[Nitro] The future of Nitro, Og and Raw
transfire at gmail.com
Thu Oct 23 16:37:08 EDT 2008
I'm still listening. After George went south on the project I haven't
had any real motivation to continue. But it's a real shame
considering. I think it would be great for you to pick up the torch
and run with it.
On Oct 23, 11:10 am, chris <prp... at gmail.com> wrote:
> Not sure who all is still listening but I have something to say.
> As Dan North said a while back. Nitro is quite feature full and could
> really use some time to settle in with some bug fixes and a really solid
> release. I may end up taking a stab at this. Once I got some fixes into
> the Nitro and Raw dependancies and require statements. A simple "nitro
> hello;cd hello;nitro" works out of the box. I also converted the entire
> projects to "newgem" projects ( no offense intended ).
If it works for you, that's fine. Since I stopped working on Nitro I
recrafted my whole build suite anyway.
But how does newgem help except to build a gem?
> The rake build tasks
> in there are excellent and "rake local_deploy" works perfectly. I have not
> touched Og, it seems to be pretty solid already. (side note: I think Og
> could stand on it's own two feet as well but if you want me to hold onto it
> for a while I will)
I separated Og into it's own project. See http://ogden.rubyforge.org.
There is a git repo. It is pretty much exactly what is in the current
Nitro repo. I figured splitting Og off on it's own was the first thing
to do in working toward a new stable release. Unfortunately the tests
were converted to RSpec just before all that (not by me), and they do
not fully pass and I have not been able to track down all the issues --
I wish we had the old unit tests instead.
The main plans I had for the future of Og were to make "enchanting"
explicit, and remove as much dynamic code injection as possible (there
is a lot of that in this lib). But a new solid release before all
that, at this point, would probably be best to help get a little fire
going under it.
> Fixed Nitro and Raw gemspec dependancies
> Fixed missing "require 'raw/controller'" in raw.rb
> Replaced lib/facets.rb with a gemspec dependancy
> Converted Nitro to a newgem project to capitalize on the rake tasks and
> rubyforge integration
> Converted Raw to a newgem project to capitalize on the rake tasks and
> rubyforge integration
> I think I found some bugs in the admin part, which in my opinion is key to
> getting a new user's feet wet with nitro and og. I'd like to fix those.
> Massive build up of examples in the rdocs
> Move good examples to the prototype site produced by "nitro project_name"
> Include more details in the actual prototype site code and startup files
> like app.rb and config/debug.rb
> I do not plan to move away from using the current unit testing. However I
> will create new tests using rspec and rbehave. They are not mutually
Ultimately it's your call, but I'd prefer not going the rspec and
rbehave route. There are much more important things to do. If you want
to move toward BDD, using minitest's mini/spec (now included in Ruby
1.9, btw) or Shoulda would be a much easier, more tempered step in
> Get "rake docs" working
> Get "rake website_generate" working
> With your permission, Trans and George I'm talking to you, I'll put some
> more effort into this and see what I can come up with. None of my changes
> will mess with the api, my entire goal is stability and usability. After
> presenting my initial work to you guys, we can decide where you want to go
> from there.
That's more than considerate. Since the project has been down for the
count, I say, feel free to take her where you want. As you can tell
from my reply I'm still interested in seeing this project succeed and
am happy to help where as I can.
> Before I forget, Robert Mela, if your out there. I'd love to get a hold
> of your cheatsheets for inclusion in the actual rdocs. Looks like your site
> has been reclaimed by the domain goblins.
More information about the Nitro-general