I'm researching implementing UUIDs in an app for the first time, and the very first thing I want to know is 'what is the probability of collisions'.

I think it's an incredibly important software engineering question, yet it seems to be closed as 'off topic'.

Counter argument: "is't it obvious that if the whole software world is willing to use the standard, then you can assume you're going to be safe too? Not necessarily, some methods for migrating existing (incremental) identifiers to UUID's merge existing incremental IDs with the UUID, so understanding how probable collisions are in the UUID helps one to understand how probable they are in an identifier that is part UUID and part incremental ID.

That's one example, but there are obvious reasons to do with wanting to understand the software you're building in the general sense too.

  • 1
    Does this answer your question? Is a question about research material on-topic? – gnat Nov 18 '20 at 9:58
  • 2
    @gnat are we really saying that "Has anybody done any real research on the probability of UUID collisions" is more than a slightly sloppy way of asking "what's the probability of a UUID collision?". To me they're the same and the question should be lightly edited and reopened. It's of clear interest to the community and directly related to software engineering – stevec Nov 18 '20 at 10:06
  • Why do you want to reopen the question? Do you want to write a new answer to this old question? – Doc Brown Nov 19 '20 at 6:45
  • 1
    @DocBrown I have nothing to add to it at this point in time (I am nowhere near knowledgable enough on the matter). I want it reopened for the same reason I would like every one of the millions of good questions on SE open; so they can be contributed to, improved, updated as necessary to name 3 reasons. False positives occasionally happen when looking for bad questions, but sometimes they happen because a good question simply looks bad, which I believe is the case here. – stevec Nov 19 '20 at 9:25
  • 1
    @DocBrown just noticed the question was once tweeted by the Software Engineering twitter account, and was considered on topic for many years, until 2016. – stevec Nov 19 '20 at 9:34
  • @maple_shaft please consider reopening. – stevec Nov 19 '20 at 9:36

The scope of this site changed over the last years. When the question in stake was asked in 2012, almost any conceptual question about programming was allowed, and questions for third party resources like books, tools, external links, or research papers were on-topic.

Since then, the community here decided to narrow the scope of the site and not to allow such questions any more. I don't remember the exact point in time when that happened, but it must have been before 2016. This meta post gives some information about the reasoning behind this decision. Also at the end of 2016, the site was renamed from "Programmers.SE" to "Softwareengineering.SE" to give newcomers a better first impression about the scope (of course, the name cannot tell anyone that 3rd party resource requests are off-topic, for this, askers have to look into our help center).

Nevertheless, we still keep many of the old questions here because we see value in them. Often such questions are protected by a historical lock, so noone will delete them. But it is pretty unlikely a moderator will reopen them - they are off-topic today, and could be used as a justification for asking other off-topic questions.

  • 1
    Thanks for replying. This addresses why it was once fine and at some point was deemed not okay. But it doesn't do anything to address my primary point, which is that the question is absolutely essential to software development, yet it has been deemed off topic, presumably because it simply asks (sloppily) "Has anybody done any real research..?", I believe once that is deleted, the question is incredibly relevant, it's madness to consider it off topic here! – stevec Nov 19 '20 at 10:24
  • 1
    @stevec: I think it is an interesting meta question if a question like the one in stake can be modified somehow to become on-topic and reopened again. Maybe worth a new meta-question on its own (though I guess the answer will be clear: you need to convince a mod). However, there is also to consider that a question which was historically locked should be edited with extreme care, since there is a high chance to invalidate some of the existing answers. Best option is probably to ask a new question about the aspects which was left unclear by the old question and it's current answers. – Doc Brown Nov 19 '20 at 16:12

You must log in to answer this question.

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