Folks: We're back on 3.76.1 because we're running GNU make on a non-mainstream platform for which the port was done some years ago -- nobody's done the update port to 3.79.1 yet (although I'm giving it serious consideration).
A colleague of mine asks an interesting question about VPATH. He claims that, within the execution of a single makefile, he has just made an object within a directory path named as one of the VPATH operands (not cwd). The very next rule within that same makefile states a dependency on that object, and make cannot find it through VPATH. Is this a bug? The documentation, as far as I can read, gives me no hint whether or not the directory search is performed *in the filesystem* for each new rule (dependency list) or not. From the behavior he claims, I'd have to guess that make builds a list of the VPATH directory contents once, and doesn't refresh it. Help? If this is a bug, I've got a case that can be reproduced. If it's not, I'll be happy to correct my colleague. I've checked the ChangeLog in 3.79.1 and no corrective items jump out at me. Thanks in advance, Jim Tison _______________________________________________ Bug-make mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-make
