2021-07-17, 03:54
You'll see the skin reloaded messages because I send the command...but I can't do anything about OSMC stopping it, and Kodi doesn't give a confirmation or anything. But it works on all the other platforms, so...
In the end, you might just have to live with that workaround as your approach...not really a big deal to do it what way really, is it? But it's some sort of OSMC weirdness for sure. I could add a note to the wiki for the tiny few folks who still bother with OSMC.
The patcher _is_ working as far as Kodi/Python is concerned - there's no error on the write....so really, what could I possibly do to help here??
The only difference is that one operation (restore) is using a basic xbmcvfs.copy and the other (patching) is using xbmcvfs.write. You could try changing around line 155 to:
I guess that would log the result, at least, which might confirm the write is false, but it wouldn't tell us _why_ still. Again, Python really always throws an exception on failed writes, so it's pretty reasonable to assume the write has succeeded at this point.
The other thing to do might be to manually check the file permissions...or do a search of the entire file system to see if the files are actually being written, but for some reason in the wrong spot....but honestly I just can't see why restoring them first works, that just makes no sense...and I can only care so much about an obscure bug on one particular offshoot platform when it works everywhere else just fine.
(....or just do yourself a favour and get an Odroid N2 with CoreElec - way faster than the Vero, and no weirdo bugs...)
In the end, you might just have to live with that workaround as your approach...not really a big deal to do it what way really, is it? But it's some sort of OSMC weirdness for sure. I could add a note to the wiki for the tiny few folks who still bother with OSMC.
The patcher _is_ working as far as Kodi/Python is concerned - there's no error on the write....so really, what could I possibly do to help here??
The only difference is that one operation (restore) is using a basic xbmcvfs.copy and the other (patching) is using xbmcvfs.write. You could try changing around line 155 to:
Code:
file_to_write = config.xml_destination_folder + "/" + ntpath.basename(file)
log(f"Writing patched file to: {file_to_write}")
with xbmcvfs.File(file_to_write, 'w') as destination_file:
result = destination_file.write(new_data)
log(f"...done...result: {result}")
I guess that would log the result, at least, which might confirm the write is false, but it wouldn't tell us _why_ still. Again, Python really always throws an exception on failed writes, so it's pretty reasonable to assume the write has succeeded at this point.
The other thing to do might be to manually check the file permissions...or do a search of the entire file system to see if the files are actually being written, but for some reason in the wrong spot....but honestly I just can't see why restoring them first works, that just makes no sense...and I can only care so much about an obscure bug on one particular offshoot platform when it works everywhere else just fine.
(....or just do yourself a favour and get an Odroid N2 with CoreElec - way faster than the Vero, and no weirdo bugs...)