"Bleeding edge" can be a concern if it has implications on for bugs that
affect user experience. I don't know if it is Haskell or the skills of
the xmonad team (probably both), but I have always used the latest
xmonad and found it remarkably bug-free. In fact, I would find it
difficult to name any other piece of nontrivial software that has a
similar level of stability.
Best,
Tamas
On Mon, Mar 02 2015, Alexander Genaud
Or, if x.x.B already has meaning (bugfix release), then rc (release candidate) notation would serve a similar purpose:
0.12.5-rc1 ==> 0.12.5 (bugfix release) 0.13-rc1 ==> 0.13 (major/minor release)
I think the difference is EVERY Java release must be a supported standard target. Xmonad has local, but no global, critical child dependencies. One would never recommend that a general user compiles bleeding edge Javac for any real work -- yet that's what some Xmonad-ers have recommended for two years.
If 0.12.3 exists, but is not considered stable, then 0.12-rc3 or 0.12b3 would seem more appropriate.
Oh, yes.. i didn't check the page. I've been too much in the java world using the snapshot descriptor, it would be cool if we had such a thing in xmonad in where 0.12 would be release notation for example and 0.12.5 would be upstream notation, and a preparation for 0.13. 0.12.5 is released as 0.13 and the darcs version is updated to 0.13.5 and so forth.. I attach the patches in case this resonates with you.
xmonad mailing list xmonad@haskell.org http://mail.haskell.org/cgi-bin/mailman/listinfo/xmonad