In gearman-0.28, configure.ac has some bad copypasta that ends up
breaking the build on platforms that do not have librt, such as
MacOS X.
configure.ac --
173 # Check for -lrt
174 AC_CHECK_LIB([rt], [clock_gettime],
175 [
176 RT_LIB="-lrt"
177 AC_SUBST(RT_LIB)
178 AC_DEFINE([HAVE_LIBRT], [ 1 ], [Have clock_gettime])
179 ],
180 [
181 AC_DEFINE([HAVE_LIBRT], [ 0 ], [Have clock_gettime])
182 ])
183
184 # Check for -lm
185 AC_CHECK_LIB([m], [floor],
186 [
187 M_LIB="-lrt"
188 AC_SUBST(M_LIB)
189 AC_DEFINE([HAVE_LIBRT], [ 1 ], [Have clock_gettime])
190 ],
191 [
192 AC_DEFINE([HAVE_LIBRT], [ 0 ], [Have clock_gettime])
193 ])
As the comment on line 184 indicates, it's supposed to be checking for
libm, not librt. This commit corrects this, which fixes issue Homebrew/homebrew#10222.
ClosesHomebrew/homebrew#10823.
Signed-off-by: Jack Nagel <jacknagel@gmail.com>
Device names under OS X are very long. Previously a short form was
available, but this seems to no longer be the case. This patch doubles
the size of the buffer for device names, allowing the real device name
to fit. If they don't fit, device autodetection fails.
See http://trac.macports.org/ticket/27819ClosesHomebrew/homebrew#10388.
Signed-off-by: Jack Nagel <jacknagel@gmail.com>
Finally, a new stable release. With this we can remove some old cruft
from the formula. Also updating winetricks.
Support for --devel is kept because we expect development releases to
continue at the usual brisk pace.
Signed-off-by: Adam Vandenberg <flangy@gmail.com>
"E is a theorem prover for full first-order logic with equality. It
accepts a problem specification, typically consisting of a number of
first-order clauses or formulas, and a conjecture, again either in
clausal or full first-order form. The system will then try to find a
formal proof for the conjecture, assuming the axioms."
[http://www4.informatik.tu-muenchen.de/~schulz/E/E.html]
Signed-off-by: Adam Vandenberg <flangy@gmail.com>
