Lots of people are making atom.io packages (which is good), but many of these packages are “taking” the most obvious names.
I would guess many packages are going to get quickly abandoned. Is there any plan to allow people to take over the name (or ownership) of abandoned packages? Or in a few years will we have to search through language-x, language-x-continued, language-x-2, etc. to find the still maintained package for a particular language or usage?
I’m new to Atom, but have already fixed bugs and added features to a bunch of 3rd party packages. I’ve submitted a couple of pull requests, though it seems like the original authors are M.I.A. I totally understand, of course. I’m a busy developer too.
The other day, I thought of a potential solution. Why not identify packages by both a name and an author? For example:
Language RSpec by a-busy-developer
Language RSpec by a-not-so-busy-developer
The clever developers at GitHub can figure out what’s the canonical author of a package. Obvious choices include the repo’s owner or some author property in package.json.
This seems at its heart to be a duplicate of this topic:
And if the Atom team wants to curate packages to the extent that people are suggesting, more power to them. I don’t think it is necessary though.
RubyGems theoretically suffer from the same problem that is being described here because there is only a single namespace for gems. Take for example the cancan gem. It has generally been considered mostly abandoned and in dire need of some bug fixing love by the community at large. So someone came along, forked it and created the cancancan gem.
These problems have organic solutions that have existed and been accepted practice in open source software for many years now. Unless there is some wrinkle unique to Atom that drives these issues up in impact, I feel like the current system is fine.