/usr/bin/ruby: no such file to load — auto_gem (LoadError)
Scott Rippee @ 8:46 pm December 17th, 2006Gentoo decided to throw me a hurdle when trying to `emerge rubygems` tonight:
CODE:
-
>>> Install rubygems-0.8.11-r5 into /var/tmp/portage/rubygems-0.8.11-r5/image/ category dev-ruby
-
/usr/bin/ruby: no such file to load -- auto_gem (LoadError)
-
/usr/bin/ruby: no such file to load -- auto_gem (LoadError)
-
-
!!! ERROR: dev-ruby/rubygems-0.8.11-r5 failed.
-
!!! Function ruby_einstall, Line 125, Exitcode 1
-
!!! setup.rb config failed
-
!!! If you need support, post the topmost build error, NOT this status message.
The solution is to clear the RUBYOPT variable:
CODE:
-
RUBYOPT="" emerge rubygems






January 30th, 2007 at 5:30 am
And this does what?
August 19th, 2007 at 11:19 am
Thanks very much.
October 5th, 2007 at 3:26 pm
Thank you! This was driving me nuts, and http://bugs.gentoo.org didn't show anything for rubygems that I found.
January 14th, 2008 at 4:07 am
This helped me repairing my ruby install... big time. Thank you!
March 12th, 2008 at 11:07 pm
Thanks! Exactly what I was looking for. I couldn't get ruby 1.9 to run on gentoo.
June 13th, 2008 at 11:18 am
Thank you very much.
It solved my problem with Ruby Enterprise Edition + Gentoo.
October 6th, 2008 at 7:21 am
Nice! thx u!
October 13th, 2008 at 2:30 am
yesss,
this solved my errors in amarok 1.4
Tnx a lot master
May 6th, 2010 at 1:53 am
thanks, this was exactly what i've searched for. here we go REE :)
November 30th, 2010 at 8:36 am
Cheers. We have to use Gentoo at work (it's either that or Windows!) and I've realised that when I 'emerge' (install) anything, it might just update Rubygems in the background, breaking it. The sysadmin guys all say Gentoo is the dog's bollocks, but I can't quite understand why. There isn't an 'ebuild' for Rails 2.3.9, for example, so we have to upgrade every app to Rails 2.3.10 before delivering it into production.
December 30th, 2011 at 2:03 am
@Rod: The only generalization that can be made, is that people who generalize, can in general not be taken seriously. (And because that includes this instance, there are exceptions to this rule too, making it a valid rule. … Or circular reasoning … I can’t decide … ;)
Don’t just run emerge like that. Run it with -atv (ask, tree view, and verbose). That way it shows you a nice tree of dependencies it installs, and asks you politely before going ahead. ^^
And for updates, this is the result of 8 years of Gentoo experience:
1. eix-sync (Requires /etc/eix-sync.conf to be set up like shown below.)
2. emerge -auDNtv --changed-use --keep-going world
3. Run any language-specific updaters. (Like haskell-updater or python-updater)
4. emerge -1 @preserved-rebuild && revdep-rebuild && emerge -a --depclean && etc-update # @preserved-rebuild requires Portage 2.2, so unmask it.
If there are any problems, you may have to go back to step 2 again, after fixing them. (Don’t worry, it will go much faster the second time.)
The /etc/eix-sync.conf:
# Order according to the script: (“*” replaced by “!layman -S”.)
!layman -S || echo "Continuing, despite errors in Layman."
# emerge --sync gets executed here
@egencache --repo=local --update # Only if you use your local portage overlay
#@cd /var/lib/layman/kde/Documentation && sh ./metadata-sync # Only if you use the kde overlay
@emerge --regen || echo "“emerge --regen” returned error status $?."
# #emerge --metadata (Not active because of FEATURES="metadata-transfer" being on and required by overlays that modify eclasses.)
# eix-update gets executed here, if eix-cache isn’t newer than tmpfile
@@eix-remote update
# eix-diff gets executed here
January 4th, 2012 at 6:32 pm
Barefoot - thanks for your reply, exactly 13 months after my post. In the meantime, I've become a Gentoo geek. I now know why Gentoo is the number one distro of Linux for robust delivery of server systems. I've written several ebuilds, complete build scripts for servers using the Gentoo 'emerge' package manager, which is the best in the business.