/Music not be local but a meta
location.
Everyone here is reading it, myself and other 'higher ups' as well. :)
As I am getting involved in it and hopefully you want me to help you
figure out the UI design and interaction the work, and provide the
mockups...that should assure you it will be reusable. (which is why I
need you guys to do a lot of leg work in terms of writing a detailed
yet concise spec...first version very simple, basics even and plan it,
feature by feature so that you know what will come with each update/
version...again, make sure version 1 is the core basic functionality
and not too ambitious)
The only thing I wanted when arguing pros and cons of mythtv
and dvbd is with which one should we start.
Ah. ok.
Since dvbd is not usable right now for me, and my time is somewhat
scarce, I'd probably only be able to help with the myth one. That
doesn't mean people can't be working on both (or a vdr one for that
matter).
The only question is that: should we build a general, abstract 'TV'
plugin that would then be subclassed to support different scenarios?
That would surely require more work to design it with enough
flexibility.
Thats a good question. It could be also, do we create a superclass
that
all pvr plugins subclass, or do does Moovida provide a set of widgets
used to implement a TV plugin. Slightly different, but subclassing the
TV plugin kind of forces a UI and behavior a bit on the plugin, while
allowing the plugin to use UI elements or not based on its choice is
different.
Very very tricky to pull off, I am confident in saying that most PVR
projects, from a UI point of view have more or less a basic formula
and even a fairly standard feature set.
From that point of view, you can at least have something workable for
any type of backend.
The difference comes if, there are specific unique features that one
project has that another doesn't that require specific UI things.
I would suggest that, you dont go insane thinking too much about it at
this juncture and consider instead the idea that, there is a core
foundation of features and techniques UI wise that they all share, so,
we get those solid.
I'd vote for the latter. We provide a set of widgets, EPG, Player, etc
which the plugin can subclass to receive the hooks it needs to
function
properly. Or it can extend it to provide additional information. Say
the
enhanced EPG that mythweb uses to show recordings that are happening
by
changing the color/border of a cell in the EPG. Some PVRs might not be
able to access that info easily.
Yeah, honestly...you need to trust that the UI I put together for
Moovida can display all these kind of features. BUT, it might not do
it exactly 100%. or copy the UI techniques of other projects... it
will however be able to use those features it has... in the Moovida
style.
Does this sound reasonable?
If we can get a response on how the TV Shows/Movies/Music/Pictures
should work, maybe I can then come up with a writeup of:
--- Generic ---
1) What Widgets from Moovida we'd need for any PVR plugin.
2) What hooks/behaviors we would need from the Widgets (Example,
Player
allowing a dialog to come up in front)
3) For Widgets not currently in Moovida, UI we'd like to see (EPG,
full/now+later)
4) How and where generic PVR menus should show up in Moovida
These ones above, you might not be able to get straight away... after
some mockups, and the UX interaction are agreed upon.
This will be clear what widgets are missing. As I've said in previous
emails... I think everything is there frontend wise, except widgets
for doing a full EPG.
--- MythClient ---
5) What the UI specifically for a Myth Client plugin should look like
Again, the UI is designed to be flexible enough to cater for different
things and I will set this whole thing up in such a way that we have a
generic UI that can do the core stuff that a PVR needs.
So, any backend can tie into what we are doing... but it ties in and
uses things we do in the Moovida style.
Some might not like me saying that, (As you might argue: I like Mtyths
EPG... He might argue: I like this one... She might argue: I want the
flashing big red and green button in that PVR software...etc.) but,
you need to understand that what is making Moovida successful and the
comments we are getting back when compared with other HTPC projects is
the fact we have something so so simple and yet powerful as it is
consistent. (this is from both the public & big companies too :))
Sound ok?
Kevin
--
Michał Sawicz <[email protected]>
David McLeod
Senior Designer & UI Lead
Fluendo S.A.
World Trade Center Edificio Norte Pl.2
08039 BARCELONA SPAIN
Tel: +34 936 002 323
Skype: Daiode
Jabber/GMail: [email protected]
Email: [email protected]
