github: reword bug or regression

[Aqa-Ib]

I have been watching and a lot of people still don't mark a clear regression. Rewording would fix this, I believe.

[MightyPlaza]

might also be a good idea to ask them to bisect and provide the wiki link https://wiki.hyprland.org/Crashes-and-Bugs/#bisecting-an-issue

[Aqa-Ib]

Let me know if it can be worded better or clearer (english is not my mother tongue).

Also, do you think that we should give "I don't know" as an option?

[fufexan]

LGTM

[fufexan]

Also, do you think that we should give "I don't know" as an option?

Probably not, as we would still tell them to bisect. It's either yes/no.

Also, we should tell users to also search closed issues, not only open ones.

[MahouShoujoMivutilde]

Maybe something like

• Bug - This is an unexpected behavior (something is broken).
• Regression - Like bug, but you also know that it used to work as expected and now doesn't (provide bisect).

?

Because "absent in a previous version" is IMO not as clear - you could read it and be like "of course it was absent, this feature didn't exist".

[Aqa-Ib]

I understand your concern about a new feature. However, in my opinion, your definition of regression is not simple enough. Feedback is welcome.

[fufexan]

@Aqa-Ib what about "Regression - something no longer works in newer versions"?

[Aqa-Ib]

Done. One thing that is not clear: if we force the users to check for a regression, one can wonder: "how many versions do I have to roll back to check for it?" If we are omitting the "I don't know" option, in my opinion we should clarify this. Maybe 1 or 2 major releases ?

[MightyPlaza]

Done. One thing that is not clear: if we force the users to check for a regression, one can wonder: "how many versions do I have to roll back to check for it?" If we are omitting the "I don't know" option, maybe we should clarify this.

i don't think this matters much
the point is that they write that it's a regression if they know it was working before and now isn't
not go looking for a working commit (this is also a good idea when debugging issues, but not necessary imo)

[Aqa-Ib]

If that is what we want, then it would make sense for the "I don't know" option to be included. Thoughts?

[fufexan]

Done. One thing that is not clear: if we force the users to check for a regression, one can wonder: "how many versions do I have to roll back to check for it?" If we are omitting the "I don't know" option, in my opinion we should clarify this. Maybe 1 or 2 major releases ?

If the user just updated, they should know which version they were on before, and start with that. If they don't, then 1-2 major versions sounds alright.

[Aqa-Ib]

@vaxerski please, as you are the one that deal with every bug report, when you have a moment do you mind sharing your thoughts on this?

[vaxerski]

rest is ok, I didnt review cuz yall were having a heated discussion

[Aqa-Ib]

@fufexan @MightyPlaza @Agent00Ming would you make a change to improve bug reports ?

[Agent00Ming]

Maybe the command paste thing is a bit obscure but I don't know if we can wrap the user input in a plain text box and treat it with markdown afterwards. Otherwise, it looks as good as it's going to get considering the difficulty that some experience reading a sentence longer than 5 words.

[fufexan]

@fufexan @MightyPlaza @Agent00Ming would you make a change to improve bug reports ?

Unsure what we could improve further here.

[Aqa-Ib]

I found a way to force the user to select yes or no.
At the moment I am exploring simplifying the command paste textbox.

[Aqa-Ib]

I wasn't able to simplify the command paste textbox. I believe that it is not possible at the moment.

[Aqa-Ib]

This is ready.

[fufexan]

LGTM, thanks!

[Aqa-Ib]

Thank you as always. Wish you all a wonderful day!