From pat.eyler at gmail.com  Wed Feb  1 10:06:30 2006
From: pat.eyler at gmail.com (pat eyler)
Date: Wed, 1 Feb 2006 08:06:30 -0700
Subject: [RPS] Ruby Quiz #64 - Port a Library
In-Reply-To: <9e272eff0601311433p3cb6dac5l1f036d31d00431f5@mail.gmail.com>
References: <9e272eff0601311302v5adb797dv8e43d81c4627905f@mail.gmail.com>
	<6fd0654b0601311325p3d9f9fbej168e4f84ba6d64da@mail.gmail.com>
	<9e272eff0601311433p3cb6dac5l1f036d31d00431f5@mail.gmail.com>
Message-ID: <6fd0654b0602010706i3122f6c3g6f8becaef7bbc4e2@mail.gmail.com>

On 1/31/06, Sean Carley <seanacarley at gmail.com> wrote:
> Pat,
>
>  I would love to do extensions to RubyInline but they feel slightly over our
> heads at the moment.  Lets put them on the schedule for next project.

Sounds like a winner.   zenspider mentioned that he wants to refactor
RubyInline to make it easier to extend anyway ... so waiting might be a
good thing.

>
>  Anything in the Games space sounds like a lot of fun.  Any ideas in
> particular?

Nothing sticks out.   Anything looks especially fun to anyone else?

>
>  Do you have languages for the other libraries you mentioned?

Everything but RubyInline was from Perl ... more details live at
search.cpan.org

>
>  Sean
>
>
> On 1/31/06, pat eyler <pat.eyler at gmail.com> wrote:
> > (Second response, sorry)
> >
> > I do have a couple of ideas that might be fun to work on though ...
> >
> > RubyInline::Fortran
> > RubyInline::Haskell
> > RubyInline::OCaml   (zenspider would like to see this one ... maybe he'd
> help)
> >    (not really ports, but extensions of RubyInline, which is the
> > analog of the scary capable Perl::Inline)
> >
> > Gedcom
> > a bunch of libraries in the Games:: space
> > Astro::Sunrise (or maybe the emacs calendar conversion stuff)
> > Quantum::Superpositions
> >
>
> _______________________________________________
> Therps-discuss mailing list
> Therps-discuss at rubyforge.org
> http://rubyforge.org/mailman/listinfo/therps-discuss
>
>


--
thanks,
-pate
-------------------------


From pat.eyler at gmail.com  Fri Feb 10 13:02:29 2006
From: pat.eyler at gmail.com (pat eyler)
Date: Fri, 10 Feb 2006 11:02:29 -0700
Subject: [RPS] New Project, when to start
Message-ID: <6fd0654b0602101002n77c2a7e3kf5dda62dbbb184ac@mail.gmail.com>

After some discussion, poking around, and general handwringing, it looks
like we're ready to get started with RWB (the Ruby Web Bench) as our
new project.   When do we want to get rolling?   We could start on the 13th,
or wait until March 1st.


--
thanks,
-pate
-------------------------


From pat.eyler at gmail.com  Wed Feb 15 12:05:47 2006
From: pat.eyler at gmail.com (pat eyler)
Date: Wed, 15 Feb 2006 10:05:47 -0700
Subject: [RPS] Ruby Web Bench
In-Reply-To: <d4e4955b0602141751x55429846m5fc6eb0b75b0f205@mail.gmail.com>
References: <d4e4955b0602141751x55429846m5fc6eb0b75b0f205@mail.gmail.com>
Message-ID: <6fd0654b0602150905g7b44bb81y91413b8d1fc5b870@mail.gmail.com>

On 2/14/06, Wilson Bilkovich <wilsonb at gmail.com> wrote:
> Unfortunately, the current CVS version and release gem totally don't
> work for me on Windows or Linux, particularly not when the "number of
> threads" and "number of requests" numbers start to get near each
> other.  The unit tests don't pass either.

Yeah, I think I broke something pretty badly in time for the last release.
When I look at my current tree I get:

pate at linux:~/lang/ruby/rwb> rake test
(in /home/pate/lang/ruby/rwb)
./test/test_rwb.rb:147: warning: method redefined; discarding old
test_get_timesLoaded suite
/usr/local/lib/ruby/gems/1.8/gems/rake-0.6.2/lib/rake/rake_test_loader
Started
....................................................
Finished in 0.451318 seconds.

52 tests, 95 assertions, 0 failures, 0 errors


can you show me the unit test results you get?

>
> I put it down, and started thinking about what I was looking for in a
> testing tool, and thought: "Huh.. it would be pretty cool to have a
> generator class for URLs (particularly Rails-style URLs), and multiple
> workers could self-register and communicate over DRb."
>
> Then I read the discussion you two had on the mailing list, and I saw
> that you were thinking along similar lines.


The Ruby Programming Shop has picked this up for our next project,
but we'd welcome anyone interested in working on it.  I think the first
step is to get all the tests passing again, and test coverage for un-
tested methods.  Then I'd like to work on extracting the http connection
stuff to a mixin (we can combine this with r43), and general refactoring.
Finally, I'd like to work on the move to multiple workers, and improved
scalability anbd performance.

If you have additional bugs/features you'd like to see, we can hash
them out in email, and track them using the RWB trackers on rubyforge.


>
> Any interest in picking this back up?  I'd like to see something like
> this, and I don't have any problem helping.
>
> Thanks,
> --Wilson.
>


--
thanks,
-pate
-------------------------


From pat.eyler at gmail.com  Wed Feb 15 14:24:39 2006
From: pat.eyler at gmail.com (pat eyler)
Date: Wed, 15 Feb 2006 12:24:39 -0700
Subject: [RPS] Thoughts about net/http and RWB (and r43)
Message-ID: <6fd0654b0602151124x70ea0f6esa6df0acf03b988db@mail.gmail.com>

>From an IM session with Sean ...

(12:10:06) pate: so, I was reading
http://kallokain.blogspot.com/2006/01/refactoring-extract-mixin.html
and thought we might be able to extract the http request code out of
r43 and rwb into a mixin (and add POST handling while we're at it) to
make it easy to handle proxied/non-proxied, GET/POST, etc. functions
across these (and similar) libraries

(12:10:28) pate: (but maybe net/http is already extracted enough ...
I'm not sure)

(12:10:47) Sean Carley: I will give it a gander.

(12:11:05) pate: I'd like to hear your thoughts

(12:14:07) Sean Carley: My first thought is that the GET handling is not
that clean in r43.  It would probably benefit from being made clean enough
to support the refactoring.

(12:15:47) Sean Carley: As far as second thoughts go, net/http could
certainly be friendlier and a simplified version in a mixin would be a good
thing but we would have to watch out for metaphor seep.

(12:15:56) ***pate nods

(12:16:55) pate: As I was reading the article, I was struck by how similar
the things I was doing in rwb and r43 were ... and that started to smell
like an unextracted mixin

(12:18:08) Sean Carley: The idea is good.  The implementation will be
more work than described in your link.  Weshould try to tackle it during
the rwb project.

(12:18:30) pate: sounds good.

(12:18:59) pate: we (or I, or something) can work on moving it back into
r43 to verify that it makes sense

(12:19:05) Sean Carley: You and I can make use of and reference to r43
for how to extract it but it should probably be extracted from rwb then we
can move that implementation back to r43.

(12:19:12) ***pate nods


--
thanks,
-pate

-------------------------


From pat.eyler at gmail.com  Thu Feb 16 21:38:03 2006
From: pat.eyler at gmail.com (pat eyler)
Date: Thu, 16 Feb 2006 19:38:03 -0700
Subject: [RPS] Ruby Web Bench
In-Reply-To: <d4e4955b0602150943q5a4311b7m2645e8e6cb2ca5@mail.gmail.com>
References: <d4e4955b0602141751x55429846m5fc6eb0b75b0f205@mail.gmail.com>
	<6fd0654b0602150905g7b44bb81y91413b8d1fc5b870@mail.gmail.com>
	<d4e4955b0602150943q5a4311b7m2645e8e6cb2ca5@mail.gmail.com>
Message-ID: <6fd0654b0602161838j6d017e0ax294faae744c9b0cf@mail.gmail.com>

I think I know what's going on.  Are you using Ruby 1.8.4?  I've
been using 1.8.2, and things work wonderfully.  When I use 1.8.4
I get the same error as shown with your script.

On 2/15/06, Wilson Bilkovich <wilsonb at gmail.com> wrote:
> Here's what I get from the latest tarball:
> E:\Projects\rwb>rake test
> (in E:/Projects/rwb)
> ./test/test_rwb.rb:147: warning: method redefined; discarding old test_get_times
>
> Loaded suite c:/ruby/lib/ruby/gems/1.8/gems/rake-0.7.0/lib/rake/rake_test_loader
>
> Started
> ....F........................................F......
> Finished in 0.063 seconds.
>
>   1) Failure:
> test_get_url(TestRWB::TestBuilder) [./test/test_rwb.rb:49]:
> <"http://localhost"> expected but was
> <"http://localhost/foo">.
>
>   2) Failure:
> test_to_url(TestRWB::TestUrlGroup) [./test/test_rwb.rb:242]:
> <"http://www.example.com/search?baz"> expected but was
> <"http://www.example.com/search?foo">.

I think these both have to do with randomization.   These tests
are both looking for randomized data to come back with a set
answer based on a specific seed.


>
> 52 tests, 92 assertions, 2 failures, 0 errors
> rake aborted!
> Command failed with status (1): [c:/ruby/bin/ruby -w -Ilib "c:/ruby/lib/rub...]
>
> (See full trace by running task with --trace)
>
>
> ..and via a simple test:
> E:\Projects\rwb>ruby -Ilib wilson_test.rb
> 10 http://google.com
> 10 http://google.com
> c:/ruby/lib/ruby/1.8/net/http.rb:1441:in `initialize': undefined method `empty?'
>  for #<URI::HTTP:0x143e43c URL:http://google.com> (NoMethodError)
>         from ./lib/rwb.rb:85:in `join'
>         from ./lib/rwb.rb:85:in `run'
>         from ./lib/rwb.rb:85:in `run'
>         from wilson_test.rb:13
>
>
> The code for which is:
> require 'rwb'
> urls = RWB::Builder.new
> urls.add_url(10, "http://google.com")
> tests = RWB::Runner.new(urls, 10, 2)
> tests.run
> tests.report_header
> tests.report_overall [0.5, 0.9, 0.99, 0.999]
>


--
thanks,
-pate
-------------------------


From pat.eyler at gmail.com  Sat Feb 25 16:31:31 2006
From: pat.eyler at gmail.com (pat eyler)
Date: Sat, 25 Feb 2006 14:31:31 -0700
Subject: [RPS] starting up in March
Message-ID: <6fd0654b0602251331h5924af5dx722a3f2b4b7c62a4@mail.gmail.com>

We're going to get rolling on RWB in March.   I'm working on a
task list now.  If you're interested in being involved, please
respond so I can get a feel for how much work is reasonable.


--
thanks,
-pate
-------------------------


From seanacarley at gmail.com  Sun Feb 26 06:07:23 2006
From: seanacarley at gmail.com (Sean Carley)
Date: Sun, 26 Feb 2006 05:07:23 -0600
Subject: [RPS] starting up in March
In-Reply-To: <6fd0654b0602251331h5924af5dx722a3f2b4b7c62a4@mail.gmail.com>
References: <6fd0654b0602251331h5924af5dx722a3f2b4b7c62a4@mail.gmail.com>
Message-ID: <9e272eff0602260307y2a550548k86dee2eb355d8e05@mail.gmail.com>

I am interested.

Sean

On 2/25/06, pat eyler <pat.eyler at gmail.com> wrote:
>
> We're going to get rolling on RWB in March.   I'm working on a
> task list now.  If you're interested in being involved, please
> respond so I can get a feel for how much work is reasonable.
>
>
> --
> thanks,
> -pate
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/therps-discuss/attachments/20060226/12af849f/attachment.htm

From ret1 at acm.org  Mon Feb 27 08:03:19 2006
From: ret1 at acm.org (Rob Thornton)
Date: Mon, 27 Feb 2006 07:03:19 -0600
Subject: [RPS] starting up in March
In-Reply-To: <9e272eff0602260307y2a550548k86dee2eb355d8e05@mail.gmail.com>
References: <6fd0654b0602251331h5924af5dx722a3f2b4b7c62a4@mail.gmail.com>
	<9e272eff0602260307y2a550548k86dee2eb355d8e05@mail.gmail.com>
Message-ID: <f53d80d90602270503n5d7d61afwaedcd246a6f22c19@mail.gmail.com>

I am interested as well.

Rob

On 2/26/06, Sean Carley <seanacarley at gmail.com> wrote:
>
> I am interested.
>
> Sean
>
> On 2/25/06, pat eyler <pat.eyler at gmail.com> wrote:
> >
> > We're going to get rolling on RWB in March.   I'm working on a
> > task list now.  If you're interested in being involved, please
> > respond so I can get a feel for how much work is reasonable.
> >
> >
> > --
> > thanks,
> > -pate
> >
>
>
> _______________________________________________
> Therps-discuss mailing list
> Therps-discuss at rubyforge.org
> http://rubyforge.org/mailman/listinfo/therps-discuss
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/therps-discuss/attachments/20060227/e41862e7/attachment-0001.htm

From pat.eyler at gmail.com  Tue Feb 28 11:20:49 2006
From: pat.eyler at gmail.com (pat eyler)
Date: Tue, 28 Feb 2006 09:20:49 -0700
Subject: [RPS] RWB -- tasklist
Message-ID: <6fd0654b0602280820j3b7ca4eds1fc4ccf46855262@mail.gmail.com>

There are several things that RWB really needs.  I'd like to toss out an
unordered list, and ask people what they think is doable over the next
month or two (hopefully in 2 week iterations).

add POST and HEAD functionality
improve test coverage[0]
improve documentation
move from a threaded model to a multiple process
   model using DRb to communicate
improve URLGroups to handle more intricate URL
   patterns
refactoring (move HTTP handling code into a module (or
   improve uri-open to do the things we need)
add more base reports
build up some integration with Ruport/Gruff


I'm sure there are things I've overlooked.   What do you
think RWB is missing?


--
thanks,
-pate
-------------------------

[0]  coverage report from rcov
Name  	               Total lines  	Lines of code  	Total coverage 
	Code coverage
TOTAL  	                 1050   	 1050   	       81.4%                
           79.2%
./lib/rwb.rb 	            135 	     135 	          55.6%              
              57.4%
./lib/rwb/report.rb 	 100 	          100 	               34.0%        
                    27.9%
./lib/rwb/results.rb 	 164 	          164 	               90.2%       
                     89.0%
./lib/rwb/url.rb 	   126 	            126 	        100.0%             
             100.0%
./lib/rwb/warmup.rb 	48 	          48 	              16.7%            
                11.4%
./test/test_rwb.rb 	 477              477 	               97.3%       
                     96.9%