Prompted by @pf_moore 's post %-style formatting method - #70 by pf_moore, but I decided to create a new thread since what I want to say is totally off-topic for that one.
I’ve been wondering whether a wiki would be a better replacement for the Ideas category here (and, to a lesser extent, various other categories where “potential changes” get discussed, even including PEPs themselves).
My thinking is thus: ideas are brought up vaguely at first, then gradually turn into proposals that get more and more refined over time, sometimes by the OP, sometimes by others, usually a combination. But all these refinements are dotted about in replies to a forum thread. Once it gets a bit longer, every reader has to separately keep track in their mind of which bits should be kept and which should be dropped, which bits are still outstanding, and filter out all this from the “noise” of unrelated back-and-forth. The agreed points get buried and forgotten about and all that people see is a mess of “endlessly relitigating the same points”, even if no one individually did anything wrong. I truly feel that when a thread turns into a long, heated, and mostly-unproductive discussion, it’s not the result of people lacking in discussion skills, but rather an unavoidable consequence of discussing a non-trivial proposal in this manner.
There is a reason why we see, in some threads, one person go to the great effort of collecting all the information, making their own judgment of what “consensus” looks like, and then post it as one big summary once discussion has cooled down, because in large part that works: nearly everyone will broadly agree with that person’s summary and is tired of arguing, so if anything, that is what will get done. And while we can certainly applaud people for carrying out this task, it shouldn’t be necessary in the first place: I can only imagine it’s a pain to have to do so, not something I’d want to do myself, and many times it won’t happen and then a potentially good idea will be avoided just because it’s so troublesome to touch it.
If we used a wiki instead, then the OP could simply create a page, provide their initial proposal - including the what, the how, and the why - and then over time, any number of separate discussions about individual points to do with the proposal could be held on the talk page, while the article page could be constantly updated with the latest consensus; a “rejected ideas” section can also be gradually established. Any interested parties would be able to see straight away from the article page what each of the accepted points are and their rationale. The article page would serve as a reminder that the majority of the proposal has been figured out, keeping focus on tidying up the loose ends - avoiding what we suffer at present from forum threads, which is the illusion that no one seems to agree about anything. People wouldn’t have to worry so much about their points going totally unnoticed just because they’re a few days late to a hot discussion. The article page can be structured like a PEP from the get-go, so that if a PEP is required, then by the time consensus is reached - bam, you have one already.
If I may provide an opinion on what kind of wiki - I personally prefer MediaWiki, and Miraheze is a thing now if we want to avoid the burden of actually hosting one, though I don’t know how long it takes to get a wiki approved there. (I’ve hosted/maintained MediaWiki instances before, though while I could theoretically volunteer to host one for Python myself, I don’t really feel it would be appropriate for me to do so, as an outsider.)
) Better interfaces are imaginable but I haven’t seen any that actually exist.