For the "Others", they have components defined but there appear to be
soooooo many components, that it's only showing the "big hitters" with a
larger number for the given component category.  I wonder if maybe reducing
the number to a smaller one or have a "top level" component would help
(i.e.Java, Gradle, Javascript, etc.) with a manager for that top level
category and then also break it down further if needed (i.e. Javascript -
Editor, Javascript - Libraries, etc.).

Maybe model after the Netcat tribes like (See
https://netbeans.apache.org/participate/netcat.html).  Maybe that is a new
direction for Netcat groups to monitor specific "tribe" based areas to help
in triage as well.
PrefixTribe

[rcp]

API Support

[cnd]

C/C++

[db]

Databases

[debugger]

Debugger

[editor]

Editor

[groovy]

Groovy

[form]

GUI Builder

[j2ee]

Java EE

[fx]

Java FX

[maven]

Maven

[php]

PHP

[profiler]

Profiler

[unit]

Unit Testing

[versioning]

Version Control

[web]

Web Client


Eric Bresie
[email protected]


On Thu, Nov 11, 2021 at 10:34 AM Eric Bresie <[email protected]> wrote:

>
> Eric Bresie
> [email protected]
>
>
> On Thu, Nov 11, 2021 at 9:56 AM Neil C Smith <[email protected]>
> wrote:
>
>> On Thu, 11 Nov 2021 at 15:44, Michael Bien <[email protected]> wrote:
>> > I do like the github issue tracker since it appears to be simpler while
>> > still doing its job. To keep an issue tracker useful, it does need
>> > maintenance though. E.g many projects automatically close issues if they
>> > are inactive for a certain amount of time. Having 32k open issues on
>> > jira like NB has, is no longer observable.
>> >
>> > Many "critical" issues i commented on asking if its still a problem were
>> > in fact already fixed. This has to be automated, nobody should have to
>> > do that manually. If the original poster cares about the issue and its
>> > closed by mistake, believe me, it will be reopened.
>> >
>> >
>> > i am for github issues, but someone would actually have to write the
>> > templates, and then also try to automate things, otherwise it will be
>> > jira 2.0 in a year or two.
>>
>> Absolutely!  I'm basically saying let's follow what Airflow are doing,
>> and tweak where needed, rather than do this from scratch.  They're an
>> ASF project that seemed to have the problems we have, and seem to have
>> addressed it in a way that we could use as a model - forms,
>> automation, etc. included.
>>
>> I'm for doing this for the same reason I was for ditching LTS - it's a
>> promise to our users we're not keeping.  It feels worse than useless
>> right now - 32k open issues is just a black hole for users to vent
>> their frustrations into.  We really need fewer, better quality issue
>> reports.  So, why not see if we can learn from other ASF projects who
>> have been through this?  It can't be worse! :-)
>>
>>
> Regarding quality reports, see my earlier emails about filters,
> dashboards, reports, etc.
>
> Having filters showing assorted results like open issues.
>
>
> https://issues.apache.org/jira/issues/?jql=project%20%3D%20NetBeans%20AND%20status%20not%20in%20(Closed)%20
>
> I think the JIRA Dashboards provide some of the quality issue reports
> being looked for to provide some guidance..
>
> I unable to to Share my personalized Netbeans dashboad so not sure this
> will be visible or not but with a dashboard like
> https://issues.apache.org/jira/secure/Dashboard.jspa?selectPageId=12333642
>
>
> From that you can drill down to the given linked element (i.e. components,
> status, etc.) where applicable
>
> [image: image.png]
>
> And most importantly, triage for the open issues.
>
> (1) Break the list up into smaller subsets of issues (i.e. by components -
> see dashboard example)
> (2) Although a majority of these open tickets are not set or set to
> "Other" so maybe the first step would be to assign a given ticket to a
> component to start with).
> (3) Alternative labels (i.e. label as "duplicate", "ready to close",
> "reject", "close with no fix", etc) or field could be used to help as well
> in the triage process
> (4) then have "experts for the given components" could monitor these
> and/or provide focus on the components to help the validity of the tickets
> (determine if relevant still, OBE, is duplicate, or is fixed and should be
> closed)
>
> It's important to be cautious about closing items after a given time, just
> because an issue hasn't been worked on forever, doesn't mean there is not
> still an issue and just no one has committed to work on it.
>
> Eric Bresie
>
>

Reply via email to