Homestar Runner Wiki Forum

I personally think we should upgrade to MediaWiki (the software Wikipedia uses), if we can. What are the reasons for not doing so?

Mostly that MediaWiki has about a thousand and a half features we don't need, I think. It's a very good bit of software, but it's also a huge package.

Anyway, we've been planning for a long while to upgrade to TWiki, but do you have any idea how much work is involved?

_________________

Heh, I'm a sysop at a wiki that's much smaller than this one and it uses MediaWiki.

Anyway, yes, I know a lot of work would be involved in making a change, but I'd be happy to help out if I could. I have MAD SKILLZ.

- Kef

P.S. I posted this in "General discussion" because I wasn't making a suggestion so much as wondering why we weren't making the move. But I have no problem with relocating the topic.

Yeah.

There's another thing to consider too, kef. I guess we also try to Keep It Simple. (Stupid?) We've got a lot of people that have a hard time with easy stuff like linking and subpages. We don't want to alienate anyone. I know I'm a little intimitaded when I edit Wikipedia articles.

I mean, we don't even use all the features of 'Tavi.


Yeah, I moved it. I took "I personally think we should upgrade to MediaWiki" as a suggestion. Oh well. It'll stay here.

I don't see why Wikipedia articles should be more intimidating to edit. About the only difference commonly encountered is that [[ and ]] are used instead of (( and )) for links, and you don't have things like DaAaAaAaA becoming links if you don't put an exclamation point before them (which does mean the brackets are required, but we almost never use CamelCase anymore except for, like, PomPom, StrongBadEmail, and EasterEggs).

I could probably do this research myself, but what does MediaWiki run on? PHP/MySQL, or does it use Perl and proprietary storage?

Our biggest concerns right now surround trolls and how to better deal with them. Twiki solves the problem by authenticating users. I originally liked that idea, but now that I think about it, authenticating users flies in the face of the Wiki Way.

I personally like the no nonsense setup we have with 'Tavi, but I don't like the code it generates. Can MediaWiki produce valid (and properly marked up) XHTML and CSS? If so, I would switch on that reason alone, if there are people willing to help. What other options would it give us?

Bottom line: What would be the direct benefits of moving to MediaWiki?

http://wikipedia.sourceforge.net/ - that should answer any questions I can't answer here.

I've never used the software itself, just wikis that use it, but I'm reasonably certain it does generate valid XHTML and CSS. It does use PHP and MySQL.

To me, the primary benefits would be having dedicated talk pages (while we can have that here, it's more automatic...just click "discuss this page" on the sidebar or whatever, even if no talk page exists), having cleaner editing (again, away with the CamelCase stuff), and having a meaningful login system, as well as automatic notification when somebody posts on your talk page, etc.

The biggest downside of course would be making the conversion in the first place...but we should make up our minds about this soon or that conversion would only get harder as more edits are made by more people. We don't have to rush it, of course, but I wouldn't put it off for months. There may be more benefits and/or possible complications than I know of as far as administration goes, again because I don't use the software itself, but that's what I see for the user side of things.

Of course, we could test it out before committing ourselves to it.

- Kef

I don't have any particular allegiance to Twiki. By the documentation, it looks like MediaWiki does indeed run on PHP/MySQL, and I'm always impressed by the apparently cleanliness of Wikipedia's output (though I've never been sure if that's a software feature or a people feature).

I guess, like Joey, I'd like to see a list of what advantages MediaWiki has over TWiki, and vice versa. We should choose the system that fits the best.

I'm just skimming the feature list now and there are a couple features that appeal to me:

• Table of Contents generation
• File upload, automagic image resizing
• Watchlist
• User contributions (which articles a user has contributed to)
• Extended recent changes (yay!)
• Namespaces
• Section editing
• Minor edits
• Talk pages, messaging
• Caching
• Permissions
• Search/backlinks that work

Plus some nice purely cosmetic stuff (e.g. word extension linking, automatic signatures).

The conversion process (which Joey knows I'll be very happy to assist in) is going to suck no matter what. Currently the Wiki has 3,179 pages, so doing each page by hand, in my estimation, is probably not an option. Even if we had 20 people working really hard on it, that would be more than 150 pages per person. If we had a more realistic 6 people working on it, it would be more than 500 pages per person. If we skip WikiFanStuff and userspaces, we might squeak by around 400. Automation is possible, of course, but it won't be a one-button process or anything like it. Since there's no standard Wiki format, whatever Wiki we choose is going to have a different format than ours, and all of those tags are going to have to be converted. It can be done, but yow, doing the scripting is going to hurt. Unless someone around here is WAY better at PCRE than me.

_________________

Does MediaWiki have RSS/Atom feeds?

That's one other thing that would be nice with Twiki.

It looks like RSS is available in the currently-in-beta v1.3. That's the version Wikipedia and its sister sites are running on right now and the web site says it's "considered stable".

_________________