ReworkingProblems

MeatballWiki | RecentChanges | Random Page | Indices | Categories

Reworking a page -- "refactoring" -- is necessary work to increase the quality of pages, but it does not happen very often.

A lot of good content comes into existence because somebody writes a little essay about something. Discussion might follow, and sometimes the original essay and the discussion are reworked into a new essay. This improves the content.

Reasons why refactorings are rare:

reworking has a low priority for most users most of the time
it is difficult work, requiring an intimate understanding of everything previously written
it takes a long time, as it cannot be done both incrementally and thoroughly
there are social risks involved, someone might feel treated unfairly, even a conflict is possible
it may be difficult to find the right time for refactoring, at least in the early stages of a page
it is boring work

Possible solutions:

Do not write content that requires reworking before anybody will consider it "good". If you cannot write a page such that it can stand on its own, do not write it at all, because there are enough people creating too many shallow pages and so many unnecessary wanted pages.
Do more refactoring work. Use "random pages" to look at arbitrary pages regularly. Look at ten pages. Choose one of them and improve it. Learn to refactor by doing it.
Encourage refactoring. Keep a list of pages, candidates for refactoring. Learn refactoring and teach refactoring. It's an art. Not only grok some issue but also how people think about it differently. Reworking is rethinking. Appreciate it when people try it. Give friendly feedback.

See also WhatIsReworking, Wiki:WhyNotEnoughEditingHappens. Currently there is a discussion at CommunityWiki:LackOfReworking about this.

Take the test ... Is it true that so many pages need reworking?

Use the Random Page link (enable it via Preferences) and do a little test now: Click it 10 times. Read the page and check whether it is "good" by your standards. Does it seem to be the work of one person, or that of many that has been reworked?

Of the ten pages I looked at: 7/10 I considered "good", none seemed to be the result of true reworking. -- AlexSchroeder
of the ten I looked at 5/10 were what I considered "good", and like Alex, none seemed to be the result of reworking. -- MarkDilley
I had a bad run. Only 3/10 were good. One seemed to be heavily refactored from discussion, another partly reworked, and the third was mostly untouched ThreadMode. -- StephenGilbert
Maybe I have different standards from others. I got 9/9 + 1 homepage (I decided not to judge the homepage). Out of those 9, 3 of them could use some slight editing, but in general they were "good". As for deciding whether these pages were the work of one or many, how would I tell? A really good rework could have consolidated many contributions into one smooth DocumentMode. More details below. -- BayleShanks

Bayle's Test

BrowserUsageStatistics: good
StanfordPoynterProject: good
JürgenHermann: unjudged
LinkCheck: good
WebOfTrustModeration: good, could use some slight editing
CritLinkMediator: good
QuickWeb: good, could use some slight editing
MetaWikiPedia: good
TextArc: good
HeaderSyntax: good, could use some slight editing

-- BayleShanks

I'm not too sure you are looking very deep. To give some examples, LinkCheck is not good. There is no detail on it, only a ForwardIndex to other pages, one with exactly the same purpose (WikiLint). Once you go to WikiLint, you will see that not only is that a bad page itself, but it was what spawned the shallow page LinkCheck. It also created the meaningless SpellCheck (by giving only the obvious definition; worse, only in the wiki context), and almost created GrammarCheck?. All of those ideas should be folded into WikiLint. I'll do that once this example has been seen.

CritLinkMediator doesn't even give a link to what it's talking about, although I am not suggesting that a simple link (a wiki quiki) would be adequate, as I still wouldn't know what CritLinkMediator is, does, or what I'm supposed to learn from it. If you look harder, you'll see that it was once again created as a shallow "define me!" page off of CritDotOrg, and the entire ball of shallow crit.org pages, and it still didn't even define what the Crit link mediator is.

Now that I've criticized these things, I'm going to have to clean them up. Joy for me. ;) -- SunirShah

CritLinkMediator does indeed define what it is talking about. "A freely available perl script, created by KaPingYee to allow finegrained annotations on arbitrary webpages from arbitrary webpages.". I'll change that anyhow.. -- BayleShanks

I know that we are having this conversation now, and I am assuming that this conversation has been had on at least on WikiWiki, if not here. . . So is there a page that talks about this as explicitly as above? (not that that will be the end all) -- Best, MarkDilley note to self, check out Wiki:AcceptableRefactoring

WHat is the acceptable level of refactoring here on MB? Is it okay to turn whole chunks of thread mode into document mode? Perhaps reworking doesn't often happen because people don't feel bold enough to edit others' words.

I think you can go ahead and rework the page, if others don't like it, they will revert it. That is what happened with the TourBus pages, it is the wiki way, MarkDilley

Same here. Just go ahead. If anybody feels some statements you removed are worth preserving, they can always add them back in. -- AlexSchroeder

Helmut, Dec 19, 2005

If I don't take the two French pages into account, it is 4/10 that need reworking.

HistoriqueComplète -- French, don't understand it
SymmetricCipher -- good, although a bit short
WikiCities -- good, although very short
InternetGenome -- needs reworking, short and not consistent
CommunityWikiArchive -- needs reworking
GroupLens -- needs reworking, also very short
DontLookAtTheFinger -- good, layout improvements?
WikiCommeJeu -- French, don't understand it
KarlPopper -- very short, should be extended
RoyalWe -- needs reworking
MetaCard -- don't know, category is missing
EditCategories -- needs reworking

Typical Thread Mode, needs reworking

This is not necessarily indicative of reworking being intractable; it could be that it simply isn't done often enough.

Yes, perhaps one is the reason for the other, but either way, it is not done often enough. Therefore let us be wary of PagesThatNeedReworking. If you cannot write a page such that it can stand on its own, do not write it at all, because there are enough people creating too many shallow pages and so many unnecessary wanted pages. -- AlexSchroeder

Query; would the MeatballWiki community be pleased if there were more reworking of pages? For example, would it be a good thing if some of us were to make an effort to rework one page for every five pages we write, and to do lesser editing one page for every seven pages that we edit/comment upon? We could start by editing the pages that we have just discovered that we don't like.

Reworking is the same as writing. If you do it hastily, you will do it wastily.

Yet we should distinguish between two scenarios. Is reworking possible (but expensive), or intractable? If it is the former, then we simply need to increase the ratio of reworking to writing (bearing in mind the preceding aphorism, and reworking with care).

CategoryWikiConventions CategoryReworking