# Is it time for setup.rb to kick the can?

**URL:** <https://www.ruby-forum.com/t/is-it-time-for-setup-rb-to-kick-the-can/134189>\
**Category:** Ruby\
**Created:** [April 8, 2008, 6:34pm UTC](https://www.ruby-forum.com/t/is-it-time-for-setup-rb-to-kick-the-can/134189 "2008-04-08T18:34:26Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Trans](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/t/dfb087/32.png) [@Trans](https://www.ruby-forum.com/u/Trans)\
**Post date:** [April 8, 2008, 6:34pm UTC](https://www.ruby-forum.com/t/is-it-time-for-setup-rb-to-kick-the-can/134189/1 "2008-04-08T18:34:26Z")

</div>

One upon a time setup.rb was THE way to install ruby packages. But  
over the last year or so, setup.rb has begun to fade from the scene.  
Installing to site\_ruby location has largely eroded in general. Even  
commonly used project assist tools have moved away from site\_ruby  
installation support.

The main reason is obvious, RubyGems has become all but ubiquiteous.  
No mystery there. But I suspect also, this has a lot to do with the  
rise of Mac users among the Ruby community --since Mac packages are  
folder based, similar to RubyGems.

Long time Rubyists still include setup.rb in there projects, they know  
doing so facilitates distribution managers in repackaging --most  
notably .deb’s.

But I wonder if it’s time to take the next step and forgo setup.rb and  
site\_ruby altogether. To throw down a gauntlet to the old FHS schools,  
that it’s time for them to adapt, rather then the other way around.

I personally think it is probably a noble pursuit. To put some bottom  
up pressure on Unix/Linux and FHS to move forward with better, more  
modern approaches.

Of course, to truly do so, Ruby itself has to embrace the design. It  
has started to do with the inclusion of Gems in 1.9, but we still must  
consider what limitations remain that prevent Ruby from fully moving  
in this direction. In other words can Gems bootstrap? Can Ruby itself  
be “gem updated”.

T.

P.S. In my personal regard, I’m coming at this from the point of view  
of wanting to access package metadata (eg. version) from within my  
scripts without having to maintain separate ruby code for it in the  
lib/ folder.

---

<div class="post-metadata">

**Author:** ![Trans](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/t/dfb087/32.png) [@Trans](https://www.ruby-forum.com/u/Trans)\
**Post date:** [April 9, 2008, 5:22am UTC](https://www.ruby-forum.com/t/is-it-time-for-setup-rb-to-kick-the-can/134189/2 "2008-04-09T05:22:13Z")

</div>

Trans wrote:

> Long time Rubyists still include setup.rb in there projects, they know  
> doing so facilitates distribution managers in repackaging --most  
> notably .deb’s.
> 
> But I wonder if it’s time to take the next step and forgo setup.rb and  
> site\_ruby altogether. To throw down a gauntlet to the old FHS schools,  
> that it’s time for them to adapt, rather then the other way around.

Thank you so very much for finally addressing this which has caused me  
so many sleepless nights. Those “long time Rubyists” who hold onto  
their “setup.rb” tighter than their macbooks think they rule the world  
and if we don’t stand up to them now then who will???

When I compile my own gems by running ruby setup.rb I get a sour taste  
in my mouth that lasts for days and my hands feel dirty no matter how  
many times I try to wash the filth away.

It has lead me to monkey patch my test suite and rename all my methods  
commence\_the\_initialization\_of\_the\_test\_cases() which I find much better  
than the default setup() which ofcourse reminds me of setup.rb

I will gladly join your campaign and fight if there is at least a chance  
that I can spare another developer from the travesty of having to type  
“ruby setup.rb”… ohh the horror of it!

ilan