(2014-03-07, 15:18)nick3young Wrote: Thanks for the replies everyone.
My Windows 8 box and Raspberry Pi are now both displaying my media correctly. However, my Pi isn't showing the ticks for what I have seen or continuing from where I left off on the Windows computer. I have added the advancedsettings.xml into this location on the Pi: /home/pi/.xbmc/userdata. The xml file contains this:
<advancedsettings>
<videodatabase>
<type>mysql</type>
<host>***.***.***.***</host>
<port>3306</port>
<user>xbmc</user>
<pass>xbmc</pass>
</videodatabase>
Maybe just a typo in your post, but you are missing the closing </advancedsettings> tag. Check xbmc.log on the Pi to make sure the advancesettings.xml file is being read correctly - it sounds like it is though.
One other thing, is your Windows PC and Raspberry Pi running the same version of XBMC? Different versions of XBMC use different database schemas and you might be using a more recent version the Pi which has created a new database schema (from the version created the Windows PC) so now your PC and Pi are no longer synced, they're using two different databases again. Solution would be to upgrade Windows to the same version of XBMC used by the Pi, or downgrade the Pi to the version used by Windows (and drop the databases created by the Pi).
(2014-03-07, 15:18)nick3young Wrote: This is working fine on my Windows computer. On the XBMC 'how to wiki', it mentions having to add this to the xml also:
<videolibrary>
<importwatchedstate>true</importwatchedstate>
</videolibrary>
Is this necessary?
No, that only applies when importing the watched status from NFO files and has nothing to do with MySQL.
(2014-03-07, 15:18)nick3young Wrote: Also, I'm a little unsure as to how all this works! From what I have read, since v12, XBMC doesn't keep the media in the mySQL database... it instead caches it locally to improve performance. Does this mean that when both are working, the Pi and my Windows computer will both have their own local copies of fan-art etc?
Yes.
The script in my sig can help optimise performance of the cache on each client. The cache on the client(s) used to scrape the library will be populated as part of the scraping process, but the caches on your other clients will not contain any of the artwork for newly added media until you display each item in the GUI - this can result in noticeable delays as the missing artwork is fetched, resized/resampled and then cached before finally being displayed (hence the script).
(2014-03-07, 15:18)nick3young Wrote: And MilhouseVH... how do you designate which is the scraper? When you add a source, you have to enter the media type (e.g. movies)... should I only be doing this on one of the devices?
I would, but you don't have to. I just think it makes more sense to have a single "scraper" client, typically your most powerful client but it could also be a Pi particularly if your scraping is fully automated in which case an always-on client is more useful.
So how you choose which client is your scraper is up to you, I was simply pointing out that you only need the sources.xml defined on the client(s) that will be doing the actual scraping. If that is all of them, then make sure they all have the same sources.xml.
My approach is to scrape new media using a Pi which adds the new movie/episode to the library, then run any addons on the Pi to grab extra artwork or metadata (eg. Artwork Downloader for logos, clearart, etc.) and then finally update the texture cache of all my other clients so they have all of the new artwork in their cache. This is all fully automated from a job running on my NAS and the script in my sig.