Carcharoth carcharothwp@googlemail.com wrote:
Indeed. There is a bot that can help index talk page archives. I'll give details below.
Well, while I see the value in raising indexing as a process, I still have to point out that we aren't talking about talk pages and organizing them topically for later ease of reference (ie. WP:OBT) , but the "refactoring" of actual "vote" discussions wherein we have to make collective qualitative discernments about the merit of individual arguments.
In that context we of course realize that IAR is not an actual solution, and we are now starting to talk about process methods for dealing with discussions in a meta way. At this point it requires mentioning that what we are really talking about is in part a rating system for comments integrated into talk pages, similar to a Scoop or Slash system. I'm not certain this is a current or even planned functionality in Liquid Threads, but in any case it seems that the LT project (or some better-thought out derivation) should be regarded as a high-priority ("usability") project that we need to put more coders to work on. AIUI, keeping things still "wiki" - such that discussions still have basic wiki re-factoring capability seems (typically enough) to be both a high principle, and an obstruction.
Aside from the rating component, we should consider comment length as a factor in how sub-comments are nested - some comments are just short votes of support for an above argument. Nesting those beneath a main argument seems necessary. In the wild, typically see four basic dimensions within a discussion: 1) long posts with lots of substance 2) short posts with lots of substance 3) long posts with little substance 4) short posts with little substance
Simplistic, true, and its often hard to atomize long posts - substantive or not (which is why I like line-by-line replies). But ranking helps get rid of the bottom two kinds of posts. Proper nesting can deal with how the first two interrelate. After that, its possible to use the tool improperly, where ranking *can indicate which of the substantive arguments are dominant, but reliance on this can raise the voting fallacy issue all over again. But what of it? At least 3 and 4 are disposed of, and 1 and 2 are put in place.
-Steven