On Mon, Jan 12, 2026 at 01:17:50PM +0100, Rene Kita wrote:
The issue tracker should only have open items that are confirmed bugs
or bug reports being triaged and feature requests where someone agreed
to work on.
says who?
i for one think that a feature request that has been closed is a clear
"not welcome here, don't bother" message. adding a comment that says the
opposite is just confusing.
i'm fairly sure that a significant portion of people, if not the
majority, feels that way. in fact, my impression is that the only ones
who ever argue differently are some (de-facto ex-)maintainers.
Closing these issues will notify the issuer. This allows complaining
about it. If no one complains, it's not that important.
my experience with actual implementations of such policy is that no-one
who has the rights will bother re-opening such issues - after all, that
would mean acknowledging that something *should* be done, and who wants
_that_ feeling? not closing the issue in the first place has the same
psychological effect, but the inhibitions are higher, because closing
would be an active action.
Also with a lack of manpower saying 'patches welcome' is a valid
option.
sure it's valid as such, but one shouldn't send mixed messages about how
welcome a patch would _really_ be.
as i said in the issue arnt is referring to, the correct way to handle
things is to add labels which indicate the issue type and the
importance. one can go beyond that and have explicit needinfo,
helpwanted, etc. labels. there is no shortage of projects which
demonstrate that.
the act of adding such labels is sufficient feedback for the reporter
that the issue was looked at and what to expect.
We should make it clear to users that the correct way to request a
feature is not to open an issue, but to contact the developers through
the mailing list and discuss it beforehand. Only after it is discussed
and agreed on that it's worth having it as a feature request someone
should open an issue.
ehm.
"let's make it as hard as possible. hopefully this will discourage
people from actually doing it."
also, how exactly does that solve the issue under discussion? under the
protocol arnt now implemented, such issues would still expire.
