[BlueCop]

mr.b: I am generally against shifting any load from a client to an intermediate server. I don't like the server dependance. I would just rather do it all in the python is possible.

I am working on a module that will cache plug-in queries itself rather then the web request. I am still coding it right now. I am doing an md5 of the parameters that the plug-in is started with. It uses this to look up to see if the request has been made before and if the cache for it has expired.

If the cache is expired then I want load the expired items and then update the cache in the background. I am trying to handle changes to the sites by executing a directory refresh when the new cache is saved so then the new information will be refreshed once processed.

I might fall back to just requesting and store the data before loading the directory but the user will have a bit of a wait while the directory loads.

That is what I am trying to work out right now. It will hopefully speed up things quite a bit for some things.

[jmarshall]

@BlueCop: It may be useful, once you come up with a good scheme, that XBMC does that for addons on your behalf? That way each addon author doesn't have to reinvent the wheel and gets it for free. Trick would be determining when the cache expires...

[BlueCop]

hulu has the advantage of alot of information. we have expiration dates on videos so I wanted to use those for some video list expirations. Show and networks listings could be cached for a longer period because they rarely change. I want a relative short expiration on popular,recent, and featured lists because they are updated daily.

I think it would be useful for plugin developers to specify the expiration..

Some plugins it might be useful to have cobtrol over the cache process in the plugin. I have been looking for a python web cache module that properly uses cache headers. one that is able to check the site for change quickly for web scraping plugins. You could then tell xbmc to return the cache from the previous run is you get a 304 not modified.

if it were to integrate it into xbmc I would imagine it as part of endOfDirectory(). It could also include an option to return cache for the not modified situation.

[Googler]

Hi BlueCop
Thanks 4 ur hard work.

I am trying to import your Hulu plugin to other programs. I have read your source code here:

Code:

http://code.google.com/p/bluecop-xbmc-repo/source/browse/trunk/plugin.video.hulu/resources/lib/stream_hulu.py

But I have trouble finding the 'video_id' on line 313:

Code:

video_id=common.args.url

I have read the source code of common.py here:

Code:

http://code.google.com/p/bluecop-xbmc-repo/source/browse/trunk/plugin.video.hulu/resources/lib/common.py

But I still dont know how to get the very important 'video_id'

Could you please outline the process for me? Which functions are involved?

Thanks a lot!

[BlueCop]

visame: what other programs are you trying to integrate with?

It is the video id. it is clearly labled in the data sources. I am not sure what you're attempting so I can't really help much.

[kmvguy83]

im using this plug-in via xbmc installed on my apple tv, and i was wondering if there are ways to activate closed captioning on the videos

thanks in advance

[lordofportals]

I just used your Hulu addon with the latest version of XBMC on a Sony Viao laptop running the Ubuntu 11.04. The plugin loaded without errors. An index of TV episodes loaded. When an episode is selected the advertisement plays. However, the episode file cannot be found.

[BlueCop]

kmvguy83: yes in the settings

lordofportals: what do you mean by latest? what librtmp version do you have? It will not work without a newer git version.

Please post to the support thread below and not use the Developer thread for support.

http://forum.xbmc.org/showthread.php?tid=45889

[toejam]

Update: just did some testing and the smil encryption is still the same.

 hulu's xml encryption of the playlist seems to have remained constant over the years.

The playlist xml contains the paths to the video urls.  Back at the time that the above post was written, the playlist xml was retrieved via a request to s.hulu.com/select, something like shown here: https://web.archive.org/web/201503240102...0/b5HfkMc0

Now, the playlist xml for the modern dash playlist is retrieved via a request to s.hulu.com/playlist.

But the data returned in both cases (the old 'select' url and the new 'playlist' url) is not in plain text.

1. Is there a standalone python script that can take the encoded data returned from a playlist xml and convert it to text?  Probably anything that existed to decode the old select xml playlists years ago would still work on today's xml playlists.  I just can't even find any code for the old system.

2. Is the xml encryption of the playlist the same or different than the encryption of the smi format subtitles?  I have found python decryption code for the smi format subtitiles which still works today, but it appears to be different than the encryption used for the playlist xml.

Note that none of the above references to encryption have anything to do with the actual content.  It is only about the xmls that an add-on would use to construct usable data for the site.  This is long before any content would be played and has nothing to do with the drm.

So, if anybody has any references to code that was used to decode the "smil" back in the day, please point me to it.  The old add-ons probably had it, but searching today has turned up nothing.

[oOflyeyesOo]

Not sure if this can assist in anyway, just posting in case.

https://github.com/michaelhelmick/hulu