Hello, On Thu, May 25, 2006 at 01:37:11AM -0700, Paul Eggert wrote: > If we can come up with something nicer that still avoids the bugs, > that'd be good. But let's not revert the patch until we have > something nicer in hand.
sure, that's a good rule. First, let me mention what I see as the main disadvantage of your patch: config.status is less readable with all those expansions. And two two levels of here-documents make status.m4 difficult to read, so looking into a generated config.status is often useful. ``something nicer'' I would like to add a check which would rule out shells with this bugs from the ``better shell'' selection. If I understand it correctly, there is always another usable shell on the affected machines. I'll prepare a patch later. Stepan Kasal
