From: Peter Rabbitson Date: Thu, 11 Feb 2016 17:07:51 +0000 (+0100) Subject: Add experimental meta flags signifying the status of parallel testing X-Git-Tag: v0.082821~2 X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=commitdiff_plain;h=a941f202a0097b40c191d707005f59743117f5cf;p=dbsrgits%2FDBIx-Class.git Add experimental meta flags signifying the status of parallel testing --- diff --git a/maint/Makefile.PL.inc/21_set_meta.pl b/maint/Makefile.PL.inc/21_set_meta.pl index 845728e..cfb12e7 100644 --- a/maint/Makefile.PL.inc/21_set_meta.pl +++ b/maint/Makefile.PL.inc/21_set_meta.pl @@ -6,6 +6,18 @@ author 'ribasushi: Peter Rabbitson (present day maintenance # pause sanity Meta->{values}{x_authority} = 'cpan:RIBASUSHI'; +# !!!experimental!!! +# +# am wondering if an x_parallel_test => 1 and x_parallel_depchain_test => 1 would be of use in meta +# to signify "project keeps tabs on itself and depchain to be in good health wrt running tests in parallel" +# and having cpan(m) tack a -j6 automatically for that +# it basically allows you to first consider any "high level intermediate dist" advertising "all my stuff works" so that larger swaths of CPAN get installed first under parallel +# note - this is not "spur of the moment" - I first started testing my depchain in parallel 3 years ago +# and have had it stable ( religiously tested on travis on any commit ) for about 2 years now +# +Meta->{values}{x_parallel_test_certified} = 1; +Meta->{values}{x_dependencies_parallel_test_certified} = 1; + # populate x_contributors # a direct dump of the sort is ok - xt/authors.t guarantees source sanity Meta->{values}{x_contributors} = [ do {