> This change is going to be a nightmare for us users.
Some of us noticed that the documentation made no promise, so always
coded to not assume sorting ...
> Things that have worked "forever" now break.
True. Then again, I'm looking forward to this as an opportunity to
catch bugs in our make files, where someone has made changes that
cause an order-dependency that I'd specifically declared they should
never assume. I definitely regard such order-dependencies as bugs,
already - it's just not been easy to police until now.
> The last thing people want to do is edit
> production code to fix things like this.
Hyperbole.
I can think of dozens of things later than this !
Starting from scratch and re-writing everything anew, for example.
I'm sure those who have this problem shall gripe, and it'll hold some
back from upgrading to 3.82, but they only need to wrap $(sort ...)
round each affected wildcard - it's far from the end of the world.
And make shall be faster.
Eddy.
_______________________________________________
Bug-make mailing list
Bug-make@gnu.org
http://lists.gnu.org/mailman/listinfo/bug-make
