Hi Hugo,
Hugo Villeneuve wrote:
Yeah, I got blocked with "bash" dependent ports (ruby-2.0 for subversion).
what is happening for you?
I have temporarily given up building on my SS20: all modules I have are
indeed unstable, both the 50Mhz and 40Mhz SuperSparcs just segfault at
will, as Miod has too.
I fixed and upgraded my SS5, MicroSparc II 110Mhz, which was running
OpenBSD reliably for years untill the gcc 2.95 switch, when I gave it up
for the moment.
I have a fresh install of 5.6 on it and started building several base
packages, which worked fine. It seems stable even if it compiles for days.
When I try compiling subversion (make package bulk=Yes) I get this:
`/usr/ports/distfiles/swig-2.0.11.tar.gz' is up to date.
>> (SHA256) swig-2.0.11.tar.gz: OK
===> swig-2.0.11p0 depends on: ruby->=2.0,<2.1 - not found
===> Verifying install for ruby->=2.0,<2.1 in lang/ruby/2.0
===> Configuring for ruby-2.0.0-p481
Using /usr/ports/pobj/ruby-2.0.0-p481-no_ri_docs/config.site (generated)
Bus error (core dumped)
*** Error 138 in /usr/ports/lang/ruby/2.0
(/usr/ports/infrastructure/mk/bsd.port.mk:2704
'/usr/ports/pobj/ruby-2.0.0-p481-no_ri_docs/.configure_done')
checking for core files, I see this:
$ find . -name \*.core
./pobj/ruby-2.0.0-p481-no_ri_docs/ruby-2.0.0-p481/bash.core
./tests/portcheck/t5/some.core
Is this what you get too?
do we really need bash to build ruby? and... why ruby for subversion?
not counting shells one ends up having perl, python, tcl and ruby! what
a mess.
Riccardo
