What pushed me away from contributing patches and improvements was the Contributor Agreement. I'm sure others feel the same as it is a pretty large barrier to entry.
Making each clojure.contrib.* library more independent makes sense, but it makes even less sense now to keep everything under CA. As Rich said, it is not a standard library - it isn't strictly affiliated with clojure proper.
It seems to be a cover-my-bases way of saying "If you contribute, we have joint-ownership of the code." Which is, I think, reasonable if you're going to contribute code to a project.
I don't care whether copyright assignment is given or not, my objection is to anything getting in the way of improving Clojure. Somewhat frequently you see people come into the IRC channel with a simple bugfix being told their fixes can't be used right now. Often they never get around to resubmitting them once the paperwork has landed. First impressions count for a lot, and I suspect the initial brush-off leads to people being less likely to contribute when they do have something more substantial they could do.
Of course, in the case of Clojure copyright assignment is hardly the biggest impediment to getting patches applied, so maybe this is moot. It is interesting to note that Clojure has only had 33 external contributors since switching to git in the middle of last year, while my own project Leiningen has had over fifty despite being a younger, much lower-profile, and less interesting project:
> Barring that, is not completion of the CA a reasonable a priori proxy for the level of commitment of a contributor, where a base level of commitment is desirable?
Wait, so now lazy people don't have good ideas? As a lazy person, I take offense at that notion.
Saying you're lazy seems absurd on the face of it. :-)
But: ideas mean nothing without execution. What if the great ideas are only offered as git pull requests? Or are based on Clojure 1.1.0, but the author is too lazy to do the work to rebase/merge? Or are pasted into a gist and linked to in #clojure?
There's lots of criteria that are applied to whether or not a patch gets in, totally aside from its technical merits. The CA just happens to be one that gets a lot of ink.
well, humor aside, there are different grades of lazyness.
I'm a geek, I feel pleasure to do stuff on computers.
Tell me to print something, ehm find some paper somewhere without coffee stains, connect the printer, print it, sign it, put it somewhere on the desk so next day (it's night now) I can send it via snail mail, buy stamps etc etc
aaaargh, no simply no, not for a 5 lines fix in the contrib lib.
This 5 line fix, might have been another 10 lines tomorrow, then perhaps something more in future, but for now they are 5 lines and this procedure is actually blocking my gradual involvement in the whole thing.
It would be different if I had a real interest, if I had project I wrote and I would like to integrate, then perhaps I would find the whole operation slightly less useless. But for a quick bug fix ....
I would recommend reading the most recent messages on the linked thread. Rich tries to address your concern and, whether or not you agree with his decision, it's clear he's done his research.
This is great. The contrib collection had this noob a bit confused when I started, thinking it was some sort of standard lib. Then I found it to be a bit less stable and organized than most standard libraries (but chock full of goodies!).
I like the approach of using GitHub and making each piece its own repo.
Making each clojure.contrib.* library more independent makes sense, but it makes even less sense now to keep everything under CA. As Rich said, it is not a standard library - it isn't strictly affiliated with clojure proper.