On Tue, Jan 15, 2019 at 10:52:50PM +0100, Patrick Storz wrote:
Actually I think there was a slight misunderstanding: Rather than implementing a full-featured new piece of support software and delaying the issue migration to GitLab indefinitely, I was actually suggesting to use GitLab to implement the basic concept of friendly feedback today!
My suggestion is to create a single subproject like https://gitlab.com/inkscape/inkscape-bugs that is used as the primary user-facing issue tracker from where reports can be triaged and sorted into the more technically oriented trackers of the suitable subprojects, e.g. https://gitlab.com/inkscape/inkscape/ https://gitlab.com/inkscape/vectors/ https://gitlab.com/inkscape/extensions/ https://gitlab.com/inkscape/inkscape-web/ https://gitlab.com/inkscape/inkscape-docs/ https://gitlab.com/inkscape/lib2geom/ and many more...
Ah, yes that does sound doable. Now that you mention it I do remember discussing this before. I'd totally be open to this.
If the incoming queue is to be general including for lib2geom, perhaps it should be named something more generally, like
https://gitlab.com/inkscape/report
Or
https://gitlab.com/inkscape/inbox
And presumably it's where we would want to direct inkscape.org/report?
From the sheer amount of these trackers it's pretty obvious it will be hard to impossible for users to find the correct tracker on their own, so the idea is to provide one single point of entry that can be used to report *any* Inkscape-related bug and doesn't require the reporter to have an in-depth knowledge of the internal workings of the project. (Let's be honest: Even the most experienced members of the project are not involved in all of the sub-projects that make up Inkscape today!)
Yes, and there is also a desire to direct wishlist issues to something else, and this approach would enable doing that.
If this additional layer is deemed to be unnecessary (i.e. the current plan) we have to expect people will use https://gitlab.com/inkscape/inkscape/ as the "catch all" bug inbox. It will be hard to impossible to keep on top of it, especially if we do not start rigorously closing issues that are not clearly bugs (something we did *not* do in LP so far; we usually supported users where we could!). Unfortunately this is necessary as - despite your optimism Bryce - we simply don't have the resources to give every report the attention it deserves without jeopardizing the central function of this tracker, which is efficient developer-facing bug-tracking for the core program itself. In other words: There won't be a lot of room for "friendly feedback".
I hope I made myself clearer this time. If so my main point is that there's a window of opportunity here to avoid having mixed opinions / feature-requests / support requests / questions / etc. into the core program's tracker all over again (status quo in LP). However this window is closing fast - so think twice and choose wisely...
Thanks, yes that is clearer. I was hoping you had a better idea on how to achieve this than me, and am glad you did.
We have some issue tickets open on how we should handle the bug tracker, I'll review through those, and see if we can reuse one of them to discuss this plan.
Bryce
Cheers Patrick
Am 15.01.2019 um 20:43 schrieb Bryce Harrington:
Yeah, I would love to see this capability, and think it'd help users and developers a lot. And you're right this would be the perfect time to introduce something. But other than roughing in the concept, progress on implementing it has not moved very far forward. I'd love to see us work on introducing a solution for this in the future, so I'd like us to keep it as a long term goal, but I don't think it is worth delaying moving ahead on using gitlab issues.
I also worry about seeing an unmanageable amount of bug reports coming in. Hopefully that doesn't happen soon. If it does start to become a problem we'll need to rethink plans. There may be more than one way to solve this problem, so it's worth keeping open to alternate ideas.
Bryce
On Tue, Jan 15, 2019 at 06:35:41PM +0100, Patrick Storz wrote:
Related to this very question: What's the status on "friendly feedback" and the idea of separate (user- and developer-facing) trackers?
During the last discussions I was involved in there seemed to be a whish to keep especially https://gitlab.com/inkscape/inkscape/issues "internal" in the sense that only developers / experienced users should directly open issues against the main project, so it would only contain valid and easily reproducible bug reports in the long run.
I think the intention was that the vast amount of "unfiltered" reports should be made against a "catch all" tracker for triaging, where the wider community could help to sort reports into the relevant sub-projects and opinions / feature requests / support requests / invalid reports could be caught early and answered accordingly (hopefully in a more friendly way than "not a bug, closing").
Is this still a goal or (to say it bluntly) do we just continue as-is, which will likely mean we just start fresh in collecting an unmanageable amount of bugs? ;-)
If we *were* to make this cut I feel like *now* would be the time to do so, before the inkscape/inkscape tracker is flooded with "random" reports triggered by the upcoming releases.
Cheers Patrick
Am 15.01.2019 um 17:54 schrieb Marc Jeanmougin:
I'll be glad to make a post in forums about this. But is this the correct link?
https://gitlab.com/groups/inkscape/-/issues
Or is it this? https://gitlab.com/inkscape/inkscape/issues
The first one shows all bugs from all the Inkscape projects (Inkscape, extensions, website, vectors…) while the second only pertains to Inkscape
Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel