[BioRuby] Ruby installation

[Joachim Baran]

Hi!

  Pjotr, what a great post! Awesome!

  Can we focus on *why* BioRuby seems to be failing in some environments?
What are the gems that are causing trouble?

  On the related side, the next release of BioInterchange is overdue, but I
solved a few dependency issues there. Basically, I noticed that I only use
a tiny fraction of the functionality offered by most gems that I relied on.
I then replaced their functionality with a few lines of specific Ruby code
that did the same for my purpose. My aim there is actually to make
BioInterchange work with mruby, but this approach might also be viable for
BioRuby.

  I agree that BioRuby should be deployable in three lines:

  $ gem install bio
  $ irb
  > require 'bio'

  That *must* work on all OS X machines out of the box. It should only
require one more apt-get (or whatever) on Linux machines. If it does not,
then we failed, because it will be too hard for biologists to install.

Kim


On 1 May 2014 00:39, Pjotr Prins <pjotr.public14 at thebird.nl> wrote:

> On Thu, May 01, 2014 at 08:21:03AM +0100, Iain Barnett wrote:
> > VMs are easy, just install the software, then run the the file.
>
> So how do you get your data files onto the VM? Yes, that requires more
> instructions. Next you need to explain how a container works inside a
> computer. Then they use 2 editors on 2 accounts to work with data?
> Come on guys, it is not that straightforward. I also use VMs all the
> time (I am writing this E-mail on one) and I contributed quite a bit
> to CloudBioLinux.
>
> You are still mixing in expert advice. What works for you is not
> always easy for others to grasp. Expert advice is for experts.
>
> I simply care about others using my software. What they experience is
> that it is too HARD to use my software. The HARD part is deployment.
> Which sucks because I write great software ;). I am not targeting Ruby
> programmers here, my first audience is people who want to run a Ruby
> tool on the command line without everythink exploding in their face.
>
> But even a newbie should be able to simply install a tool and program
>
>   require 'bio'.
>
> It is amazing how many fail around me. I can only guess to extrapolate
> what that means on a mondial scale. Ruby is going to lose out. Better
> have them learn some easier to deploy language if we remain in this
> mind set - that is what I am thinking.
>
> When you actually read
>
>   https://news.ycombinator.com/item?id=6891214
>
> Ruby is mentioned specifically. Yesterday, on Slashdot the same
>
>   'Ruby makes sysadmins cry. I tried updating a legacy server yesterday
> that is running a ruby app. After two hours of trying to make it work, I
> gave up.'
>
> in
>
>
> http://ask.slashdot.org/story/14/04/30/1344224/c-and-the-stl-12-years-later-what-do-you-think-now?utm_source=rss1.0mainlinkanon&utm_medium=feed
>
> I am not making this up. These are both non-Ruby threads! We have a real
> and
> acute problem if you start reading up on this shit and see people struggle.
>
> rvm's and VM are actually the problem, not the solution. We are happy
> in our own little boxes ignoring the outside world. Remember some
> facts about evolution?
>
> We need to come up with a sensible protocol that I can give to any
> student for his laptop and any system adminstrator on a compute
> cluster.
>
> Software is software. We can solve it. The only reason we are not
> solving it is that *we* do not really care about other users. We are
> ignoring them.
>
> Pj.
>
> _______________________________________________
> BioRuby Project - http://www.bioruby.org/
> BioRuby mailing list
> BioRuby at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/bioruby
>