[prev in list] [next in list] [prev in thread] [next in thread] 

List:       openjdk-openjfx-dev
Subject:    Re: Not really a nice comment but a real issue?
From:       Michael_Strauß <michaelstrau2 () gmail ! com>
Date:       2021-03-28 1:14:29
Message-ID: CAJEpuXQZuyr+bEbi3HLs4mMsCutUpH0jWe1dmor-e020g+Zrgw () mail ! gmail ! com
[Download RAW message or body]

Part of the reason why it's hard to reproduce a bug using a one-shot
report is that there's no way to have a low-effort open discussion and
no way to ask questions that would help clear things up. Not only is
having open discussions of issues beneficial for quickly zeroing in on
a workable solution, it also helps with getting a better understanding
on where common pain points are for users. Even if a user wrongly
identifies or poorly describes an issue, you'll often get a better
understanding of an underlying issue just by allowing discussion.

Am Fr., 26. M=C3=A4rz 2021 um 22:05 Uhr schrieb Nir Lisker <nlisker@gmail.c=
om>:
>
> There's no question that the JBS is the way to go for handling existing
> reports.
>
> As for incoming reports, I dislike the GitHub issues as they are just a
> text dump, and many reporters are not careful about giving you versions
> (Java/JavaFX/OS...) or working environment details. You can help GitHub
> with templates like Gradle does, but I find it still a bit shaky and
> overall produces much less quality reports.
> There's also the issue of transferring them from whatever system we use t=
o
> JBS, how to do it and who exactly will do it? Will the triage team at
> Oracle that work on the webform work here too? It takes a lot of time to
> try to reproduce a bug using (sometimes cryptic) submitter instructions.
>
> Let's enumerate the webform issues:
> 1. Bugs don't immediately reappear on JBS since they are triaged internal=
ly.
> 2. Can't add images/attachments.
> 3. No submitter name for those who want the credit.
> 4. ???
>
> Suppose we create our own webform:
> 1. We technically have credentials to forward them directly into JBS
> because we have Authors (needless to say it should not be dependent on a
> specific person, but for now...). Will we lose Oracle's triage team in th=
is
> case or can we tell them to have a look at these bugs (we can adda label =
to
> mark them)?
> 2. We can allow attachments and transfer them directly to Jira attachment=
s.
> 3. We can't credit the submitter because the Reporter field in Jira is no=
t
> editable (at least at my permissions level) and I'm getting the impressio=
n
> the people who are in charge of this will not make an exception for us, b=
ut
> we can add a text line with "submitted by" as a workaround.
>
> Is it worth it? Can we do better?
>
> N.B. As John Neffenger said about issues still coming to the old repo,
> maybe we should close it so they won't be misled.
>
[prev in list] [next in list] [prev in thread] [next in thread]