On 5/4/21 8:54 AM, Halla Rempt wrote:
On Tuesday, 4 May 2021 16:26:22 CEST Nate Graham wrote:
Again, my point was not that everything about GLI is universally better
than everything about BZ. Just that most things are mostly better in
most ways that most of us care about.
List them? I cannot think of a single thing that's good...
* Far simpler interface means that it's harder to get confused and
accidentally add information in the wrong place
* Inline images makes it much easier to describe a bug visually compared
to BZ where images are attachments that you have to go out of your way
to view in another tab or whatever
* Being able to edit the text allows for the correction of typos and
adding additional information rather than useless extra comments
* Mentioning an issue in another issue automatically creates a link to
it, rendering the "See also" feature of BZ obsolete and removing the
manual chore of having to manually link issues
* Tighter integration with the GitLab Merge Request feature without the
need for a fragile bot
* Tighter integration with other GitLab features such as milestones
* Many more "closes X" keywords are accepted, reducing the chance that a
merge request will try to close a bug report on merge but fail due to
not getting the syntax quite right
* Per-project templates
* One fewer dedicated service for our overstretched sysadmins to host
* Responsive upstream
* Possibility of improvements over time
I could go on for longer if you want, but this is just what I came up
with after a few moments of thinking about it. I won't pretend that GLI
is perfect, and there are unresolved questions regarding how we would
handle something like the "plasmashell" product that aggregates bug
reports for user-facing components that live in many different repos.
The search is less powerful too (though much faster). But let's all try
to have an open mind. I don't think we'll get anywhere if anyone says,
"no way, no how, never ever, over my dead body."
Nate