Serviio

I installed the newer version of Serviio and noticed that it doesn't recognize series and put them in series folder. All the other folders (sorting movies by producers, actors and whatnot) are ok. I uninstalled it and installed the older version and everything was ok again.

there should not be any difference between the two versions in this regard. You could check detailed log from the 1.0.1 version to check if it found the online metadata - might have been a problem with the online service when you ran the scan.

this happens because the owner of thetvdb.com has implemented a browser check feature on the site. When serviio tries to fetch online metadata it sends (in my case) "User-Agent: Java/1.6.0_31" header and site replies with 403 Forbidden. Dear Petr, could you please change the header to some of well-known browser identificators or let us to do this ourselfs in java properties. Thanks in advance.

Do you have any source for this claim? It works for me and I doubt they would do a browser check for public API,which is supposed to be used mainly outside a browser.

Yes, I've tried to catch traffic to thetvdb.com via tcpdump. Result is attached.

Cut of the logfile:
2012-09-23 18:03:58,851 WARN [LibraryManager] Cannot read metadata of file /mnt/storage/video/Торренты/Активные/Alphas.S02E08.rus.LostFilm.TV.avi via extractor ONLINE_VIDEO_SOURCES. Message: Cannot retrieve episode details (series = 210841, season = 2, episode = 8): Server returned HTTP response code: 403 for URL: http://thetvdb.com/api/235C8CA4529142E9 ... 2/8/ru.xml

Also I've tried to replay this issue manually. I've used lynx to introduce a request with the default User-Agent and with Java/1.6.0_31. Very interesting things. The default requests in series return correct response. But requests with custom User-Agent return 403 Forbidden on second and following iterations.

Sample requests where:
default: lynx http://thetvdb.com/api/235C8CA4529142E9 ... 2/8/ru.xml
custom: lynx -useragent="Java/1.6.0_31" http://thetvdb.com/api/235C8CA4529142E9 ... 2/8/ru.xml

So for me it means some kind of restriction technique, but probably works just for my IP address (Russian IP block). I'm using Serviio v.1.0.1 and didn't ever use the older versions.

I knew I am not crazy. The same thing happens to me. I've reported a bug on bitbucket and it is being resolved. :thumbsup:
It is really strange that they are putting a request filter by user agent on an API

FWIW I've noticed an inconsistency in the API. On the one hand it always works when I open one of the metadata links in browser or with curl using one of the standard user agent strings. On the other hand when I try to curl with Java/1.6.0_23 user agent string it sometimes fails and sometimes it works.

Just executed some tests. Requests with "Java" user-agent to php script need approx. 2 seconds to be unbaned, but requests to .xml file need more (didn't measure exact time). So, when file is adding to library there are 2 requests (1st - get xml path, 2nd - retrieve xml info). In this case while executed in series 2nd request fails 'cause 2 sec interval didn't meet. And subsequent requests also fail, 'cause there more time to be unbaned after failed xml request. IMHO.

Do you see the same behaviour with a 'Mozilla' user agent?

No. The use of -useragent="Mozilla" doesn't reproduce this error. Tried 8 subsequent requests to different URLs in 5 secs - all was successful. But, for example, -useragent="MSIE" fails.

Well, for now the problem has gone and I can see all series again! It seems that the content owner fixed his filter on browser ID.
I hope that such a problem will not repeat and annoy again )) And thanks for fix to all of you.

thanks for letting us know, I'll close the ticket