[tpm] MetaCpan VM

Matthew Phillips mattp at cpan.org
Wed Oct 30 12:14:22 PDT 2013


To MetaCpan's credit the added complexity of a virtual machine is actually
removing complexity of an entire stack required to run metacpan, which is
solving a very complex multi-faceted problem. Using the VM saves you from
setting up
-an OS
-a perl installation
-elasticSearch
-metacpan itself
  -both the web part, and the api part
  - its cpan dep tree, including catalyst,moose,elasticSearch
-the minicpan logic, its scraping
-probably some other things I haven't thought of

It's not a simple system - the vagrant / vm solution saves a fair bit of
effort.

Cheers,
Matt


On Wed, Oct 30, 2013 at 12:59 PM, <arocker at vex.net> wrote:

> >
> > I don't have a real opinion on that, but complicated problems don't
> always
> > have simple solutions.
>
> What I was grumbling about was complicated solutions to simple problems.
> "Solutions" shouldn't be more complicated than the problem they set out to
> solve.
>
> > What we're trying to do here is give you access to a service with
> multiple
> > layers without forcing you to know about everything that is happening
>
> I didn't mean a particular criticism of MetaCpan's process; it seems quite
> normal by today's standards.
>
> Adding a program used to be a simple matter of putting a file in an
> appropriate directory. Now, it's unpack a tarball, initiate an install,
> hope that it uses compatible libraries, has no unfulfilled dependencies,
> and doesn't do anything ugly to the existing directory hierarchy.
>
> _______________________________________________
> toronto-pm mailing list
> toronto-pm at pm.org
> http://mail.pm.org/mailman/listinfo/toronto-pm
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.pm.org/pipermail/toronto-pm/attachments/20131030/cdd09912/attachment.html>


More information about the toronto-pm mailing list