[BioRuby] Ruby installation

[Pjotr Prins]

On Thu, May 01, 2014 at 05:31:17PM -0700, Joachim Baran wrote:
> On 1 May 2014 17:18, Iain Barnett <iainspeed at gmail.com> wrote:
> 
> > You *provide* the VM. You *provide* the files. You *provide* the tools you
> > used to build the VM. You *provide* the process you used. Then anyone can
> > jump in at any level, either just load a file, or load a file and
> > reconfigure, or tweak your process and dump whatever isn't wanted.
> >
>   I think only a few bioinformaticians would actually go through that
> process -- probably many more if everything were very easy and well
> explained. I would bet that simply no single biologist or statistician
> would be willing to load their data this way.

Yup. VMs are available - I particularly like CloudBiolinux since I
added flavors to that - and more is coming. VMs, indeed, do not solve
the problem I am referring to.

> >
> > Either you want things configurable and bespoke, or you want them simple
> > and straightforward. You can't have your cake and eat it.
> >
>   That is true. I still would like to hear from others the reasons why 'gem
> install bio' fails. Are there dependencies that should be culled?

Actually, BioRuby itself is pretty good since it does not depend on
other software. But there are plenty other examples, SciRuby -> Atlas,
Bioruby-blastxml -> libxml, etc. And then gems depending on each
other. I think SciRuby is the real test case we should sort.

Pj.