On Mon, Feb 25, 2008 at 6:59 PM, Otto Hemmi <ottoman01 / gmail.com> wrote:
> My work environment is very controlled. We validate certain
>  configurations of servers on our staging/QA servers before moving to
>  production. Before updates to gems that are different than our
>  validated version can be deployed we must first test them on the
>  staging and QA servers. As a result i'd like a better method to
>  control the versions of ruby gems that are installed on the server.
>  The best idea for this is to check local copies of gems into the
>  source repository and include these in the release (thus freezing a
>  versioned gem list). One thing i've noticed is that if you attempt to
>  install local gems gem cannot resolve local dependencies inside a
>  directory. So building a directory with all the gems and then running
>  "gem install *" or the flag "--include-dependencies"seems to not
>  work...
>  Am i going about this the wrong way? is there a way to get gem to
>  resolve local dependencies?
>  suggestions or recommendations? My experience with ruby is defiantly
>  limited so if i'm asking a stupid question please be nice....

You can run your own gem_server, and install gems from there using
--source=http://whatever/