It can be hard for subcommunities to leave large Wikis. See WikiMass.
If subcommunities stay on the same wiki, they need PageClusters.
When subcommunities leave and found a new wiki, they interlink to the old wiki using InterWikiLinks.
The way you say "splintering the community" makes it sound like a bad thing. I am not sure what would happen is bad (or at least, worse than what we have now). The system I use most, CIX, is effectively a collection of overlapping communities. What approach do you think is better? -- DaveHarris
I'm not sure it is such a horrible thing provided that each subcommunity has enough critical mass to sustain itself. After all, in the end the pages will interlink. You would like some level of quality between them. Also, you might lose consistency between the subcommunities. What is very likely to happen is certain subcommunities being of extreme poor quality. Right now, the quality is more or less even throughout Wiki because people are forced to at least acknowledge the presence of other pages. That's one reason why RecentChanges is oldest to newest--it makes people keep looking over older pages.
It may be superior to do what Cliff and I have done here with MeatballWiki: create a separate Wiki with a different focus and interlink between them. But you can see the number of pages we know from WikiWiki that we naturally link to that don't exist here. For instance, you linked to AnalysisParalysis on ObviousNextStep which really exists at Wiki:AnalysisParalysis.
"What is very likely to happen is certain subcommunities being of extreme poor quality." - Yes, I agree. I am not sure that's a bad thing either, so long as the subcommunities are able to die. If something keeps them alive, I would hesitate to make the "poor quality" judgement call.
It follows that part of the trick of scaling is allowing the subcommunities to die cheaply, without the corpse taking up too many resources. A (tentative, possible) problem with RandomPages is that it necessarily does not discriminate between poor quality pages and good ones, and so may re-animate monsters best forgotten.
I think most people are quite happy to let monsters die. Some people, however, are attempting to achieve some sort of "popularity" on Wiki. These are the kinds that spawn pages pointlessly in order to get attention back on RecentChanges. DemaGogues, definitely. -- SunirShah
I think most of Wiki has fairly strong consensus. Just a few rotten apples who've spoiled the bunch. Indeed, I think many people are interested in talking about Wiki, just not so much nowadays. --ss
I know what you mean :-) Wiki will calve off sub-communities whatever happens to recent changes. I think it will often be better to support sub-communities within a single Wiki by jiggery-pokery on recent changes, then to set up an entire new site. -- DaveHarris
Over at UnrealWiki, we're a facet of the larger Unreal editing community, which is already spread over about a dozen forums hosted by various news or tutorial sites. Being new to wiki in general, I don't know if this has implications on the way the wiki will function. We already appear to have gone through the WikiLifeCycle several times. The plot thickens because we had a URL move which was widely reported in Unreal news sites, which prompted a flurry of new writers. Some have stayed and have posted & taken part in refactoring and building the site. I'd say we're maybe back in stage 4: close knit pioneers.
Another interesting difference is that the average age is younger. (I say this very roughly, as I have no idea what the average age is in other wikis, or indeed if there is one.) We may be more at risk of vandalism. The majority of the members of the supercommunity don't remember the earlier days on the net when you carved up a rough html page with a grey background. They're used to giant commercial sites with flash graphics, and wiki seems a little lo-tech to them. Furthermore, as my co-founder Mychaeel put it, "they're used to the web as television". Not only are they not accustomed to the idea of creating content oneself, when they do they try & ape the giant sites. (I could link this up to the malaise of our times, StyleOverContent?...) I probably had a salient point to make here, but I seem to be losing my way. Retreat & rethink... :-) -- Tarquin of UnrealWiki
It's become more and more common to see wikis as a tool (content management, discussion area, sandbox) used by a larger community group rather than as the primary community itself. Places like MeatballWiki are in the minute minority, taking back seat to projects like SeattleWireless, UnrealWiki, and countless thousands of other project wikis. Even this place originally was only meant to be a smaller facet of a larger effort, but it turns out that the wiki serves every facet of Meatball that is currently important except for the rarely used mailing list. -- SunirShah
I don't think that means that wikis aren't suited as communities, but rather that they are also suited for projects. I bet a small proportion of paper is used for newspapers, but that doesn't mean that paper isn't a great medium for newspapers. It is also true that wikis are a small proportion of all online communities, but I think it just hasn't caught on yet. -- BayleShanks
There are ProjectWikis?, but sometimes they engender a community: WikiPedia. In some cases, Wiki can be more that a project tool, it can be an extra platform. I communicate with friends face to face, on the phone, with email. Similarly, some online communities use several platforms in parallel, eg web pages, web forums and IRC. I think that's what I'm trying to achieve with UnrealWiki.
I'm not really sure that this is CategoryOnlineCommunity; this page is about online communities in general, not a specific online community. -- BayleShanks