Crazy Things That Might Save Wiki

Discussion of ideas to save wiki from the WikiProblems. Some argue that ThereIsNoTechnicalProblem and that social solutions can work at ReconcilingIrreconcilableDifferencesOnWiki despite their cost.

This is NOT a new topic and there is plenty of background material to assimilate before one can speak intelligently and constructively on it. Dozens of prior online communities have come up against the exact same problems and most of them have ended up documenting them. Some of these are referenced in ClayShirky's A Group Is Its Own Worst Enemy (http://www.shirky.com/writings/group_enemy.html).

Goals

Have Wiki be an open society, much as it has always been, but more tolerant against WikiSpam, WikiVandals, and EditWars (see WikiProblems).

The phrase "...more tolerant against WikiSpam [etc]..." is confusing - are you suggesting tolerating spam and vandalism in the sense of letting them be? Perhaps a better phrase (if this is what's meant) is "more resistant against". Otherwise this doesn't seem like a laudable goal. :)

 Another person

I have a problem here. Who wrote the goal statement? If Ward did, that's the end of it, except for asking for explanation so we can comply better. If he didn't, it might be available for grounds-up type explorations. -- dl

I have thoughts that "may have some relevance" to this discussion. Basically, the behavior and expectations of todays "wiki newbies" could be very different from the well-mannered and considerate initial "invited guests" in 1995. If we accept that, behavior management mechanisms may have to respond differently as well.

Ten years ago, WWW was fairly new; a lot of high school students then did not know about it. Today, primary school students could be attracted to it, as I perceive a lot of people in the education sector have embraced the concepts of collaboration. Being the original wiki, this site draws in a lot of much younger participants whose technical skills could be more developed than social skills (compared to ten years ago).
Pace of society may have significantly increased. I know FutureShock theory came up decades earlier, but the culture change of the masses lags behind. Today's twenty-five year old have seen media images of "falling berlin wall", and a twenty-five year old in a few years time will have seen "9/11" images while they were in primary school. The sense of "self-importance" in a fast changing and uncertain world has never been stronger.
Respect for authorities and institutions have diminished a lot in the past ten years as well.

Ideas

The best ideas are ones that take advantage of the distributed nature of the internet and the general good will of most of its inhabitants.

Fix it with politics. This is not likely to work.
Incremental improvement of the existing wiki engine. See WikiSpamSolutions and WikiVandalismSolutions.
Switch to a better wiki engine. Have WardCunningham hand pick some administrators or else call a vote like WikiPedia.
Design a new medium to supersede the wiki. The present situation requires a shift in the balance of political and social forces in order to resolve. Political compromises are embedded in any groupware implementation. See PrinciplesForTheNextWiki. DecentralizedWiki seems to be a suggestion along this line.

In terms of NeverWriteaLineOfCodeWithoutaFailingTest/TestDrivenDevelopment, we already have numerous failing UnitTests, even if only informal ones. Some of them have been semi-formalized on WikiSpamSolutions and WikiVandalismSolutions.

My suggestion is to SeekFirstToUnderstand. What do we want to "save" anyway?

I have been reading and sometimes writing on this Wiki since 2000, I agree that in past year or so, this Wiki's "mind" seems to have "drifted" away from programming. E.g. a programming focused Wiki would have seen more discussion on SOA, web services, or even more about the recent improvements of the job market, etc. Instead of focusing on EditWar and other spam.

So, is a more "programming focused" Wiki what we want here? Or merely one that is no longer focused on EditWar?

To refocus wiki on programming, we have to first take the focus off edit wars. Even if we don't refocus on programming, presumably we should still take the focus off of edit wars. For the first step, we know what we need to accomplish.

Put all the on-topic HolyWar fights in FooBarDiscussion? or FooBarRants? for the FooBar topic, for example. That way cruft can exist without interfering with the primary topic. In other words, put the funny stuff in a dark corner room of the house with a marked door instead of outright deleting it. (IOW, ThereIsNoTechnicalProblem)

That assumes that the participants in such acrimony would voluntarily stay there. Too often, they don't. In some cases, a gentle reminder fixes the problem. In other cases, people may want a wider audience. And in some cases, persons convinced of their inherent rightness confuse their opinions on a topic with established fact - and thus claim that their rants are properly placed on the main topic page, not the discussion page (which, after all, is where all the opinionated nonsense goes).

{There are not a whole lot of established facts in software engineering, in which case everything should be in the rant pile?}

Put the Edit function in an icon down at the bottom of the pages among many icons. One would have to read the FAQ page to know that one of the icons is the editing link and which one it is. This would reduce casual visitor vandalism. (IOW, LetsInsulateOurselves, but from newcomers too)

A humble suggestion for a better Wiki: pull the large content blocks out and put them into their own, smaller wikis, and use Ward's existing SisterSite technology to link them together. WikiWiki then becomes a sort of giant nexus where it all overlaps. For example, the PortlandPatternRepository could be extracted in its entirety to be itself again, and all links to pages within it would still work, as well as outbound ones to Wiki. Likewise with the ExtremeProgramming material. No deletions required, everyone is happy. Hopefully. -- EarleMartin

That's already been done. It's just so seamless and transparent that you can't tell.

There is no easy nor semi-automated way to split out a wiki given the current state of the technology. And regardless, I don't see how it would affect C2 wiki's major problems. I know that it would itself create some major problems, including the annoyance factor of keeping track of multiple recent changes, the constitutional bickering every time a new site is created (they all have different policies), and the possibility that these sister sites go under as WhyClublet and ReformSociety have both done.

Sorry, I guess I wasn't clear enough. My idea wouldn't be an automated change; gnomes would have to do the work of extracting content, as usual. The new sites would also still be run by Ward, live on the same server, and have the same policies apply as the mother Wiki. I think this would fit neatly with Ward's style of minimal extra codework. I agree that creating entirely new sites would be a nightmare, as you can imagine; plus, I didn't mean it in a sort of TimeForWikiMutiny? sense.

An integrated RecentChanges for all three sites would be trivial to produce, and I would really like being able to see specific RecentChanges pages for each of the subtopics (in effect, FilteredRecentChanges). Yes, there would still be the major issues that affect Wiki today; however, it would be harder to justify an allegedly OffTopic page's existence in the PortlandPatternRepository, for example than on the wider WikiWiki. The actual experience of using the trinity (I use the word in the same sense Christians do, to mean three that are one, and yet also distinct) should be a seamless one. The one modification to existing WikiWiki behaviour that I might suggest would be for SisterSite links to appear distinct from standard ones in some way, but that is not essential for my idea. -- EM

It seems to me to be a massive effort for very little payoff. I was hoping you had some other advantage in mind than eliminating off-topic bickering. Note that the more seamlessly integrated the different sites are, the less they function to provide people separate spaces protected from each other.

I've been thinking about that since I originally wrote. The more seamless it is, the less obvious any difference from the original setup would be - clearly pointless! So perhaps it should be extremely obvious when you move between sites; this could be achieved with obvious differences between the appearance of each site, and clear delineations of inter-site links. For example (although an extremely bad example, for accessibility guidelines specify not indicating differences by color), links to WikiWiki are blue, links to the PPR are green, and links to the XP wiki are red. Or a small icon (with ALT text, of course) could appear before the link text for SisterSite links. It should be possible to use all three together without difficulty but also without any confusion to where you are.

As to potential benefits, I think eliminating off-topic bickering is a plus! But the major advantages would be being able to use FindPage/FullSearch/RecentChanges more accurately within the scope of each site (in addition to globally). I would love being able to use a pure PPR without having to dodge around OffTopic stuff and flame/edit wars, which I would like to believe would tend to limit themselves to WikiWiki. -- EM

There's a wiki engine that puts little icons next to links depending on the type of link. I don't recall which.

Yes, off-topic bickering is annoying. But what you're proposing would be a massive undertaking. The effort expended would be at least as great as the payoff.

And then there's the fact that FindPage and FullSearch just aren't that useful so multiplying their usefulness by a factor of 2 or 3 isn't a big win. They would be vastly more useful if it were possible to track all edits by a given contributor (like in WhyClublet), or if a small selection of RandomPages appeared on the search page (again, like in WhyClublet). If this were implemented, then I could see the point of splitting up C2 wiki into itty bitty pieces. -- RK

A CurrentAffairsWiki as a SisterSite would certainly be one of the one of the WikisIdLikeToEdit.

Deleting Off Topic pages has a merit, but please consider WikiSig. (After all, this page is called Crazy Things ...) -- HansWobbe

I argue against this idea on WikiSig, but don't consider the suggestion inappropriate within the confines of this page. -- WardCunningham

CategoryWikiProgress