Ralf Wildenhues wrote: >> with all deliberate speed. > > Not sure what you mean here. If you mean applying the patch without > a testsuite addition, then my humble reply is: if we had abandoned > that strategy just longer ago, we'd have avoided a few embarrassing > regressions.
Not this. Rushing ahead blindly would not be deliberate. > If you mean not waiting on MinGW to change: sure. This. Waiting (interminably) on MinGW changes would not be very speedy. I know most of these the libtool/mingw-cross changes under complaint were mine. However, I did at least try to ensure that they were a strict improvement in regards to performance on existing tests. If those tests didn't catch some problems, well...yeah, better test coverage is (mostly [*]) good. -- Chuck [*] except that it takes over three hours to run the testsuite natively on cygwin, on my 3GHz machine...and under Vista, I have to *attend* those three hours, thanks to this: http://cygwin.com/ml/cygwin/2008-04/msg00681.html http://preview.tinyurl.com/5dofbf Ditto on native mingw...