Can we warn new users when they edit their question to add the words "fixed" or "solved" to tell them that they should add their solution as an answer instead of that the bottom of their question? Sort of like the warning you get when entering "best" as part of the question.

Here is an example: Modifiers (SHIFT + (Random Key)) in Key Binding. (Java) EDIT: SOLVED

  • 3
    Oh man, I can't wait to edit the title of this question once this is implemented. Irony! POW! – Bill the Lizard Dec 5 '12 at 21:09
  • 4
    Do we really need warnings and filters for everything? If you see such a question, simply edit it. And especially "fixed" has several legitimate uses that might not make it an ideal candidate. – Bart Dec 5 '12 at 21:10
  • I guess the point is @bart, that if a question has been "solved" then there should be an answer... just adding solved to the title doesn't better the internets at all. – ben is uǝq backwards Dec 5 '12 at 21:11
  • 3
    @Bart: The whole idea is prevention (and education). Why waste users time with something that can be done or prevented automatically? I thought we were programmers? – user159834 Dec 5 '12 at 21:11
  • 4
    @bart well, providing automated help for common problems is a way of reducing load on human exception handlers. It's just a question of how common this is vs the side effects. Remember "problem" in titles? I am sure you do... – Jeff Atwood Dec 5 '12 at 21:12
  • 3
    @JeffAtwood Oh yes, glorious word filters. Those are a great addition. ... I guess I'm just not in favour of too many such filter. "Don't add fixed", "Don't add solved", "You said thanks, did you mean to accept?", "You said terqzziuz. Did you sneeze? Want me to call you a doctor?". Are we really trying to solve a problem here that is currently too big to handle? – Bart Dec 5 '12 at 21:15
  • 3
    @Bart: "Do we really need warnings and filters for everything?" No, of course not. However, is this issue something we want them for? That's what's on the table. I'm leaning towards "maybe". – user159834 Dec 5 '12 at 21:16
  • 1
    @WesleyMurch Sure. And I guess you know my answer to that one. ;) – Bart Dec 5 '12 at 21:17
  • 1
    EDIT: S0LVED.... – user102937 Dec 5 '12 at 21:28
  • @RobertHarvey The more common workaround might be to just put the answer after the question and not change the title, or worse yet just not edit in the answer they found because "it's too much of a bother". – Servy Dec 5 '12 at 21:30

The comments make it clear that I didn't want to advocate that this is a huge problem. I see it often enough as a moderator, but it's not my biggest editing complaint.

If this is determined to be a big enough problem, rather than a hard stop, perhaps putting the detected problem edits into a queue for review would be a better solution to this "problem".

A full-on machine filter seems a blunt tool for this ailment and false positives would be damaging to the innocent.

I would bet most of the people prone to making this editorial mistake

  • will not see or care if context sensitive help warnings
  • won't listen if they do see and understand the warning
  • will work around any filters with unicode or worse terms than EDIT!!!

Would asking the reviewers to approve self edits when the first word of a new text added to a paragraph was "edit" or those strings were seen in a title change once edited? Anyone doing human reviewing would be able to clean this sort of thing up / revert the change and it wouldn't need to be an all or nothing, but more like something that could be flagged for low quality if people really are being standard about how they edit a question once it becomes "solved"

I'll leave the determination of existence of an actual problem to someone closer to the data.

  • 1
    ^^---This. Love the idea of queueing them, instead of summarily acting on them with a machine. – user102937 Dec 5 '12 at 21:49
  • @robert yes, because the current queues are so well serviced, why not add a few more things to queue up in line for? What could it hurt? Except everything. – Jeff Atwood Dec 6 '12 at 0:34
  • 1
    @JeffAtwood: Right. Let's automate it instead. We already know how well that solves pr0blems. – user102937 Dec 6 '12 at 0:36
  • 1
    @robert I'm just saying, both approaches have clear downsides. Queues create work. I'd rather have a few vocal, er, "special" users complaining about edge cases for millions of mythical questions about "the halting problem" versus the actual endless "problem with SQL" questions we get in reality. Depends on the data. – Jeff Atwood Dec 6 '12 at 0:41
  • 1
    @JeffAtwood: That's why I'm not a big fan of involving machines (whether by queue or by dialog) in what amounts to an educational problem for people who don't read anyway. – user102937 Dec 6 '12 at 0:43
  • @robert machines are very effective at helping people who don't read. That said I'm not sure the data here (e.g. is there trash all over the city?) is that compelling. – Jeff Atwood Dec 6 '12 at 0:59
  • I'll just say the review queues as implemented are a big step up in community moderation for Ask Different. I don't see any negative from another feed or more data, but I'm not footing the bill for the developer, so the value of it is hard to tell from where I sit. – bmike Dec 6 '12 at 1:04
  • @JeffAtwood: That's what I meant; this use case is not worth the effort, IMO. – user102937 Dec 6 '12 at 1:04
  • @bike well, imagine a thousand flags per day. Now add up the time to read, process, and handle them. It's a different problem on Stack Overflow. – Jeff Atwood Dec 6 '12 at 1:14
  • @JeffAtwood So I've heard - the review queue alone there boggles my mind. I'd think the payback for good tools there is the real motivator for developing tasks that distribute moderation to the base as well as minimize a category of flagged behavior. – bmike Dec 6 '12 at 1:23
  • Please. No more queues. Well, maybe a few more, cause we could do better at getting some eyes on things like deleted posts, but... it seems like since the new /review was rolled out, a solution to just about any problem has been "let's add another queue!". That's unsustainable. When anyone can edit or submit an edit to remove "solved" from the question title and post the answer as an answer, it seems like adding an extra step, possibly error-prone detection, and a special place where people have to go to find these problems is just creating busywork. – Adam Lear Dec 6 '12 at 2:13
  • 1
    @AnnaLear I appreciate what you are saying about not adding a queue, but this could (only if it's needed) be a second feed into the existing edit queue, no? – bmike Dec 6 '12 at 2:24
  • 1
    @bmike Yeah, let's assume for the sake of argument that this is a widespread issue. Using the suggested edits queue would solve a part of the problem, but the other parts - notifying the editor and reliably identifying which edits should land in that queue - would still be an issue. Everything else aside, at the end of the day I'd rather have a question that has "solved" in the title and the answer in the body of the question than a question that's left unanswered because the asker didn't notice that their edit got rejected or didn't understand why. :) – Adam Lear Dec 6 '12 at 2:31
  • @AnnaLear I think you make a great point. The documentation, the "mind share" of people to adapt to any time you change things, and confusion are always easy to underestimate whenever the "fix this problem" brigade is out with a great idea to solve a few isolated cases. – bmike Dec 6 '12 at 2:34
  • @JeffAtwood Of course you can refer to those vocal non-believers in word-filters as "special", but I assume you're aware of this as well, right? My point being that I think users are more likely to circumvent such filters than truly address the quality problems. If that makes me/us "special", so be it. ;) – Bart Dec 6 '12 at 19:43

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .