On Thu, Jun 18, 2015 at 09:05:14AM -0700, Paul Rogers wrote:
> But there are other errors in the /js/ branch that seem to have the same
> perlish names in them.  That's why I was thinking it could be the
> perlmods were incompatible versions.  Getting just the perlmods to play
> nice was a "dependency hell".

Getting a new perl module to fit correctly in the build order is always
'fun', and the same thing can happen when one of them has a new
version - at least if you want to run the tests on each module.

At the moment I'm playing with the new release of biblatex-biber (2.1,
instead of the 2.1a I produced when it's dev indicated he was not
intending to produce an updated tarball - but apparently texlive has
required the new tarball) with perl-5.22.0 (that causes some failures
in biblatex-biber's tests), and sundry new versions of some of the
perl modules needed to test the required modules: those pulled in
three or four more modules (as in 'A needs B to run its tests', 'B
needs C to run its own tests').

Also, I allow for some conditional updates where a shipped core
module might be too old [ interestingly, those usually get a
warning during 'make', and the build - sometimes also the tests,
completes.  So looking at the logs after the event is necessary. ]

All part of building a distro :-(

ĸen
-- 
Nanny Ogg usually went to bed early. After all, she was an old lady.
Sometimes she went to bed as early as 6 a.m.
-- 
http://lists.linuxfromscratch.org/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Reply via email to