XBMC4Xbox

Just curious: what's the benefit of this for xbmc4xbox?

allows scripts to work on both platforms without changes. i'll test more, ticket 91 should not happen.

i mean ticket 99

I've reverted this in r30567 since it caused major issues with several scripts. I could reproduce it by running Navi-X as a script which totally crashed my Xbox.

weird that it crashes, not sure where the script is, could you post the call that loads the window.

i think a better solution maybe to have this patch modified to also include the check for skin.xml first.

even better though is to have the scripts updated.

if these are scripts you host post a link and i will update the files.

don't take this as a threat, it's not :), but my scripts updated will not work unless this patch is in. that includes amt and any that use a gui, like amt lites showtimes, home theater experience...

i just don't see the need to keep two versions.

Replying to [comment:7 nuka1195]:

weird that it crashes, not sure where the script is, could you post the call that loads the window.

i think a better solution maybe to have this patch modified to also include the check for skin.xml first.

even better though is to have the scripts updated.

if these are scripts you host post a link and i will update the files.

don't take this as a threat, it's not :), but my scripts updated will not work unless this patch is in. that includes amt and any that use a gui, like amt lites showtimes, home theater experience...

i just don't see the need to keep two versions.

What do you mean by "could you post the call that loads the window." ? Of course the scripts should be updated but on the other hand, it should not crash XBMC and it would be huge win if we could keep backwards compatibility (which I think is possible by just detecting the type of argument used). I surely understand that you don't want to create seperate versions for XBMC4Xbox so I don't take it as a threat, no worries there. Maybe you could change the patch with your (and mine) suggestions? I surely like to see this sorted before we release 3.0

could you point me to the script that is crashing xbmc4xbox, so i caqn test the exact issue.

yes i think a simple check for skin.xml first and make sure the 4th passed variable is a string or a bool should work.

but i agree, crashing xbmc is weird, just wondering if it's this change or not.

Use Navi-X as a script (NOT plugin) from here http://code.google.com/p/navi-x/downloads/list. It crashes right after executing it.

Isn't just "skinInfo.Load(basePath);" missing in the new code?

haha, i just found it and was going to post a patch :p, i will anyways.

Fixed properly in r30570.

@Nuka1195: Please inspect what I did in winxml & winxmldialog. Not entirely sure the legacy stuff is correct.

@arnova, you may want to check my patch, it only affected the one file, it was something i missed from xbmc. of course it's based on your pre revert.

about the legacy stuff, i'll see what i can find.

Replying to [comment:15 nuka1195]:

@arnova, you may want to check my patch, it only affected the one file, it was something i missed from xbmc. of course it's based on your pre revert.

about the legacy stuff, i'll see what i can find.

Maybe I'm missing something but I don't really see why this would help in case *res is not properly initialized? My changes specifically init RESOLUTION variables to INVALID everywhere now, so we can check that in GetSkinPath. Mind elaborating?

And if would use your patch, shouldn't it be:
RESOLUTION tempRes = INVALID;
if (*res != INVALID)
res = &tempRes;

?

no my patch is what i missed directly from xbmc. it must rely on not initializing res, like you did.

Ok, so we might as well leave it the way I've modified it (?)

sure, they are different enough from xbmc, that any changes in xbmc won't directly patch anyways.

but you should mention it to jmarshall or fix it in xbmc.

i mean fix the not initializing res.