[AnalogKid]

Guys.... (and gals if there are any),

Is there any hope that we could adopt a consistent approach to 'special' artwork not inherently supported by XBMC?

This is NOT a criticism of work done so far... it's been very innovative indeed, but the same problem has been cracked half a dozen different ways....

So...

If seems to me that the crux of most 'special' art lies in knowing the path of the media file itself (so that the skin can pull a side by side art file). In many circumstances, this can be difficult, or impossible to know e.g.

U2\
U2\Greatest Hits\
U2\Greatest Hits\01 - Some Track.mp3
U2\Greatest Hits\somespecialart.png

SOME skins have opted to replicate a hierarchical structure under the skin\media folder, and some have opted for a flat structure under the skin\media folder...
OTHER skins have asked the user to define the path to the media files (in a skin setting screen), then they've been able to locate the special art side by side with the media.

I would like to propose that the latter technique is the better technique (because it allows users to keep their special art with their media, and it's not skin specific).

If you buy into this argument so far...


I would like to propose the following:

ALL 'special' (non XBMC native supported) art goes side by side with the media file and is prefixed with the string "special-"

e.g.

U2\
U2\Greatest Hits\
U2\Greatest Hits\01 - Some Track.mp3
U2\Greatest Hits\special-cdart.png
U2\Greatest Hits\special-clearart.png

in the case of multiple image collections (backdrops, movie frames etc), then a subfolder prefixed with 'special-' is used... e.g.

U2\
U2\Greatest Hits\
U2\Greatest Hits\01 - Some Track.mp3
U2\Greatest Hits\special-cdart.png
U2\Greatest Hits\special-clearart.png
U2\Greatest Hits\special-fanart\
U2\Greatest Hits\special-bandshots\


I know this isn't a perfect solution, but it would at least bring SOME sanity to what's going on, it's relatively easy to implement and really helps the end user and 3rd party media managers. Plus.... should any of these special art types be considered good enough to be adopted by XBMC natively, it's fairly easy to write a script to remove the 'special-' prefix.

I also realise it requires users to have consistent structure in their media organisation in order to locate the special art... but I still think it would be worthwhile getting the skins to behave consistently.


Is this doable/sane/insane/pita?

[rwparris2]

There isn't really a point to the special- prefix.

[AnalogKid]

There isn't really a point to the special- prefix.

The point of the prefix is to clearly identify that such elements are not part of XBMC per se. That whole artwork situation is already a nightmare with just the officially supported stuff, let alone the extras.... I believe it's useful to clearly identify that this art is 'you're on your own with this solution'. It absolutely serves no benefit other than that, but I believe it's worth noting it.

[jmarshall]

There isn't any point to any of this. No matter what scheme you come up with, trying to get to images that XBMC doesn't know about via path-based hacks WILL NOT WORK.

Just say no.

Cheers,
Jonathan

[fekker]

AnalogKid -

I understand that your looking to make things where they work under multiple skins, a standard so to speak.

Maybe what should be done is a well defined outline (what image is what) should to be written up, a track ticket submitted, with a patch that adds the support (if you can) so xbmc supports the images in a sane method and there's not every skinner making up diff image support in each skin, mod of skin, skin copy, etc.

And when it's something that's in xbmc, the 3rd party tools can add support for it as well, making this cleaner, easier, universal and working with all skins and platforms without hacks.

just my 2 penny's of course
- fekker

[AnalogKid]

Well...I tried. I have had a ticket for a comprehensive overhaul of the art system open for ages (if you're gonna fix it, might as well make it a comprehensive fix), but the motivation for an overhaul is probably low when 90% of users needs are adequately met with fanart and a thumb. It's just those 10% of weird and wonderful artwork hacks I'm trying to address.... the devs saying "don't do it" and the skinners "have already done it"

The rush to get ever more ambitious skins and mods doing more and more graphics has resulted in a mess for the typical end user, and a nightmare for noobs.

My philosophy is.... "if we can't stop you making ticking time bombs, at least do it in a nice well defined play area!"

*Waves white flag and surrenders*

[joebrady]

I kinda have to agree here, these hacks do work (maybe not well, or in the best way possible, but they do WORK), as skinners are already using them and have been for a while now. And not just skin mods, but full skins that have been featured on xbmc.org. (Transparencey!, cdArt and ClearArt, which is already doing it in the way AnaglogKid suggested)

At least this would bring some sanity to the chaos until these special images are either incorporated into XBMC or deemed to never be incorporated.

[AnalogKid]

I kinda have to agree here, these hacks do work (maybe not well, or in the best way possible, but they do WORK), as skinners are already using them and have been for a while now. And not just skin mods, but full skins that have been featured on xbmc.org. (Transparencey!, cdArt and ClearArt, which is already doing it in the way AnaglogKid suggested)

At least this would bring some sanity to the chaos until these special images are either incorporated into XBMC or deemed to never be incorporated.

I admire the way the Core XBMC engine is dealt with (on the whole), and it's right that changes aren't made without careful consideration.
As a rule, the skins tend to be a little less well controlled.... you could view these skin hacks as 'upstream' developments...

If we could just align the skins to use the same techniques consistently, then in my view this would actually strengthen the case for XBMC to incorporate them (or some of them) into the core. It's not a "skin police" effort, it's more a case of "if the rules have to be broken, let's limit the damage, so the end user gets an easier ride.

I still think it's a legitimate development effort from the skinners who just happen to be using controls in a manner not originally envisaged by the core team. Viva la evolution!

[jmarshall]

The hacks don't work. This is the whole problem.

They may work 90% of the time, or even 99% of the time, but they do not (and never will as currently implemented) work 100% of the time.

Given that this concerns primarily how images are handled, and given that the images currently handled by XBMC are a mess anyway, you can see the reason why I constantly say: Don't do it.

I realize the temptation, and am not opposed to skinners getting together and developing something more unified, but until it gets XBMC support (which will not happen until the image caching stuff gets cleaned up so that this sort of thing is possible in a 100% consistent manner) it will remain unsupported, and thus will never be anything more than a hack. Skins that require such hacks will always be the domain of users who put up with such hacks - they will never become mainstream. There's nothing wrong with this ofcourse - they serve a purpose for those users, after all!

In the end, there's a limit to how much useful additional image data is realistically available for a particular movie or tvshow, and a limit to how much you can show without visual overload.

Cheers,
Jonathan