Nice project. I like seeing people make their own tools.
The part that jumps out at me, though, is that this had to be specifically targeted at GitHub Issues. Why is bug tracking still a bunch of silos? I hear programmers speak of "open standards" and "interoperability" and "separation of concerns", and still the otherwise smart engineers at Bugzilla/GitHub/BitBucket/Trac/Jira/Launchpad/FogBugz/... go and build these systems which are 98% identical but provide no common interface.
I can share events across calendars, vector images across graphics editors, audio files to hardware players with different codecs, source code history to different version control systems, and nearly every other impossible-sounding cross-platform conversion, but with bug tracking, the shoemaker's children go barefoot.
(The one platform that gets partial credit here is Debian, because they use plain email, and you can rsync yourself a copy of the entire database. It's not exactly pleasant but at least I can process everything with existing standard tools.)
What's the logic here? Those principles of software architecture are good for every type of software except bug trackers? If we keep your issues hostage, you're less likely to leave our service? We're each going to try to become dominant so our proprietary API is the standard?
The part that jumps out at me, though, is that this had to be specifically targeted at GitHub Issues. Why is bug tracking still a bunch of silos? I hear programmers speak of "open standards" and "interoperability" and "separation of concerns", and still the otherwise smart engineers at Bugzilla/GitHub/BitBucket/Trac/Jira/Launchpad/FogBugz/... go and build these systems which are 98% identical but provide no common interface.
I can share events across calendars, vector images across graphics editors, audio files to hardware players with different codecs, source code history to different version control systems, and nearly every other impossible-sounding cross-platform conversion, but with bug tracking, the shoemaker's children go barefoot.
(The one platform that gets partial credit here is Debian, because they use plain email, and you can rsync yourself a copy of the entire database. It's not exactly pleasant but at least I can process everything with existing standard tools.)
What's the logic here? Those principles of software architecture are good for every type of software except bug trackers? If we keep your issues hostage, you're less likely to leave our service? We're each going to try to become dominant so our proprietary API is the standard?