Hi

I'd like to propose to truncate any trailing s/\(\s*update.*?\)\s*$//i from node-titles in replies.

Reasoning:

Many, me included, flag important updates to their posts by adding an (updated?) to their titles, which is fine.

But this becomes nonsensical when replies are also named (updated) because it's inherited.

I could provide a patch, now that I've identified the add_re_to_title-code responsible for the Re^n: -parts in the titles.

Examples:

?node_id=3989;HIT=%28update (click search)

Opinions???

Cheers Rolf
(addicted to the Perl Programming Language :)
Wikisyntax for the Monastery

Replies are listed 'Best First'.
Re:Proposal: Truncating trailing '(update)' from node-titles in replies
by jdporter (Paladin) on Mar 12, 2021 at 21:58 UTC

    I am sorry that you're having this issue. However -- my opinion, is that title fields are not reliable. That is to say, they're not under enough control that you should rely on them for any given purpose. Ultimately, node titles are essentially just keyword lists. They are what gets searched when one does a simple search. Anything aside from that, any rules we might try to impose (such as "OP") are just conventions and hacks. Even the "threadingness" of titles is not reliable.

    I reckon we are the only monastery ever to have a dungeon staffed with 16,000 zombies.
      > Even the "threadingness" of titles is not reliable.

      Well, I know ... I've seen the code.

      > are just conventions and hacks.

      it has become a tradition to expect from posters to mark important updates in their titles.*

      IMHO will my proposal do more good than harm.

      Granted, it's a hack, but regexing Re: into Re^2: isn't much better.

      This mechanism could be used till we have a proper last-updated-field or updated-count in place, which might honestly never happen.

      my 2 centimes...

      YMMV of course! :)

      Cheers Rolf
      (addicted to the Perl Programming Language :)
      Wikisyntax for the Monastery

      *) and that's not a keyword.

        it has become a tradition to expect from posters to mark important updates in their titles.*

        I don't believe that's true. A tradition for you personally, maybe. Anyway, my whole point was, traditions, conventions, kludges - they're unreliable. You're asking for something which assumes, or imposes, a reliability/consistency which simply isn't there, and won't be without a significant change to our architecture.

Re: Proposal: Truncating trailing '(update)' from node-titles in replies
by haukex (Archbishop) on Mar 13, 2021 at 11:18 UTC
    But this becomes nonsensical when replies are also named (updated) because it's inherited.

    Actually, I use the "(updated)" tag in the title slightly differently: Since I only add it for updates that I consider "significant" to the OP, I think it's useful to see for myself (and hopefully for others) whether someone is replying to my original node or my updated node. So in that sense it would actually be a disadvantage to have the tag removed, since I'd think the responder may not have seen my significant update, so I'd have to /msg them or something. What does happen occasionally is that when I then reply to a reply, I forget to remove the "(updated)" tag from my new node's title. So I'd probably agree with jdporter that titles are just unreliable.

    Just spitballing, and this is probably not a good idea, but a theoretical solution to this would be to do s{ \(updated \K (?:\^(\d+))? (?= \) \s* $ ) }{ "^".(($1//0) + 1) }xei to the titles ("(updated)""(updated^1)""(updated^2)" ...). With that, to mark my node as updated, I can remove the ^\d+ part.

      > whether someone is replying to my original node or my updated node.

      But you know there are timestamps? ;)

      Cheers Rolf
      (addicted to the Perl Programming Language :)
      Wikisyntax for the Monastery

        But you know there are timestamps? ;)

        That would require me to either record the timestamps of my updates manually, or remember when exactly I've made every update, and at least the latter most likely isn't happening ;-)