On 5/22/07, edlich / gmail.com <edlich / gmail.com> wrote:
> Hi,
>
> > 5. Is there a (free) UML modelling Tool for Ruby on the market?
> > -> I need to have the architectural UML Overview!!
> > (How do all the gbig ruby projects do this??)
>
> So partially I figured out for myself why a UML modelling tool for
> Ruby
> is not present yet.
>
> The reason is not (as some here mentioned) that ruby code is so small
> and ruby is so agile (although both is true) that nowone needs a
> modelling tool.

That's probably the main reason.  Actually Ruby's properties tend to
attract practitioners of agile development "methodologies" where the
focus is on developing and evolving code rather than superflous (and
often out-of-date/sync) artifacts like diagrams.

This means that the ruby community is less likely to be motivated to
write a UML tool.

The debate over the usefulness of case tools and diagrams in agile
development has been going on for quite some time, even pre-dating
terms like agile, or extreme programming, back in 1993 we were more
likely to call it rapid or iterative development.

http://www.informatik.uni-trier.de/~ley/db/conf/oopsla/oopsla92.html#DeNataleBCTT92

>
> The reason is more rubys dynamic nature.
> * Types are not that detectable
> * Variables can be changed and (re?)declared everywhere
> * And ruby contains such cool / weird (?) things as instance-class
> variables:
> Have you ever seen this:
>   class << self; attr_accessor :sides end
>
> No? Then have a look at this site
> http://railstips.org/2006/11/18/class-and-instance-variables-in-ruby

Written by someone who was admittedly a ruby newby at the time he wrote it.

> So now I guess you have better understanding why your boss should pay
> you
> double income if you have to build a ruby UML modelling tool.

There's a difference between applying UML to Ruby and writing a UML
tool in Ruby.  There are probably difficulties in modeling advance
Ruby features in UML, but that seems to me to be a deficiency in UML
rather than Ruby.

On the other hand there's nothing inherent in Ruby which make it
unsuitable for writing a UML tool should  a rubyist with a motivation
to do so were found.  Dynamic languages like Ruby make fine tools for
implementing such.

In fact Grady Booch wrote the original Rational ROSE in Smalltalk, but
it got most of its use by C++ programmers.

-- 
Rick DeNatale

My blog on Ruby
http://talklikeaduck.denhaven2.com/