[dmdsoftware]

- override player playback (no longer displays "opening stream")
* With smaller files, it's ok. But with larger files, there's a long pause and I think it's better to show a dialogue that it's opening the stream (otherwise it feels like XMBC is frozen).

I was thinking maybe of some indicator. Will at least allow the user an option to toggle this on/off.

- fix playback default for never transcode + always cache
* This is working fine now. Defaults when just selecting a video. The caching pop up shows fine as well when its caching (although it shows the filename.original).
* I tried the other settings like offline or download, but couldn't figure out where the file gets saved to. Maybe it's in the temporary directory where the cache is.
Perhaps a file download location setting is needed.

It gets saved to a subdirectory under the cache folder selected in settings. It should be directly playable from selecting play offline.

-- Google Drive API 2 module - add force cache to single file option
* I like this. It should actually be the default setting in my opinion. I noticed when you pull up the menu when it's playing though, it shows the filename as cache.mp4 (maybe you could show the original selected filename instead)
* Is there anyway to automatically delete the cache file after playing? Or give an option to delete it when exiting or option in the settings to auto delete it?

I try to avoid deleting things. Also, an action to delete can only be tied to either the player action STOPPED or ENDED. Just because a user clicked stopped doesn't mean they won't click play immediately afterwards, trying to playback the content.

* Also, I realized the technique used for caching is really downloading the entire file. It's fine with smaller files, but I accidentally had the cache setting on when trying to play a large 8gb HD movie and I completely filled up my Fire TV stick and crashed Kodi (I tried to cancel the caching/downloading, but even though I hit cancel, it still downloaded until it crashed). Maybe put in a check routine of the amount of space available in the temp directory and if its too big, default to streaming with no cache & tell the user this is the case (or just say you can't cache this file b/c it's too big).
* Is there a way to make this a true fixed size buffer cache (i.e. 500mb or 15 seconds ahead) vs. downloading the whole file in the background and just playing it based on the % cache in the settings?

XBMC doesn't allow you to directly manipulate its buffering, so that is why I use this workaround. ffmpeg probably won't recognize a media file if it is split across multiple, as only the header of the file will have key details needed for playback. If it can play back segments, then it might possible to playback segments in a playlist, but of course, there would be pauses in playback as XBMC transitions across segments. A lot of online streaming that use Adobe HDS use segments of a few seconds log, that the client weaves together seamlessly during playback. This is still lacking in XBMC (which uses ffmpeg behind the scenes). I saw plugins that need to deal with stream segments actually download the segments and stitch them together before playing back.

This is important for other cloud storage plugins as GDrive is the fastest, but you really need some buffering for other ones like Onedrive. Is Amazon Cloud support with unlimited on the roadmap?

With 0.5.x, I moved to sync gdrive with the cloud services framework that I setup that the other plugins are using. The idea there is a feature that is implemented for one can easily be moved to the other. I have plans on bringing the OneDrive plugin up-to-speed with the gdrive plugin. The only feature that has to be setup manually is the OAUTH2.

Amazon Cloud Drive has some design issues. Firstly, there is no web interface, user needs to install a Windows or MAC client to access their files. There is a developer's API, but it is closed access is highly controlled by Amazon. For some odd reason, they treat it like an app store, and applications that use any of their API need to be reviewed by them, etc. This really didn't deter the idea of making an app through the official channels, but I also noticed that they limit you to 8 devices, and there is no way to unregister a device. Apple does the same with itunes, but they at least let you remove a device. From what I read in the forums, once you have hit your 8 device limit, your account becomes permanently locked and only accessible by one of those registered 8 devices. The guy who developed the Amazon Cloud Drive fuse program mentioned he lost accounts due to this. Seems like a major oversight. Finally, uploading content seems to be a major hurdle with Amazon Cloud. Their API shows that you simply POST the entire file in a single request which is going to be prone to fail for large files. All the other provides provide ability to upload files in chunks, which make it easy to move files between cloud services. I have a custom-made script that backups up my files in a reliable and timely manner between both OneDrive and Google Drive. Never a good idea to put all your eggs in one Cloud-basket. I really wanted to add Amazon Cloud Drive as a third basket, but I'm not about to go through the task of trying to get content synced to it with such limiting options. BTW, their file size limit is 2GB.