Front page | perl.perl5.porters |
Postings from June 2002
Senior Analyst Paul Green wrote: > Could we also suggest that people who are not on one of the "known popular" > platforms file SUCCESS reports as well as bug reports? We don't need to > hear about Linux (or VMS or VOS or MacPerl, etc) but if someone shows up > from one of the more obscure OSes, it would be nice to know. I guess this > means including a list of "known popular" OSes. I think I mean the OSes > that p5p regularly deals with (whatever they are). On an only remotely related subject, the RC1 release still triggers a core-dump out of ext/List/Util/t/shuffle.t on SGI boxes with relatively recent compilers (MIPSpro Compilers: Version 7.3.1.3m for sure, and I observed it with a previous compiler as well). This isn't a real surprise, the problem has been around for a long time, and the best guess is a compiler error. (It passes all tests with gcc.) But the combination of SGI and a fresh compiler is a "known pretty popular" platform. So should the problem be acknowledged in some fashion to head off a slew of bug reports? A few feeble attempts to circumvent the problem with creative compiler options did (me) no good. On the other hand, I have never seen another unexplained core dump from any application I have run with the 5.7 family, so, although any compiler error is pretty scary, this one seems as benign as possible. -- jplThread Previous | Thread Next