Planning Game For Wiki Refactoring

It occurs to me that a lot of the mess here is due to a lack of prioritization. What are the worst messes? Well, that'd be the PagesToAvoidAtAllCosts. But some of them are pretty unimportant here. Which are the highest priorities to fix?

The PlanningGame is a reasonably well accepted way to weigh priority against difficulty. This page is intended as a starting point for using it to FixOurWiki. If we each put down at least one page we'd most like to see refactored, and add votes for and against the others our peer wikizens mention, this might even be useful.

RefactoringCandidates:

MemesShmemes

BinarySearch (done, now BinarySearchCodingChallenge? needs work) CanadaCountry ChemicalVsDigitalCameras CommentTheWhy ShieldPattern


If each page to refactor is a UserStory, then it should not only be given a value (nomination and votes), but also an estimate of difficulty to refactor. A few measures: page length in bytes, number of signatures, number of horizontal rules, number of bullets, depth of bullets... In the spirit of the PlanningGame, the estimates should be sorted into three bins: Low, Medium, and Hard.

Of course, there are already some good lists of user stories out there: RefactorByMerging, MergeMe, RefactorMe, TooBigToEdit, TooLargeToGrasp, ThreadMess...

(This is kind of a cool XP analogy. I wonder if we could get Ward to play coach or customer? -- IanOsgood)


Perhaps we could value pages along a second axis as well: investment. I presume the Pro/Con votes are how much wiki folk think the page needs refactoring? Investment would show how much a page matters to the wiki community. I posit that a page that needs refactoring but has no investment has the easiest refactoring of all: swift deletion. Sometimes it is easier to burn down the forest to allow new trees to grow.

If a long page is named to inspire controversy, my first impulse is to delete it: GrossDeficienciesOfUnix, HowCanSomethingBeSuperGreatWithoutProducingExternalEvidence, IfSmalltalkIsSoGoodWhyDoesNobodyUseIt, LanguagePissingMatch, MemesShmemes, MistakesOfRogerPenrose, ObjectOrientationIsDead, OriginalGreatFailureOfXp, PdfSucks, ProgrammersAreNotProfessionals, WimpIsBroken, and XwindowProtocolShouldBeStabbedAndBurnt. -- IanOsgood

I'm also tempted to delete pages full of RK invective: KuliszContraSparks? isn't really an example, just poorly named. -- IanOsgood

I also kinda wish TopMind would try to condense some of his contributions. There are a lot of "OO vs Table" pages in the long pages list. -- IanOsgood


Pages I would add to the list show up in http:longs.cgi but which ought to be short.


As of 20060921, there are 31958 pages in wiki. 2482 of them are OrphanWikiPages. 1401 of them are OrphanNonHomePages. This is a huge amount of material that remains unreachable except through FindPage, LikePages and similar mechanisms. This is not a large amount of pages to deal with if a few people are involved.

I believe that great value could be brought to this site if we attempt to ensure all these pages are either correctly linked or deleted if they are obsolete. I've already made a start, and intend to continue over the coming weeks. -- EarleMartin


See also: PagesToAvoidAtAllCosts


EditText of this page (last edited September 21, 2006) or FindPage with title or text search