IFWiki and Baf's Guide links (and the future of IFWiki)

First, thank you, David, for all you have done with IFWiki. You have put a ton of work into it.

The impression I got from chatting with various people about this (admittedly a small sample size) was that, when people want information about a game/work, they more often tend to go to IFDB, not IFWiki. I think one reason for this is that IFDB has the advantage of having many more contributors, so things tend to be kept more up to date. It may be that the answer to the question “What are people more interested in generally” is different from the answer to the question “What are people more interested in when they go to IFWiki.” But I’m only going by a few conversations, so maybe that doesn’t reflect the majority opinion.

And, to be clear, I’m not trying to suggest that works entered in major comps are more “worthy” than other games. It’s only that, my understanding was that cataloging individual games was not the original intent of the IFWiki–and also is very time consuming–and, with the existence of IFDB, may not be the most valuable thing the IFWiki has to offer–and so it might make sense to catalog individual games only when necessary (i.e. if there is an existing page that refers to the game). What I was thinking with jams is that entries don’t come out all at once, so pages have to be updated more often to be kept up to date, and some jams are very big (e.g. Twiny Jam). No offense meant to anyone.

Re: automation efforts, I did write a python script to automatically generate individual work/game page stubs based on info from IFDB, and autogenerated stubs for all the IFComp 2015 games. But I don’t know if anyone has actually used these. So the impression I got was that, even if the means is there, adding game pages is not something that seems terribly important to people. Or else maybe it’s just that it’s time consuming compared to the edits people seem more likely to make (adding review links to comp pages, for instance).

Also, I’ve been looking at Wiki bots and working on trying to set one up to deal with dead links, at least. Running into silly errors at the moment though.

Yes, this is all true. Originally there was no plans to document any individual games. We started with the Glossary category first, copied from another source, and the Speed-IF comps came soon after, which were originally documented on Gunther’s site. It was my hope that the wiki would concentrate on all the sorts of IF info that Baf’s Guide didn’t provide, such as programming and storytelling and puzzle construction tips for authors; tropes and playing tips for players; and various other stuff like ‘agency’ and ‘mimesis’ for the theorists.

But I don’t think many people understood that. People wanted to know about the games themselves and when the next competitions were. I was told it was “stupid” to have an IFWiki that didn’t document the games themselves. And since I wanted to be a benevolent dictator administrator, I tried to give the people what they wanted.

And we did try to go beyond the duplicating effort when we could find the time for it. I wanted to show a version history for the games, which is something neither Baf’s Guide nor IFDB attempted to do. Adding who tested what game seemed a good thing to do. And I really wanted the Notable Features sections to take off, summarizing the points made by reviewers; I was really hoping having healthy Notable Features sections would help support the original intent of the wiki more. But it takes a LOT more work to write up new content instead of copying content. It was really easy to add 10 more game pages at a time instead of making a page like “compass rose” or “infotater”.

I love the betatester listings and other info not found at other places. Combined with the ease of searching (after all of these years, I still can’t do an advanced search on ifdb without looking up how each time), I refer to the ifwiki almost whenever I’m looking for something.

I think the main problem is that it’s a wiki but most of it was the burden of one person; I don’t blame David for getting burnt out. While getting an editing account on the ifwiki can involve a couple extra steps (because of a bot problem years ago), I encourage anyone who cares to take those steps and start editing away.

First, I agree on huge props to David for his work.

Like Roody, I’ve found that when I am curious about additional info like betatesters etc, IFWiki was the place that had it. And when a game’s reviews are spread out across multiple blogs, which keep dying or being moved, often there was a list of review links for the game in IFWiki. When those reviews aren’t there, the Google process of digging up what’s left feels pretty Klunk.

I’ve been a very intermittent editor, and mostly attended to games I tested or made, using prior or personal knowledge to make the process easier. It does take an age to add even a game you know a lot about if you try to complete most of the current template.

When bg made game stubs for the last IFComp (in light of multiple onlookers’ horror that IFComp game stubs weren’t created by automation), I hoped to do some, and ended up fleshing out only one, and for a game I tested.

I will be glad if all Baf links are hidden at least, because seeing them and also attending to putting in placekeepers for them when they’ve never done anything during my whole time with IFWiki (6 years?) has been variously tedious or offputting.

-Wade

Hmm. I don’t want to prevent people from adding details (testers, review links, or what have you) if they want to. And the existing template is nice for a really thorough page. But what if we were to say somehow that the required minimum info required to make a game page was much, much less? As it is now, I never want to make game pages because–even if the stub is automatically generated–there lots of little pieces of information that must be checked to make sure they’re accurate. I’d rather have lots of tiny game pages that point to IFDB with a few details that can be quickly checked, than red links all over the place. And people could still add testers, reviews, etc. if they wanted. Any thoughts on this? (From anyone?)

Fortunately the Baf’s Guide links are done with link templates. I’ve now changed the templates for Baf’s guide games and people so that it will point to Wayback Machine results for 2012, which appears to be the last full year Baf’s guide was fully functional. It would be better if there were a way to return an actual snapshot within the given time frame, rather than making you choose a date, but so far I haven’t figured out how to do that with just a link.

Is anyone opposed to deleting or hiding the output of the “Baf wanted” template? ifwiki.org/index.php/Template:Baf_wanted

I’m sorry people gave you grief over this.

Do you think it’s worthwhile to do a poll or something to see if people’s expectations/desires for IFWiki have changed at all?

There’s some info here about doing MediaWiki backups: mediawiki.org/wiki/Manual:Backing_up_a_wiki

Could I just ask a quick question?

Looks to me like there’s some effort in not letting go of Baf’s Guide entirely - masking links instead of deleting them, and whatnot. The IFDB also does this, there’s lots of pages linking to the Baf’s Guide entry (dead links, naturaly).

If that’s true, then… why? IFDB has superseded Baf’s Guide in every way, hasn’t it? The guide certainly deserves an entry in the IFWiki for the wonderful thing it did while it was active, but right now it doesn’t serve an actual purpose, it’s not even functional, so presumably every thing that links to it can just be axed, right?

If that’s not true, then just ignore me and continue talking. :slight_smile:

For IFWiki backups: I think you should contact Carl Muckenhoupt (aka Baf). He’s got the server with the IFWiki and the necessary permissions, and he once told me on the ifMUD that he could easily set up a “cron job” to send a dump regularly.

I’ll send you Carl Muckenhoupt’s e-mail address by PM, bg.

Thank you!

Thanks! I’ve emailed him.

I know I should really do this on my own, and that’s why I haven’t brought it up, but one thing that would be really useful is including 2014’s (and after today, 2015’s) xyzzy winners on pages like ifwiki.org/index.php/XYZZY_Awards_(Best_Game . I look at these pages all the time. Again, I know I should do this myself, but I haven’t after a long time, so I thought I’d bring it up.

Update: The Baf’s Guide templates should now automatically send you to a snapshot of the page from sometime around May 1, 2012. I suppose it’s possible some individual links won’t have a snapshot around then, and maybe those will need to be fixed individually (because for them the nearest snapshot might be a year later, so you’ll get the “Baf’s guide is down” page). But if anyone sees this happening a lot, please let me know, and I can move the default date earlier.

Also, I think I’ve got the bot working. I’ll continue to work on this to see if we can replace dead links in a more automated way.

Glad to help.

I’ve linked to a survey about IFWiki in this thread.

That’s probably not a good approach because the data is changing over the years. I would prefer a continuous mining effort to a single one, some script that would keep the wiki articles at least partially fresh.

I don’t think having a giant pile of … (works) is a good suggestion, sorry. Categorization is one of the strong points of any wiki.

You’re implying that only the works entering the competitions are noteworthy, which is a very dangerous thought.

IFDB is just another wiki project. If IFDB pages are less a pain to keep up-to-date, we should think about making IFWiki even easier to add pages onto, not ditching these lists altogether just because there’s another more successful wiki.

Again, you want to kill off a part of a project just because there’s another one that’s more accessible to you. That’s not a good idea even in the short run.

You know why IFDB is better than IFWiki? Because it’s more automated and formalized. Instead of one giant text area it has a nice form to describe the game or a competition. It compiles all the lists and polls automatically.

IFWiki could do that too using existing extensions. This is not hypothetical: Russian IFWiki uses some of them already. And IFWiki can do much, much more - IFDB can’t hold the articles and can’t even do links between games.

I’ll just leave these two links here:

  • Semantic Mediawiki - provides autolists (list all the games by this author, all the games in this genre, in this comp…) - essentially Data Mining For Dummies
  • Semantic Forms - provides easy-to-use forms for pages. Think about having an IFDB-like form for describing new works.

I just remembered–it has been pointed out before that the captcha question asked of new users on IFWiki isn’t something that all IF fans will necessarily know the answer to. I’m in favor of this changing to something more widely known, though I don’t know what, or how such a thing gets changed.

In retrospect I worded this very poorly. My intent wasn’t to disallow people from creating works pages. (I don’t want to disallow any of these things we’re talking about.)

But maybe you’re also pushing back against the idea that not-yet-created works pages that have existing links to them should be a different priority level than not-yet-created works pages that don’t have existing links to them (or that have fewer existing links).

I do take your points, thanks.

Thanks!

Well, the previous captcha was another IF question but slightly more google-able, and spammers were taking the time to look up the answers, so we’re not just trying to outsmart bots here.

Wow, those are dedicated spammers. Begs the question, though, of how far you can go. Reminds me of all the copy-protection software that caused problems for the people who actually bought the game. In particular, I had to download a NoCD crack to even get Discworld Noir to play when I upgraded to Windows XP.

Also, it depends on what the captcha’s supposed to do. Is it trying to keep out bots or spammers?

(I also am really curious, what sort of spam were they doling out? Editing a Wiki is not quite the same as posting on a forum thread!)

I perhaps shouldn’t be commenting (I’m a casual user of the wiki, at best) but I’m in favour of the Wiki doing things that other sites don’t do - i.e. more in the way of theory and craft than comp listings which would be duplicated by IFDB.