On Tue, 2006-11-14 at 17:58 -0800, Russ Allbery wrote: > This is basically why I think the best approach is to standardize on SUSv3 > plus test -a/-o (with a more complete specification) and local. Our > experience with previous rounds of this discussion is that everyone seems > to be able to agree that this is a reasonable feature set, it's already > implemented by the two primary shells we care about, and anything short of > this feature set would probably break enough packages that it's easier to > improve the shell than it would be to change all the packages. I'm a little confused. When I use "test -a", I'm not using a "bash feature"; I'm using a *test* feature, supported by /usr/bin/test for as long as it's existed, IIRC. Happily, Debian bash does not blow away the functionality of test when it makes a builtin. Some shells, sadly, do blow away that functionality. No shell is obliged to implement test with a builtin, and Posix makes absolutely no mention of it. Some shells choose to do so, and some of them do so in a way which breaks the functionality. Sad, but true. So, I think I have to take back my earlier agreement that we can separate this problem from the one you are attacking. It seems to me now that they are, in fact, the same issue. If you are proposing that anything installed as /bin/sh *must* support feature such-and-such *if* it has a test builtin, then I'm ok with that. (Since I think that Policy *already* says that it must support *all* the features of Debian's standard test, *if* it wants a builtin at all.) But now it sounds like you are saying that things installed as /bin/sh *do not have to support* such-and-such a feature. I cannot support this, for the above reasons. Can you live with only the positive requirements, and not the permissions to randomly override other parts of the behavior of Debian programs? Thomas
Attachment:
signature.asc
Description: This is a digitally signed message part