Ghyll:Lexicon discussion
If you've any questions or suggestions about the wiki and its syntax, the Lexicon rules, Ghyll continuity errors, letting us know you're gonna miss a turn, etc., use this page to wax poetic. Be sure to sign your name (using either the second - from - the - right toolbar icon, or typing two hyphens and four tildes), which also includes the timestamp. Similarly, we'll probably end up using horizontal rules to seperate "threads" that develop. We'll cross that bridge when we get there. --Morbus Iff 11:32, 20 Aug 2004 (EDT)
The random page algorithm is clearly crap. Even though we only have like five pages now, I still got Agony Uncle roughly ten times in a row. @@ XXX FIXME XXX etc. --Sean B. Palmer 12:59, 20 Aug 2004 (EDT)
I disagree. A truly random algorithm could display the same thing 100 times in a row - that's the uniqueness of something being "random". Any algorithm that remembers previously displayed pages creates an expectation: that the next page you see will be one you've not seen before. When you can expect a certain result, that's decidedly not random. Regardless, I'd say this is very faw down on the list of FIXME's, solely because a) I don't really think it's broken and b) the longer the wiki is up, the less it'll be an issue. Thus, entropy rules. --Morbus Iff 13:45, 20 Aug 2004 (EDT)
Bah, talk about meta meta. What would a "Lexicon discussion" discussion be about? /me grumbles. --Morbus Iff 15:36, 27 Aug 2004 (EDT)
So I added a phantom category. I did so because I thought it'd be nicer that way to be able to directly see alphabetically which pages need filling in, though I must admit it's not much of an improvement over using the old use of Special:Wantedpages instead. On the other hand, I don't think there are any disadvantages: for example, it's still possible to find out what pages link to a phantom by using the what links here link in the page's toolbox; on the other hand, using a category and phantom boilerplate for each new entry tells the entry's creator what the initial citation to that entry is, as opposed to just links. For example, (Folktown Records) has two inbound links already, but which was the first one? In fact, it was (Agony uncle) and not (Quezlarian numerals), but its 'what links here' page, and hence the old Wantedpages approach, won't reveal that. The new boilerplate texts, on the other hand, do. --Sean B. Palmer 20:16, 27 Aug 2004 (EDT)
The alphabetical is nice, so I've changed the template over to it. I'm worried, however, that we're adding another "oops, I forgot" step for maintenance. Not only will people have to define their phantom, they'll also have to remove it from the Category. Similarly, someone who is creating a new entry now has the onus of "not yet dib"bing their created phantoms, which can easily be forgotten. Wantedpages, at least, handle all this stuff automatically. As for the "what links here" page, it looks like the order is, indeed, when the link was created - reference, for instance, the "what links here" page on Folktown Records. The first link was Quez, then my Agony, then this page. So I'm not sure why you're questioning the "which was the first one?" --Morbus Iff 20:33, 27 Aug 2004 (EDT)
I'm also worried people will forget the "what links here", thinking that the manually "cited from..." entry is the ONLY thing to worry about. Although I like the alphabetical page, I'm a little worried that this is gonna cause more work in the long run - at least from Wantedpages, there's an immediate and visual "closeness" to the red links and the (# references). I wonder if it'd just be easier for me to custom code Wantedpages into an alpha, not popularity. --Morbus Iff 20:39, 27 Aug 2004 (EDT)
AlllLLrighty. In a fit of "undo all your hard work", I've undid all your hard work. I modified MediaWiki's includes/SpecialWantedpages.php to sort the Wantedpages screen alphabetically, regardless of the number of links to a post (in a game of Lexicon, popularity is secondary to turn order, as you've demonstrated). As for the "when the link was created", I finally understood what you meant. Yes, I think that can be a potential downside, but gamewise, the scholar has to create an entry based on ALL incoming links to the phantom, not just the first, or "canonical." So, yes, knowing what is first could be important, but I don't think it's vitally important to gameplay. Like you mentioned, the phantom category wasn't "much of an improvement", but the breaking of Wantedpages, I think, was its biggest disadvantage - I'd rather break "order of citation" as opposed to "default page in the wiki that I can't delete." --Morbus Iff 21:06, 27 Aug 2004 (EDT)
Well, the first/canonical link is also very important if we're disallowing forward self-citations. For example, (Folktown Records) is a term that I coined, but I'm not sure whether I can dib it or not since I'm not sure where it was cited first. Looking carefully reveals that it was you that first cited it, so apparently I can, but it's confusing. Of course, in the process of researching all of the inbound links, one should also discover which is the original citation--but if you're choosing between lots of phantoms for a particular letter, you want to be able to know which ones are yours and which ones aren't, without a lot of research for each of the terms. I think that's something that needs to be answered before I can fully except your undoing of all my hard work, as you put it (grin), but I otherwise accept that it was a bit of a stretch. --Sean B. Palmer 21:21, 27 Aug 2004 (EDT)
You and your damn forward self-citations! The only time a "coined term" is "claimed" is when you use it in your per-turn citations. Since you coined it, but didn't cite it in your per-turn on Quez, it's not your phantom. Since I cited it as part of my phantoms on (Agony uncle), I can't ever define it, but you can, even though you coined it (but not cited it). AaahHh, the joys of terminology. Ultimately, if you know you "coined" a term, but you've never, in your entries, cited it, you can define it's entry when the relevant turn is played. --Morbus Iff 21:28, 27 Aug 2004 (EDT)
