PSF News: CPython is switching to GitHub issues

[user4678]

On Wednesday, 29 May 2019 at 18:47:52 UTC, Robert burner Schadek 
wrote:
> On Wednesday, 29 May 2019 at 14:49:37 UTC, Seb wrote:
>> http://pyfound.blogspot.com/2019/05/mariatta-wijaya-lets-use-github-issues.html
>> https://www.python.org/dev/peps/pep-0581
>>
>> Most of the arguments, ideas and solutions apply to D and its 
>> Bugzilla too.
>
> https://github.com/dlang/projects/issues/43

This bring several problems, leading to the conclusion that it is 
not a good idea.

1. ownership of migrated issues, accounts

How can you guarantee that the creator of a Bugzilla issue will 
have admin rights (e.g at least the ability to close) on the 
migrated issue ?

Do you realize that not everybody has a GH account ?
So that's it ? people would have to follow or stop reporting ?

2. Maintenance.

Everybody can volunteer to maintain issues:

- close invalid.
- close issues fixed but not well referenced.
- change the labels, e.g from "enhancement" to "normal".
- and more.

think for example to the "advant of bugfixes" initiative.

moving to GH means that only members of the organization will be 
able to do this maintenance. regular registered users will be 
more limited than on Bugzilla. unregistered users will be fucked.

3. performance

GH is slow, is subject to be down because it is constantly 
attacked while D Bugzilla is dedicated to a single task, is fast 
and quite confidential (i.e not targeted).

4. Categories

There is no categories in GH issue system, only labels.
Bugzilla have keywords, see also, importance, etc.

5. Other

BTW you already have an example of what it would be like: DUB 
issues.