She got her data in my pbwiki; help us get it out!

February 21, 2009 4:02 PM

How can I migrate away from pbwiki before they do their "upgrade" in March?

I tried to come up with a cute title. I hope people get it.

A colleague of mine has a very large (hundreds of pages) wiki using pbwiki. It is niche-academic-related and really the only source for this information on the web. It is so comprehensive it is bewildering. She just got so sick of those of us in the field who should have been disseminating this information *not* doing it that she did it herself, and it has become a tremendous resource. Unfortunately I have no familiarity with the platform.

I can't figure out pbwiki is flat file or db driven, but it seems that it has a proprietary db backend that they won't let us access (they do offer an API that would let you pull from your own MySQL db, but I digress).

Ideally, we'd suck down all of her information and dump it into something like MediaWiki or one of the MySQL based GNU/GPL options that we would "own" so that we wouldn't be subject to the whims of a company that decides to "upgrade" or price her out of a very expansive Wiki.

Anyone have any ideas? I know a bit about this stuff, but I'm in way over my head. And I've Googled the heck out of pbwiki, and I can't seem to figure out much of anything, other than the fact that there are a lot of questions asking how to migrate away that have been left open on forums (no answers).

MediaWiki would be my choice because I could do a scripted install that would take me about five minutes. I'd be willing to entertain other Apache / MySQL GNU/GPL options.

Thanks!

posted by dirtypants to computers & internet (7 answers total)

Their features list says you can download a .zip file of all your wiki's pages and revisions, for backups. (Even if you decided to stay with pbwiki, it would probably be a good idea to snapshot your data onto your own machines periodically. You never know when a company might suddenly vanish.)
posted by hattifattener at 4:34 PM on February 21, 2009

Seconded. They offer an archive feature. Use it.
posted by yclipse at 7:06 PM on February 21, 2009

Just to provide some background for other folks, PBwiki has been developing a "2.0" version of their service, and among other things it includes a WYSIWYG editor instead of a plaintext editor. Many are bristling at that, especially power users. When I heard about this last week (I have a wiki there set up for internal use by a small non-profit) I looked into it and stumbled across a massive shitstorm of people screaming about the impending change. Now that seems to be gone, hidden away. Hmmm.

Few things suck as much as having to migrate wikis. This is why I insist on "works with regular HTML" when evaluating wikis. One tool that I've found invaluable is WikiMatrix.
posted by intermod at 7:54 PM on February 21, 2009

OP here. Maybe I wasn't clear.

I didn't choose the software, nor do I maintain the site. I would like to help her get the info out of it, parse it, and get it into another wiki. Preferably one that is MySQL based and open source. I'm asking for ideas about how.

Yes, she can archive it. It archives with out the applets and without other proprietary code. So it doesn't render the way that it does now, but at least we'd have the non CSS text.

Sigh.

If I posted a question saying, my friend is upside-down in her Chevy, and it is due for a new transmission, and I'd like to get her into another car that is similar to hers, I would hope your advice wouldn't be "She has a transmission; fix it."

Please re-read my post. We'd appreciate some ideas.

Intermod, thanks for having our back. And that is a great link.

dp
posted by dirtypants at 8:49 PM on February 21, 2009

They did help you. An archive is a pull down of all that data which is exactly what part of your question was. Converting that data into another wiki format is hard work as most wikis use unique syntax. I did it once using a custom parser in Python, though my parser was not for PBWiki. You are unlikely to find a conversion program to do exactly what you want.

Whining about bad answers won't encourage anyone to help you.
posted by chairface at 8:22 AM on February 22, 2009

There's a perl script here that can mangle a pbwiki archive zip into a MediaWiki xml import file. It may get you started, and at least it'll get it into MediaWiki import files. (Although probably with a lot of borked up markup.)
posted by PantsOfSCIENCE at 9:27 AM on February 22, 2009 [2 favorites]

Wow, PantsOfSCIENCE, that is fantastic. Thanks!
posted by dirtypants at 10:34 AM on February 22, 2009