Archive/Media Wiki Conversion

User:Nibbler: Please change to using the Media Wiki software, as many people already know how to use it.

User:Gus: My current feeling is that this would be very disruptive to the community. However, if the community feels that being more Wikipedia-like would be the proper play, I'll look into converting the data and see what it will take to change such things as Random Item and the Generators to work with the new format.

I do not see the value -- opposed to the cost -- myself, but if that is what the Tropers want, that is what they will get.

User:Bluetooth The Pirate: If someone can't figure out how this system works -- after a bit of practice, of course -- I'd say there's bigger issues there than just unfamiliarity. I love the system the way it is, and I personally dislike the MediaWiki system. Then again, I used this site before I used MediaWiki.

User:Looney Toons: I've never used MediaWiki -- I'd really hate to have to relearn how to code things and use the site.

User:Ununnilium: Yeah, the current system is just fine. Many people already know how to use this, too.

User:Nibbler: Well I think it's a question of "where do you want to go with the TV Tropes Wiki?". Do you want fresh blood or not? You guys who discuss on this page are the "insider circle people", and of course you like the the current status. We should get answers from people who are not "regulars". And if you look outside, there are like 100 Million english speaking internet users out there who don't know the TV Tropes Wiki yet, and of those, about 1.5 Million have had some experience with the MediaWiki software, while none had some experience with your wiki software (because it's unique, if I'm correctly informed?). So if you want fresh blood, the Media Wiki could offer some advantage, because people could "dive right in".

User:Bluetooth The Pirate: New people show up and junp in all the time; after all, I did once, as did you. The only barrier I'd see is someone so used to updating MediaWiki-based sites that they trip over the differences, and have a few hiccups on their early edits. Many mechanisms have been incorporated to deal with this, notably the lockout, and later the unlock button, plus documentation included and written by the users; Looney Toons wrote the set of excellent page format guides, and that helps a lot of people, if they bother to read them. As to the software being unique, I don't think Gus wrote PM Wiki itself, but he did add a lot of extensions to improve usability and include specialized features and shortcuts. (Thanks, man!)

The question is, do we want to make the entire userbase re-adjust to a different system, and do a lot of work re-coding and adapting the extant pages, just to eliminate the learning process for a set of users that might take fifteen minutes to get the basics down? They'd need to learn the context of the site content before doing any major edits anyway, so I'm fairly sure it's just a huge hassle with no major benefits.

User:Ununnilium: Not to mention, this is the first complaint I've ever seen about the wiki software we're using. (And it's not even a complaint about how it works or what it does - just that not everyone uses it.) And where do you get that 1.5 million statistic, anyway?

User:Looney Toons: Plus it's a complaint from someone who hasn't really been a part of the community here for very long. He's metaphorically just walked in the door and practically the first thing he says is that he doesn't like how the place is run. Nibbler, your suggestion would have a lot more weight if you had more history here.

User:Morgan Wick: Actually, I agree with Nibbler in some respects. I see some upsides not related to "expanding our number of contributors", mostly from my Wikipedia experience: On the other hand, there are drawbacks: Media Wiki's pitch for your services is here. A very technical comparison is here. Note that this may refer to a version of Pmwiki this wiki does not run on. (I would LOVE if this wiki had a "Preview" function.) Of course, maybe I don't know anything and what we have is perfect. Or it may just be that I'm the first person writing on here who started on Media Wiki instead of Pmwiki, and everyone thinks whatever they started with is the greatest thing in the world and so incredibly easy to learn (and everything else is so difficult). (In any event, this is more Wiki Tech Wish List material.) 
 * Page histories could be researched all the way to the beginning. (Note that this assumes we would be able to port page histories, and rescue the histories that are now being cut off.)
 * May be more customizable. Right now, every time a change needs to be made, it's time consuming and can have unintended consequences, and often doesn't work as intended. Switching could make Gus's life easier in the long run.
 * Broader possibilities for page names. Right now, if you throw anything that's not a letter or a space into a page name, the chances are about fifty-fifty if it's a number, and zero if it's anything else, that it'll be valid. The result: weird page names like 24 and Ranma ½. The Wiki Word convention screws up capitalization schemes as well; a title intended to be "Recycled IN SPACE" comes out "Recycled in Space", which took me a while to figure out what it was saying. (Granted, it wasn't all that great a title in the first place, but that didn't help.) Related is the dropping of one convention for linking to multi-word titles and another for single-word ones. It might also be easier to link to plurals. But I get into "make it easier for newbies" territory.
 * Trope Talk entries could be moved into their own "namespace", as would contributor pages. In a radical form, entries could be sorted into namespaces based on what media they are, or what type of entry. However, this usually shows up in the title bar. Interlinking between namespaces and making them show up on RC would be fairly easy.
 * Greater recognition of, say, the creation of new pages.
 * The language of Wikipedia and the language of PM Wiki are very similar.
 * Edit summaries would likely decrease the use of the discussion pages for little more than clarifying edits. There would still be a need to use those pages for more complex cases.
 * Need to insert a certain type of very rigid style, for example, format and genre info? Or need to blare out a certain "flag"? "Templates" have you covered. In a related story, built-in caption support.
 * Are you sure it's that easy to learn? Maybe it's just being used to Wikipedia, but I've had quite a few struggles with the code even knowing some of the basics. The help pages are not the most visible things in the world (it's not so much a matter of "bothering to read them" so much as knowing they exist, the boilerplate on the right side of the edit screen may be ignorable and it's a hassle to go to pages when you've got a lock clock on you), and I don't think there's a quick and dirty thing with everything you need to know on it. (Other than the aforementioned right side, which is VERY minimal.) This is more a problem with what this Wiki itself has decided, and not with the software.
 * "Categories," in every example I've encountered so far, do not use navigation buttons for scrolling through a category.
 * Likely a need to require a login procedure to append a name to your edits.
 * More problems I've probably forgotten or have been mentioned above.

(ETA: According to http://en.wikipedia.org/wiki/Special:Statistics, there are 1.8 million registered users on the English-language Wikipedia alone, not counting users who never log in.)

User:Ununnilium: Now, see, that's a good argument for it.

That said, I'd rather get the wiki upgraded with these features than change over to a wholly different system. I've played around in Wikipedia before, and I really do think this one's easier to edit, at least for the purposes we use it for. If we wanted to do a lot of detailed, technical editing with special links and all kinds of stuff, MediaWiki would probably be better. But with PMWiki, or at least the flavor of it used here, it's easier to dash off a quick edit.

(1.8 million? Wow.)

User:Dark Sasami: Couple more things:


 * MediaWiki would fix a lot of the problems I tend to run into every time I come here, such as searching for titles or losing histories.
 * With MediaWiki, it would have a more trusted and authoritative look and feel. That's a pro and a con.

Personally, I'd love to see it go to MediaWiki, but I can understand the reasons that it might not. Full disclosure: I do a bit of dabbling in Wikipedia and am very happy with the range of tools that it provides.

User:Ununnilium: I think the more casual look and feel we have is definitely an advantage. Wikipedia looks more official and authoritative; TV Tropes Wiki looks more cozy and approachable.

That said, we could change the look and feel without having to change the underlying system, couldn't we?

User:Morgan Wick: Marvel Comics has a wiki-style database of character profiles on its web site. You'd never know it by looking at it, but it's on Media Wiki.

User:Ununnilium: Huh, awesome.

User:Nibbler: I'm sorry if I offended anybody, I think I might need to clear up a few things and introduce myself: I'm a writer from Europe and found the TV Tropes Wiki in November 05. I spent a few hours in it, and its breath and depth, and the value it offered just - well, blew my mind! :-) I absoutely love this wiki and have great respect for you guys and gals. I'm a regular contributor to several Wikis using the Media Wiki software (including wikipedia en, de, wikibooks en, de, wikinews, and several smaller Wikia sites) and altogether, in the past 3 years, I've made close to 5000 edits on them.

Media Wiki offers me to jump right in with all those sites. I'm an eager user of the watchlist feature (gives you an aggregates summary of the edits of the pages you selected to "watch"), the "new pages" feature (shows you what's new and being worked on), the full history of every page and every user with the ability to see any page the way it looked at any point in time, the markup that eases nested discussions, and of the exportability (download all of the english wikipedia with 1 click if you wish: http://download.wikimedia.org/enwiki/20060717/ :-D)

I've also used Media Wiki internally in some companies. I was able to set it up in about 30 minutes, and I'm no hardcore techno geek, mind you :-)

Media Wiki seems to become the standard for many wiki implementations, because it's easy to administer (really!), has every feature you might need, has a gentle learning curve, and (most importantly): you might want to be sure that your wiki engine is still being developed further in 20 years time, and Media Wiki with its big momentum probably has the highest chance of all the Wiki engines to still be around at that point.

After I first found TVT, I read about 65% of all the articles over the course of a few weeks. But because of its differences (features + markup) I made only half a dozend edits (under a different alias I unfortunately cannot remember anymore). I then suggested the move to Media Wiki on some page (can't remember which one) back in December. Since then, I came back a few times, but those missing features kept me being reluctant.

Case in point: Lack of database dump. I asked myself if the time I put into this project will be well spent, or if suddenly the site together with my edits will just vanish. With a database dump, I still have got the information, and if the dump is formatted mediawiki-like, I could just set up another instance, feed it the dump and get back to the data in an hour or so.

That's why, in the past 8 months, I visited the wiki every so often, but put my free time into other, different projects. You're of course very welcome to keep the wiki the way it is, and I will still love and adore your work, but I would probably also keep standing on the sidelines.

I'm just 1 person, but I really think that a move to the Media Wiki software would greatly enhance your audience and also quickly give all the current users access to great timesaving functions you wouldn't have thought to be able to live without after only a few days. Cheers! :-) Nibbler

User:Gus: There seems to be a case for Medi Wiki. Certainly, I should be glad to get away from the horrendous data format and procedural coding that Pm Wiki uses. The next step toward understanding the impact would seem to be a plan. In the next few days I'll float a feature list (things we still want that are not offered by Media Wiki), then a list of steps needed to get from here to there. At some point, we'll need additional technical hands on board. There is no way I'll be able carve enough time out of the day job to do everything.

Oh, and ... Nibbler, we have not disappeared in the last 30 months, or lost any more than 10 edits in all that time, so while the change is underway please feel secure in adding your bits. ;-)

User:Red Shoe: Well, I'm agin' it, as I'm curmudgeonly. But more importantly, I don't find "Wikipedia uses it, so everyone else should too" to be much of an argument. Next, they'll be calls for IE-specific features, since IE is the most popular browser. There's more than one wiki platform, and that's a good thing.

Also, I could never get used to Media Wiki style, and use Pm Wiki myself for my own wiki-based projects.

Not that "it would mildly annoy Red Shoe" is much of an argument either.

User:Ununnilium: Yeah, it doesn't really appeal to me too much either. I don't want to just blindly be against change, but OTOH I don't want to be for change for the sake of change. We'll be gaining features, but what features will we lose? Looking at the lists, I don't see anything I really want; there's some stuff that'd be nice to have, but IMHO it's not worth overhauling the entire wiki.

User:Tabby: As someone who's still more or less "fresh blood," I have to say that this is the first wiki I've contributed to and I've never had any problems with the current setup. That which I couldn't intuit or glean from the info on the side of the 'edit' page I figured out in no time at all by looking at other people's contributions. I don't have the technical experience to argue much else, but I think complaints about the 'learning curve' are unwarranted.

User:Looney Toons: I have to agree that unless there is some major benefit to changing over, some killer feature that will kick the entire wiki up a notch in some way, there's no point to it. Just one (new) user who doesn't like PMWiki. (Oh, and while I'm here, my hat's off to you, Tabby -- I've seen way too many new contributors (three or four this last week or so, in fact), who don't seem to know or care that there are easy ways to mark up the text for best effect. As the author of some of the guide and template pages, it's so gratifying to see someone who takes the brief time needed to learn the few things that make the Wiki work. Thank you, Tabby!)

User:Gus: Maybe the question needs to be re-framed, a bit. There were a number of features pointed out -- especially by Morgan Wick's excellent breakdown contribution above -- that would be of use. Perhaps what is needed is to enumerate which of those we care about, then to take a look at what least-effort path is to arrive at the functionality.

Jumping ahead of that process -- as I shouldn't -- I'll guess we can get to the desired functionality incrementally, rather than making a wholesale change to another engine. In fact, we are only a few hundred lines of code away from having a locally-grown engine, as it is.

I'll extract that feature list, though, so we can touch that base. It will be useful on either path. That list will show up soon in Wiki Tech Wish List.

On the philosophical tip ... after a day or so study of Media Wiki, I came away with a sort of bureaucratic aftertaste. Things are looser around here. It seems like that bureaucratic thing is built into the technology with Media Wiki. Most of that seems to arise from the need to control what I can only describe as 'spurious' posts. A problem that comes with a very large body of contributors. As a technoid, I would rather be trying to provide more tools for working with the topic, rather than working with the wiki-ness. I would like to see things like -- to throw out some notions:
 * Being able to flip through a trail of entries related by a tag, like 'Emo' or 'Empowerment'. Better yet, a continuous display of those entries.
 * Being able to display all the "This series provides examples of..." on a continuous page at the touch of a button.
 * Similarly, being able to see a continuous page of the entries that refer to a trope.

OK. I'm getting off into feature-list items. My point is that it seems more valuable to me put tech-time into the things that add up to "Fun" in my head, rather than into controls.

User:Ununnilium: Yeah, I agree. That's more important unless the user interface is actively blocking contributions.

User:Seth: I like the way it is TV Tropes is a very enjoyable website, i use a few Mediawiki sites but this site has a better feel than any of them. The colour scheme the size everything is perfect for what i use it for. Wikipedia has a lot of people but it is very prone to vandalism the more people you get the more likely one of them is an (Expletive Deleted) who will try to mess things up. Also websites like Encyclopedia Dramatica which are mediawiki based comedy sites dont work. The feel is all wrong they are too stuffy for a comedy website (which is what i consider tropes to be, perhaps comedy is the wrong word though).

That being said a plus of Mediawiki is the preview function mentioned above, if this site had a preview function i would see no reason to switch. I guess this is just a dont fix what isn?t broken mentality though.

User:Looney Toons: We used to have a preview function, long ago when dinosaurs walked the wiki, but I'm not quite sure why it disappeared. We changed software, maybe? I don't remember.

And while I'm at it, let me just do my little bit to be curmudgeonly, and note that Nibbler -- who seems to have started all this -- appears to have vanished as quickly as he arrived; after dropping his disguised advertisement, and making his claim to have edited a few pages under a forgotten pseudonym, he never did post anything again.

User:Gus: Yes, the previous preview funtion was part of the first engine. It was about the only feature that worked. ;-) Nibbler notwithstanding, this discussion has highlighted that there are some features we could use. Nothing does it all the way we want to do it, so we'll roll our own.

User:Random 832: My 2 cents on the matter of "differences between MW and here that I like better there" - camelcase. I don't like it. I especially don't like (just ran into) the fact that you have to use even in places that don't make sense to require it. Why doesn't pagename text to link  work unless pagename is a Wiki Word? I would prefer if there was just one kind of link - if a page title you want is one word, deal with awkward Capital Ization - or the better solution would be to get _rid_ of Wiki Words. I don't really care that much about mediawiki itself, but moving towards mediawiki-ish features like no-Wiki Words, namespaces, and MW-ish markup would be nice. The one feature I've missed the most: edit summaries/comments.

User:Ununnilium: But Wiki Words are one of the most distinctive features of this wiki! It'd be blasphemous, or at least kinda meh, to get rid of them!

User:Looney Toons: Besides, the point is moot. Gus is building a custom Wiki software from the ground up, more or less. Media Wiki is no longer a contender.
 * User:Heatherly: Thank God. I've just spent forty-five minutes trying to edit a page on Wikipedia, carefully reading through all of the coding info for that particular template, only to end up with page salad. I ran screaming back over here where I can intuitively figure everything out just by, y'know, looking.