There are a number of ListMode
pages on Wiki whose main content is a purported list of Great Foobars, for some value of Foobar
These pages all seem to go through the same evolution. Initially the page contains one person's ideas about the Great Foobars; then a few other people add theirs; and so the list grows and grows, with more and more marginal or downright ridiculous entries -- and nothing is ever removed
. [Because people don't DisagreeByDeleting
Some names were just removed from WorldGeniuses, since their presence had been questioned. Ideally, the questioners would have removed them themselves. If enough people do this then, hopefully, the list will converge to some fairly steady state in which no-one much wants to delete any of the names.
Another observable behaviour is on some lists, once having grown overly long, then get refactored into shorter sublists for various qualities of Foobar. Examples include MoviesToConsider and WorldGeniuses. One potential sublist could be "membership in question", which is one step away from having a name deleted. It's easier to shuffle names into various sublists than it is to write useful but ConciseComments?, and sublists provide more meaning than a simple accumulated vote. So, if you see a GreatFoobarList? which has grown too long, try sub-sectioning it.
There are at least two reasons why, in general, nothing is ever removed from a Great Foobar list:
(See also ObscureGreatFoobars)
- DeletingContent? provided by others is widely regarded as rude, especially when a lot of the point of the page is letting people express their opinions.
- Some of the candidates for deletion are relatively obscure (which is one thing that makes them good candidates for deletion), and no one wants to take the risk of deleting an entry that really does belong in any list of the World's Greatest Ever Foobars just because they happened not to have heard of it/him/her.
- "DisagreeByDeleting" sets it out ... deleted material cannot benefit even the few.
These are, on the whole, good reasons. But the consequence is that these pages rapidly become almost completely pointless. Not completely
pointless; there's still some content to the assertion that X is regarded by at least one WikiContributor?
as a Great Foobar. But it would be nice to have some way of telling which entries really represent the Mind Of Wiki, and which just indicate that one or two people have strong opinions and no one else has seen fit to disagree.
Of course, these lists are sometimes accompanied by discussions of the entries. But, again, I think people are often diffident about writing a paragraph with no point other than to disagree with an unknown other person's assertion that X is a Great Foobar. And, after all, one of the points of lists is that they're meant to be concise.
Which I have failed to be so far. Let me get to the point. I propose the following simple convention for the annotation of WikiGreatFoobarLists
, which provides a way for a reader who disagrees (or agrees) with an entry to express that fact briefly, inoffensively, and in the right place. The easiest way to illustrate it is with an example. Here's a silly list of Great People Whose Names Begin With B:
* [AAB] LudwigVanBeethoven
* [A] JohannSebastianBach
* [ABBCC] GradyBooch
* [AAB?] KentBeck
* [BCC??] EdmundBlackadder? ()
* [AAC??] StefanBanach?
* [ACCC] WilliamShakespeare
The typographical conventions are presumably clear at this point. The intended interpretation of the annotations is A for "yes, definitely belongs in the list", B for "hmmm, maybe", C for "no, shouldn't be there", and ? for "who?" or "what?". Annotations should be kept in order: "A"s first, then [...], then "?"s.
Obviously, adding more than one annotation to a single list entry is rude. So is deleting or changing other people's annotations.
Great suggestion (from the man who launched foobar= RespectedSoftwareExperts and didn't know what to do next). But I think we need more than three values plus "?". How about a single digit, with 0 representing no perceived relevance for this list, 9 meaning a must have, the rest meaning something in-between and the "?" retained. Same rules on VoteEarlyVoteOften.
s. Y = yes, N = no, ? = who. A maybe is a non-vote anyway, so we can discount it. Actually, you can condense the information using numerals ala [Y:3 N:2 ?:4] While you could combine the yeses and nos, it helps to know what the weight of the vote is (i.e. how many people voted in total), so you might as well separate them. If you're going to eat up space, you might just as well go [Yes:3 No:0 Who?:493]. Besides all this knocking about, this is a very good idea. I like it muchly much.
I'd say that there are cases for Y/N/? and for 9/8/7/6/5/4/3/2/1/0/?. If foobar=composers for example, Bach for me would be a 9, Mendelsohn a 6, Stockhausen a 2 (if he's lucky). A simple y/n doesn't seem adequate for all domains in other words.
On condensing: once voting becomes popular digits should be condensed simply to sum/"?"s/totalvotes and y/n to something like above. So
* [968?758] LudwigVanBeethoven
* [?59699?] JohannSebastianBach
* [??45??0] CharlesIves?
would be "condensed" to
* 43/1/7  LudwigVanBeethoven
* 38/2/7  JohannSebastianBach
* 9/4/7  CharlesIves?
Too complex. How about Delete this page? (Y/N) [YYYYYYYYY | NN] ... which would result in *what?* the item remaining; voted a keeper.
But here's the weirdness in my view that leads to the empty : I believe that more people will vote if it means simply adding the letter or digit of their choice rather than adding to a cumulative number. It's that many less button presses (hence the lower case y/n) and no arithmetic. So a HumbleRefactorer
should come along and total up from time to time and use his/her own algorithm to sort the list by the result. (Do you ignore deduct "?" before averaging or not? Does Bach beat Beethoven above? I say LetThePeopleDecide
in the normal WikiWay
, even on the rules of the election as we go along. But voting itself must be consistent to work at all.) --RichardDrake
You shouldn't underestimate the people if you're willing to LetThePeopleDecide
. I'm sure they can handle simple arithmetic. And I don't understand why you put the numbers outside the brackets. Besides, the slashes don't have any meaning: it looks like a date to me. [Y:4 N:54 ?:3] is much clearer; you essentially want to embed the semantics of the process into its visible output so people can immediately grok its function and rules. If you want to weight your vote, VoteEarlyVoteOften
(i.e. add 7 or 10493 or -4). That's how WikiVoting
The method sounds good, and lets anyone that wishes to do the arithmetics to do it, and others can just write their numbers. But there is a problem: the meaning of a "0" rating is that this item should be excluded from the list. It should be a negative number in the total count, and not be counted as a zero. -- AmirLivne
I'm not sure about the voting, and usually go by the comments.
I'm prone to refactor these pages by moving the "good" ones to the top and the "bad" ones to the bottom, based on what seems to be general agreement among contributors.
Then, I'd put lines between the different levels, eventually leading to a bottom category of marginal or bad entries that could be deleted.
I agree about ContinualReordering?
. But how can even the most HumbleRefactorer
know what the overall consensus is to govern reordering if it's down to verbose verbal comments that don't encourage every reader to vote on every item of the list (if they want to). We do need a convention for this. I'd go further and say that establishment of such a convention and the creative use thereof (if we can avoid abuse) could help towards the SecondWikiRenaissance
. But let's not forget TipsForBeginners
for an older attempt at voting that doesn't seemed to have lasted long (or the WikiVillage?
is even smaller than I thought!). --RichardDrake
If people just moved the ones they most agreed with up a slot or two, and/or lowered the ones they least agreed with, we would probably get a pretty well-ordered list fairly soon.
Let's have a vote on the probability of that.
Out of 10: 
It seems to be against the PrimalFearOfRefactoring? among many WikiZens (including me). It just seems too much like deleting other's opinions from the page. -- AmirLivne
tallies votes already cast.
Wouldn't it be better if everybody's favorites lists or top ten lists were separate, rather than all squished together? If I find a single author's top ten list that has six of my favorite programming books in it, I know that the other four are worth looking into. (See SteveMcConnell
's real home page, for instance). --StanSilver
I consider the GroundBreakingLanguages
list a moderate success in refactoring a ballooning list. It follows the following policy:
- Every entry needs to be motivated
- There is a RunnerUpList? containing InsufficientlyMotivatedItems?
- InsufficientlyMotivatedItems? can be moved to the RunnerUpList?. This doesn't mean the item doesn't belong in the list: it just means that it is not yet clear that it does.
Seems like the discussion has settled down to CategoryVoting
What's the point of having these lists in the first place? On an individual level I can see the utility. If, say, AlistairCockburn
has a list of super-useful books on his page, and I think Alistair's a brilliant guy, then maybe I'll follow his list. But when you start aggregating everybody's opinion into some complex ranking scheme it seems not so useful. Even if we do figure out how to sum up everybody's book opinions into the S
uperUltimateWikiBooksListOfAllTime, so what? Sounds like just a popularity contest to me. Maybe there are people who will go see a movie just because it won an Academy Award. But I don't know any of them.
Public Wikis - http://sknkwrks.dyndns.org:1957/writewiki/wiki.pl?PublicWikis
Again, lots and lots of discussion with an underlying assumption:
What is Foobar?
Foobar = FUBAR (Fouled Up Beyond All Recognition) The etymology of foobar in Computer Science usage is more complicated than that. See the JargonFile.
How does Wiki foster recognition / rethinking?