-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 On Mon, Nov 17, 2008 at 12:58 PM, Eric Kow wrote: > I intend to apply these, thanks! Do you have any particular refactors > or tidy-ups in mind for the future? Just curious :-)
Only those which are suggested and which come to mind as I look through the tests. My motivation here is that refactoring the shellscripts makes it easier to see what's going on in the tests, and allows an evolution into some better solution like Haskell tests. (First, one could convert everything in 'lib' to a Haskell script; then one could rewrite tests one by one into Haskell tests; and so on. No need for a dramatic Big Bang.) > +tests/lib.sh: library for test scripts > tests/*.sh: factor out definitions of portable_pwd to lib.sh > tests/*.sh: factour out ~61 of the 'not' definitions to lib.sh > tests/*.sh: factor out windows abortion > --------------------------------------- > I trust that these are unchanged since Trent's review Yes. > Add --disable-unit to configure script > Remove mention of unstable branch from the homepage > --------------------------------------------------- > Patches from the now defunct unstable > > mv lib.sh -> lib & update callers > --------------------------------- > Whole bunch of replace patches > > tests/lib: rm shebang header > ---------------------------- > As requested by Trent > > tests/*.sh: remove "set -ev" from all callers of lib > ---------------------------------------------------- >> alias pwd=hspwd > > On a random note, I thought I should point out that portable_pwd can now > be replaced by something like this. Presumably putting this alias into > the lib should help. Also, I think this means we get to get rid of the > silly IFS hacks I thought this might be true as well, but I am chary of messing too much with the Windows-related stuff. I simply don't know if hspwd actually does work as well as portable_pwd (and presumably better otherwise it wouldn't be used at all), or whether it's just that hspwd and portable_pwd suck in different ways. > tests/README.test_maintainers.txt: add a section on the use of lib > ------------------------------------------------------------------ >> - -Output like "good" and "OK, passed 100 tests." is good. Output like >> - -"*** Gave up! Passed only n tests." is a shortage of quickcheck test >> +Output like "good" and "OK, passed 100 tests." is good. Output like >> +"*** Gave up! Passed only n tests." is a shortage of QuickCheck test > > I definitely appreciate the README cleanup! For future reference, we > generally like to keep orthogonal changes in orthogonal patches, as > requested in > http://wiki.darcs.net/index.html/DeveloperGettingStarted Ah. I'm a little used to wiki-editing where one throws in non-controversial changes while one happens to be at something important. - -- gwern -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) iEYEAREKAAYFAkkhvvMACgkQvpDo5Pfl1oKd+gCeKEhDVJRB5e2Gjr6zdnXEvYwk Ro4AoJQIDJDRu6sZUZ83SX/PZfjVYebw =D9bs -----END PGP SIGNATURE----- _______________________________________________ darcs-users mailing list [email protected] http://lists.osuosl.org/mailman/listinfo/darcs-users
