8

I just noticed that one of my questions was tagged . I think that's a meta-tag, not a proper Meta tag:

Avoid meta-tags

Do not use meta-tags in questions. Here are some tips to help you determine whether a tag is a meta-tag:

  • If the tag can’t work as the only tag on a question, it’s probably a meta-tag. Every tag you use should be able to work, more or less, as the only tag on a question. Meta-tags, like [beginner], [subjective], and [best-practices], are not helpful by themselves – they do not communicate anything about the content of the question.

so I've rolled back the edit (conveniently making room for a red status tag).

There have been a lot of regressed bugs in the history of Stack Exchange; I can't imagine someone explicitly searching for those, but perhaps he/she will be served by just searching for the word regression or one of its synonyms. Such questions will usually contain a link to the original ; it might even be possible to write a SEDE query finding those.

Since it's only five questions remaining, normally I'd burninate on sight. But a couple of users were involved (it was an approved suggested edit), so I thought I'd let others chime in first.

  • cough cough go ahead. – user474678 Sep 10 '19 at 21:08
1

Let's down the . I am not much active user on stackexchange but silent watcher .

SO tag for

Regression is a common applied statistical technique and a cornerstone of machine learning. Various algorithms and software packages can be used to fit and use regression models.

In other words, regression is a statistical measure that attempts to determine the strength of the relationship between one dependent variable (usually denoted by Y) and a series of other changing variables (known as independent variables). Typically the dependent variables are modeled with probability distributions whose parameters are assumed to vary (deterministically) with the independent variables.

Don't think it's related to SE. Just burn down.

| improve this answer | |
  • 2
    This argument doesn't hold that much water. SO's definition is different, but in the world of bug tracking, "regression" refers to a bug that has cropped up again after initially being fixed. It can easily be defined in the MSE tag wiki to mean that. – Sonic the Masked Werehog Aug 15 '19 at 6:42
0

I think the tag should stay. (For transparency, I was one of the users who approved the suggested edit you mention.)

On this site, I've noticed that a lot of users tend to vote to close questions that are actually regressions of prior completed bug reports, as duplicates of said prior bug reports. Time and time again, I have to explain to those voters that the SE team has explicitly stated that new bug reports should be filed in case of a regression, rather than removing the tag from the older one. Having this tag would make it easier to explain, "hey, this is tagged as a regression".

I think it's a useful category for searching for bugs that crop up again, for users looking to analyze how often prior bugs crop up again. Not everyone uses the word "regression" in their post, but by using tag synonyms we can allow them to be searched for. Merely searching for bug reports that contain links to prior bug reports would result in a lot of false positives, in my opinion. It's also useful if one wants to search for the bug report and wants to find the new report without sifting through other search results - they can search for keywords [regression].

Shree's answer makes an argument based on Stack Overflow's definition of the word "regression" - in bug tracking, the word "regression" is also used to refer to a bug that has arisen again after previously being fixed, so their argument doesn't hold that much water.

Finally, while this tag could be considered a meta tag, from what I've seen, this policy is more relaxed here on MSE than on main Q&A sites. For instance, the tag is (excuse the pun) essentially a meta tag, and so are the status tags.

In summary, I think this tag should remain to allow for easy categorization, searching, and analysis, and to better serve as a pointer to other users the fact that it's a regression and should not be closed as a duplicate.

| improve this answer | |
  • 1
    On this site, I've noticed that a lot of users tend to vote to close questions that are actually regressions of prior completed bug reports, as duplicates of said prior bug reports. Yes, I agree that is a problem (as is voting to close [status-completed] bugs as 'can no longer be reproduced'). I'm in the process of writing another question about that. Not my downvote, but I feel like using a tag for regressed bugs isn't really worth it. – Glorfindel Aug 15 '19 at 6:52
  • Also, the transparency is appreciated! I've deliberately not linked to the edit since it didn't really help to clarify the question. – Glorfindel Aug 15 '19 at 6:54
  • I'm not sold on the argument that adding a tag would make wrong closure (in your opinion) less frequently. I'm also not sold on the assumption it woud make searching these reports any easier, for potential askers or answerers or even SE staff. If a report is a regression it should say in the post and explain how. That can be as easy as linking to the former report. That has the added benefit the post gets linked. That is value in one go, without the need to maintain a meta tag and educate the close voters about its purpose. This explains my down vote. – rene Aug 15 '19 at 7:09
  • 1
    the tag meta is not a meta tag. It is useful to make a distinction between policies (for example about meta tags) that only apply to this very site. Even used by itself on a question it adds a fair categorization. – rene Aug 15 '19 at 7:11
  • I've noticed that a lot of users tend to vote to close questions that are actually regressions of prior completed bug reports. Please qualify a lot. Is that 6 to 8? – rene Aug 15 '19 at 7:12
  • @rene To be clear here, this tag won't replace adding a link to the question and explaining that it's a regression in the body, but be an addition to it. – Sonic the Masked Werehog Aug 15 '19 at 7:14
  • 1
    @SonictheAnonymousWizHog sure, still no need for the tag. It is noise. – rene Aug 15 '19 at 7:15

You must log in to answer this question.

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