On Tue, 2007-01-09 at 16:21 +0100, John Graham-Cumming wrote: > Dave Korn wrote: > > BTW, if you ever revise it, I noticed one minor omission in the > > discussion of using the -p option to mkdir: as well as building a > > whole path in one go, it also prevents mkdir from erroring out if the > > destination dir already exists, meaning that you don't have to use > > "-mkdir" and potentially ignore real errors as well as (irrelevant) > > EEXISTs. I'm sure you're aware of all that, but your target audience > > might not spot the implication if you don't make it explicit. > > That's a good point and worth mentioning. I'll make a note for a future > revision (currently trying to interest a publisher in a 'GNU Make > Cookbook' of all my articles > (http://www.cmcrossroads.com/content/category/8/147/268) on GNU Make stuff).
One other minor nit; the article implies that any change to any file in a directory modifies the directory's timestamp. That's not the case: only changes that modify the directory entry table, such as adding a new file, removing a file, or renaming a file (or subdirectory) will cause the directory's timestamp to change. That's one reason this issue is a bit tricky: sometimes it will work fine but other times it will rebuild everything. BTW, these semantics are true for POSIX systems; I'm not sure about Windows. Cheers! -- ------------------------------------------------------------------------------- Paul D. Smith <[EMAIL PROTECTED]> Find some GNU make tips at: http://www.gnu.org http://make.paulandlesley.org "Please remain calm...I may be mad, but I am a professional." --Mad Scientist _______________________________________________ Help-make mailing list [email protected] http://lists.gnu.org/mailman/listinfo/help-make
