Legion Wiki talk:Community Portal

The status quo

Okay, probably best to start with "where things stand right now":

1. All pages/revisions from 2008-2010 that anyone reading this can't restore should be considered "Lost for good"
2. The images uploaded in that time exist on the server, but the out-of-date database doesn't know they're there.
3. The temporary database files are incompatible with the old database files, for the cherry on top. NC's trying to make them fit, but I don't think there was much there anyway.

Summation: over - Reboot (SoM) talk page 23:38, 6 June 2010 (UTC)

Is it possible to copy over pages from the old database to a newer version of MediaWiki, basically create everything again in a new database? If possible, would images be recognized by the new arrangement, and is it worth the effort? I'm resigned to the fact that we will have to recreate everything from the last couple of years. I think I have quite a bit of the research, just will take take to remember what was there and then rebuild it. My current questions are:

1. What can we do to insure against anything like this happening again? I know that I will be saving copies of large important pages that I work on, but we can't save backup copies of every page we work on.
2. How much of a pain is it to make backup copies of the database on the regular basis? If this were a company, I would suggest we store backups with Iron Mountain, but I don't think that's a path we will be going down. However, we could burn a copy to DVD and send to a couple of people or upload a copy of the file to a commonly accessible location. I'd be willing to pick up the tab for any incidental costs in doing this, as long as they don't get out of hand.
3. How do we handle the image files that exist but are not recognized? Should we re-upload or is there a way that we can get them recognized again by changing the wiki code or pointing the main settings file to a different folder?

Gopher 01:31, 7 June 2010 (UTC)

1 and 2 are really for NC to answer.

On 3, image uploads aren't enabled yet - but once they are, there's no way to fully automate it since the image descriptions (i.e., the pages associated with the files) are lost and will need to be retyped like any other lost page - the server can't tell that [1] this is a cartoon screencap, nor what episode its' from, nor what characters appear in it. Given the URLs to the images, we could tell the server to download the files from itself, but otherwise it's all manual. - Reboot (SoM) talk page 02:18, 7 June 2010 (UTC)