http://qs321.pair.com?node_id=1034252


in reply to 5.18.0 is available NOW!

ouch I see lots of problems on cygwin. Does not build with make -j 3 but maybe this is expected... Still make test fails for me. Some seem to be issues with file perms but the re ones seem worrisome. :/

make test output gives:

hth --steph Failed 26 tests out of 2336, 98.89% okay. ../cpan/Encode/t/CJKT.t ../cpan/Encode/t/at-cn.t ../cpan/Encode/t/perlio.t ../cpan/Pod-Simple/t/corpus.t ../cpan/Pod-Simple/t/pulltitl.t ../cpan/Text-Tabs/t/Tabs-ElCid.t ../cpan/Text-Tabs/t/Wrap-JLB.t ../cpan/podlators/t/man.t ../ext/XS-APItest/t/call_checker.t ../lib/File/stat.t io/shm.t op/bop.t op/taint.t op/threads.t re/fold_grind.t re/pat_rt_report.t re/pat_rt_report_thr.t re/regex_sets_compat.t re/regexp.t re/regexp_noamp.t re/regexp_notrie.t re/regexp_qr.t re/regexp_qr_embed.t re/regexp_qr_embed_thr.t re/regexp_trielist.t uni/overload.t

a rebase at a fixed address should probably be done just after the build... as some dlls gave a fork error. Most tests could use a "spawn" instead of fork-exec. Please p5p give us a spawn this would be useful on cygwin. These days I maintain a rebase db for all my perls. A given dll gets the same base address. I don't use the cygwin rebase db, it does not fit my needs for many perls and a third of cpan installed ;)