From mghaught at gmail.com  Wed Sep  5 23:18:57 2007
From: mghaught at gmail.com (Marty Haught)
Date: Wed, 5 Sep 2007 21:18:57 -0600
Subject: [Boulder Ruby Group] BRG - September Meeting - Sep 19th
Message-ID: <57f29e620709052018r4c766c48i42cb5ab9039a68e1@mail.gmail.com>

The Boulder Ruby group will be holding our monthly meeting in
September on the 19th at 6:30.  I'm still in the process of lining up
speakers for September and October.  If any of you have something
you'd like to present on or share, please contact me offlist.  Check
our website for the meeting details as we get closer.  I'll update it
first.  I will also send out the usual last minute reminder a few days
before the meeting with the details as well.

Cheers,
Marty

From mghaught at gmail.com  Fri Sep  7 13:49:06 2007
From: mghaught at gmail.com (Marty Haught)
Date: Fri, 7 Sep 2007 11:49:06 -0600
Subject: [Boulder Ruby Group] RubyConf 2007 registration open
Message-ID: <57f29e620709071049q216fe7bak1ea1cd85affb8f38@mail.gmail.com>

Shared by Chad it looks like registration is open:

http://www.regonline.com/Checkin.asp?EventId=145820

Cheers,
Marty

From tim.pease at gmail.com  Fri Sep  7 15:23:40 2007
From: tim.pease at gmail.com (Tim Pease)
Date: Fri, 7 Sep 2007 13:23:40 -0600
Subject: [Boulder Ruby Group] RubyConf 2007 registration open
In-Reply-To: <57f29e620709071049q216fe7bak1ea1cd85affb8f38@mail.gmail.com>
References: <57f29e620709071049q216fe7bak1ea1cd85affb8f38@mail.gmail.com>
Message-ID: <69f66b790709071223w61b66940oabf6ae42fac5ede7@mail.gmail.com>

On 9/7/07, Marty Haught <mghaught at gmail.com> wrote:
> Shared by Chad it looks like registration is open:
>
> http://www.regonline.com/Checkin.asp?EventId=145820
>

Is anyone planning on attending and looking to share hotel accommodations?

TwP

From tony at clickcaster.com  Fri Sep 14 00:12:39 2007
From: tony at clickcaster.com (Tony Arcieri)
Date: Thu, 13 Sep 2007 22:12:39 -0600
Subject: [Boulder Ruby Group] Actor model in Ruby 1.9
Message-ID: <c7e6b2b00709132112n9fb979ck450229169cb23991@mail.gmail.com>

I did a simple implementation of the Actor model (i.e. Erlang-style
concurrency) using Ruby 1.9's Fibers mechanism.  You can check it out here
(if you have Ruby 1.9 installed):

http://pastie.caboo.se/97050

Fibers are coroutines, which means they're like procs, except at any point
they can call Fiber.yield and return control back to the process that
started them.  The controlling process can resume them with Fiber.resume

This implementation adds a mailbox to a Fiber, and an Actor.receive method
which runs a filter across the mailbox, wrapping Fiber.yield and calling it
if the filter fails to match any of the messages in the mailbox.

Here's an example of how to use it:

http://pastie.caboo.se/pastes/97049

The basic idea is that you spawn Actors from the main process.  Each Actor
has an associated mailbox, and you can asynchronously send it messages (in
this implementation, using <<).  When you spawn an actor, you pass it a
block to run.  This actually runs within a fiber.  In an Actor, you can call
the Actor.receive method to read messages from the mailbox.  You specify a
filterset for reading the messages: either an argument to be directly
compared with ===, or a proc that returns true on a match.

A run queue is maintained which contains every single Actor which has
received a message.  When an Actor receives a message, its Fiber is resumed,
and then the specified filter set is checked against the mailbox.  If
there's a match, the corresponding block is invoked, otherwise the Fiber
yields until its gets another message.  If all our Fibers are sleeping,
control is returned back to the main program until it sends a message to one
of the actors.

Actors can make other Actors with the Actor.spawn method and can also send
messages to other Actors, provided they have a handle to the Actor object
they want to send a message to.

If you dump the implementation into actor.rb then paste the example code
into irb, you can play around with it.  Just use << to send a message to any
Actor object.  Messages are totally arbitrary and get dumped into the
Actor's associated mailbox:

irb(main):040:0> actor2 << :cat
I hate cats... you take it!
I love cats
My buddy got a cat
=> :cat

Here we have messages being sent from actor2 -> actor1 -> buddy.  actor2
receives the :cat message and forwards it onto actor1.  actor1 receives it
and sends a :gotacat message to buddy.  Each line was printed out by a
different fiber.

Here's an example of the Proc-based pattern matching:

irb(main):041:0> actor2 << "Foobar"
I hate strings... you take it!
I got a string: Foobar
=> "Foobar"

Here actor2's proc-based filter detected it received a string, and decided
to forward it back to actor1.  actor1's filter was set up to print out
strings it received using a proc to match them.

-- 
Tony Arcieri
ClickCaster, Inc.
tony at clickcaster.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/bdrg-members/attachments/20070913/f4922489/attachment.html 

From mghaught at gmail.com  Mon Sep 17 19:05:00 2007
From: mghaught at gmail.com (Marty Haught)
Date: Mon, 17 Sep 2007 17:05:00 -0600
Subject: [Boulder Ruby Group] BRG - September Meeting - Sep 19th Reminder
Message-ID: <57f29e620709171605x5b6e020dsf02a4c255fd4db9b@mail.gmail.com>

Our September meeting will be held this Wednesday, Sept 19th at 6:30pm
(directions below). Nathan Witmer will handle our first presentation
on Rspec.  We'll break for 20-30 minutes to mingle.  Afterwards
Jameson Watts will present the second talk on Radius MS ? RESTful
texting. Radius MS is a RESTful api for text message data relay, and
infinite depth menuing. As a substitute for WAP, Radius provides a
stateful request/response for SMS interaction with a server. Should be
an interesting set of talks. As always we have a good time conversing
as a group. After the meeting, we will migrate over to the pub for
some refreshments and continued good company. Look forward to seeing
you all there.

Cheers,
Marty Haught
http://boulderruby.org

Directions:

Collective Intellect
1601 Pearl St, Suite 200
Boulder, CO 80302

The office is above the restaurants Aji & Leaf. The main entrance is
on 16th Street, just North to of the Leaf restaurant. Come up the
stairs to the second floor.

URL to google maps: http://rubyurl.com/112

From ara.t.howard at gmail.com  Wed Sep 19 19:16:25 2007
From: ara.t.howard at gmail.com (ara.t.howard)
Date: Wed, 19 Sep 2007 17:16:25 -0600
Subject: [Boulder Ruby Group] rfc
Message-ID: <01195B4F-39B6-4380-A0F8-F536F5340B66@gmail.com>



unless defined? SaneTransactions

   <<-README
     NAME

       sane_transactions

     SYNOPSIS

       the standard activerecord transactions are not sane.  consider  
the following

         ActiveRecord::Base.connection.transaction do
           Model.create!
           raise
         end

       in this case the Model creation is NOT rolled back.  the  
reason can be
       summarized thusly: ActiveRecord never starts a transaction  
when one is open,
       but commiting a transaction always occurs.  this means that,  
no matter how
       hard you try, one errant save|destroy|create from deep in the  
bowels of a
       plugin will unravel your carefully laid plans - throwing data  
integrity out
       the window, along with part of your salary.

       for the most part people simply ignore this fact in rails, and  
live with
       the possibility of inconsistent data after a failed complex  
controller
       action.

       sane_transactions address this issue, and more, in a small,  
furry, rails
       munging way.

     USAGE


       * POLS transactions

           require 'sane_transactions'

           ActiveRecord::Base.connection.transaction do
             Model.create!
             raise
           end

           # Model creation is rolled back


       * nested transactions work, only the top level transaction  
executes sql

           require 'sane_transactions'

           ActiveRecord::Base.connection.transaction do
             Model.create!

             ActiveRecord::Base.connection.transaction do
               ActiveRecord::Base.connection.transaction do
                 ActiveRecord::Base.connection.transaction do
                   raise
                 end
               end
             end
           end

           # Model creation is rolled back


         * shortcuts are given, db code is important so certain  
methods are
           available at the top level

             require 'sane_transactions'

             transaction do
               Model.create!
               raise
             end

             # ahhhhhh


         * throw/catch based switches enable you to bail out and  
commit or
           rollback a transaction early and from any depth

             require 'sane_transactions'

             transaction do
               Model.create!
               rollback!
             end

             # Model creation is rolled back


         * a commit switch is provided too

             require 'sane_transactions'

             transaction do
               a = Model.create!
               commit!
               b = Model.create!
             end

             # a IS created, b IS NOT


         * savepoints are implemented, and can be individually rolled  
back

             require 'sane_transactions'

             transaction do
               savepoint do
                 a = Model.create!
                 rollback! :savepoint
               end

               b = Model.create!
             end

             # a IS NOT created, b IS

     INSTALL

       - put this file into RAILS_ROOT/lib

       - add "require 'sane_transactions'" to RAILS_ROOT/environment.rb

     DISCLAIMER

       this is pre-pre-pre-pre-alpha code

     AUTHOR

       ara.t.howard [[at]] gmail [[dot]] com
   README


   class Object
     def transaction *argv, &block
       ActiveRecord::Base.connection.transaction *argv, &block
     end
     def transaction? *argv, &block
       ActiveRecord::Base.connection.transaction? *argv, &block
     end
     def savepoint *argv, &block
       ActiveRecord::Base.connection.savepoint *argv, &block
     end
     def rollback! *argv, &block
       ActiveRecord::Base.connection.rollback! *argv, &block
     end
     def commit! *argv, &block
       ActiveRecord::Base.connection.commit! *argv, &block
     end
   end

   class Module
     def thread_safe_attribute name, options = {}
       options.to_options!
       prefix = options[:prefix]
       prefix = "#{ prefix.name.underscore }_" if Module === prefix
       key = "#{ prefix }#{ name }".inspect
       code = <<-code
         def #{ name } *argv, &block
           if argv.empty?
             Thread.current[#{ key }]
           else
             if block
               value = Thread.current[#{ key }]
               begin
                 Thread.current[#{ key }] = argv.first
                 block.call
               ensure
                 Thread.current[#{ key }] = value
               end
             else
               Thread.current[#{ key }] = argv.first
             end
           end
         end
         def #{ name }= value
           #{ name } value
         end
         def #{ name }?
           #{ name }
         end
       code
       module_eval code
     end
   end

   module ActiveRecord
     module Transactions
       class Savepoint < ::String
         class << self
           thread_safe_attribute 'initial', :prefix =>  
ActiveRecord::Transactions::Savepoint
         end

         def initialize initial = self.class.initial
           reset! initial
         end

         def next
           succ!
         end

         def reset! initial = self.class.initial
           replace initial
         end

         initial "rails_savepoint_0"
       end

       class << self
         thread_safe_attribute 'using', :prefix =>  
ActiveRecord::Transactions
         thread_safe_attribute 'are_active', :prefix =>  
ActiveRecord::Transactions
         thread_safe_attribute 'savepoint', :prefix =>  
ActiveRecord::Transactions
       end

       using :transaction
       are_active false
       savepoint Savepoint.new
     end

     module ConnectionAdapters
       module DatabaseStatements
         def transaction *argv, &block
           method = "transaction_using_# 
{ ActiveRecord::Transactions.using }"
           send method, *argv, &block
         end

         def transaction?
           ActiveRecord::Transactions.are_active?
         end

         def transaction_using_transaction *ignored
           return yield if ActiveRecord::Transactions.are_active?

           finish = nil
           returned = nil
           begin
             ActiveRecord::Transactions.are_active true
             ActiveRecord::Transactions.savepoint.reset!
             begin_db_transaction
             finish = catch(:transaction){ returned = yield; nil }
             returned
           ensure
             ActiveRecord::Transactions.are_active false
             if finish
               finish == :rollback ? rollback_db_transaction :  
commit_db_transaction
             else
               $! ? rollback_db_transaction : commit_db_transaction
             end
           end
         end

         def transaction_using_savepoint *ignored
           raise TransactionError, 'no transaction in effect' unless  
ActiveRecord::Transactions.are_active?

           name = nil
           finish = nil
           returned = nil
           begin
             name = begin_db_savepoint
             finish = catch(:savepoint){ returned = yield; nil }
             returned
           ensure
             if finish
               rollback_db_savepoint name if finish == :rollback and  
name
             else
               rollback_db_savepoint name if $! and name
             end
           end
         end

         def savepoint
           raise TransactionError, 'no transaction in effect' unless  
ActiveRecord::Transactions.are_active?
           ActiveRecord::Transactions.using :savepoint do
             transaction{ yield }
           end
         end

         def rollback! target = :transaction
           raise TransactionError, 'no transaction in effect' unless  
ActiveRecord::Transactions.are_active?
           throw target, :rollback
         end

         def rollback_savepoint!
           rollback! :savepoint
         end

         def commit! target = :transaction
           raise TransactionError, 'no transaction in effect' unless  
ActiveRecord::Transactions.are_active?
           throw target, :commit
         end

         def commit! target = :transaction
           raise TransactionError, 'no transaction in effect' unless  
ActiveRecord::Transactions.are_active?
           throw target, :commit
         end

         def begin_db_savepoint name =  
ActiveRecord::Transactions.savepoint.next
           execute "SAVEPOINT #{ name }"
           name
         end

         def rollback_db_savepoint name
           execute "ROLLBACK TO SAVEPOINT #{ name }"
           name
         end

         def release_db_savepoint name
           execute "RELEASE SAVEPOINT #{ name }"
           name
         end
       end
     end
   end

   SaneTransactions = 42
end




a @ http://drawohara.com/
--
we can deny everything, except that we have the possibility of being  
better. simply reflect on that.
h.h. the 14th dalai lama




From ara.t.howard at gmail.com  Thu Sep 20 01:30:19 2007
From: ara.t.howard at gmail.com (ara.t.howard)
Date: Wed, 19 Sep 2007 23:30:19 -0600
Subject: [Boulder Ruby Group] rfc
In-Reply-To: <7968d7490709192204te9c1970m8f350fc2071826cd@mail.gmail.com>
References: <01195B4F-39B6-4380-A0F8-F536F5340B66@gmail.com>
	<7968d7490709192204te9c1970m8f350fc2071826cd@mail.gmail.com>
Message-ID: <6FC8B071-066E-49EF-BBED-177E58148CD2@gmail.com>


On Sep 19, 2007, at 11:04 PM, Lee Marlow wrote:

> This definitely looks more sane and is what I would expect.  I wonder
> if you submitted it as a patch to rails whether it could make it into
> 2.0.

i haven't looked at the 2.0 source, but i imagine quite a bit has  
changed - probably too much for this patch...


> Is this not in rails already because the savepoints and nested
> transactions aren't standard across most databases?

well - it doesn't really support nested transactions per-se.  that is  
to say that rails wraps alot of stuff in transactions but, if you are  
already in one, it's kind of moot.  so these transactions are really  
ruby transactions - where the many of them stack into only one db  
transaction.  as far as savepoints it doesn't really matter if they  
are implemented or not.  in sane_transactions you can simply not use  
them for unsupported dbs...  but you are right that only pg/oracle  
and a few others support them.

cheers.

a @ http://drawohara.com/
--
we can deny everything, except that we have the possibility of being  
better. simply reflect on that.
h.h. the 14th dalai lama




From ara.t.howard at gmail.com  Thu Sep 20 11:57:17 2007
From: ara.t.howard at gmail.com (ara.t.howard)
Date: Thu, 20 Sep 2007 09:57:17 -0600
Subject: [Boulder Ruby Group] rfc
In-Reply-To: <6efbd9b70709191647l1a5826bauba43c57c78dd25c6@mail.gmail.com>
References: <01195B4F-39B6-4380-A0F8-F536F5340B66@gmail.com>
	<6efbd9b70709191647l1a5826bauba43c57c78dd25c6@mail.gmail.com>
Message-ID: <DDBFC8DB-D9AA-4309-BCC6-79685B497C4F@gmail.com>


On Sep 19, 2007, at 5:47 PM, Chris Riddoch wrote:

> On 9/19/07, ara.t.howard <ara.t.howard at gmail.com> wrote:
>> unless defined? SaneTransactions
>
> Thanks for this! I was just reading up on Rails' transactions for a
> project I'm doing, and it's very helpful to know how things are
> broken... doubly so with a fix, too.
>
> Is there any chance of getting this solution into ActiveRecord?
>
> -- epistemological humility
>   Chris Riddoch


dunno.  i'm not 100% there isn't another workaround - but i sure  
haven't found it yet.  there are also some similar patches proposed  
but, afaikt, none that let transactions and savepoints mix so easily.

i've sent it off to marcel molina (rails core) for review - i'll get  
his input and, if positive, submit to rails-core.

thanks alot for the feedback!

ps.  if anyone can drop this into their projects and see if things  
work it'd be great.  there are few things with fixtures that *should*  
break if you've done the standard workarounds but, other than that,  
things should work unchanged.  if not send me a stack trace!


a @ http://drawohara.com/
--
we can deny everything, except that we have the possibility of being  
better. simply reflect on that.
h.h. the 14th dalai lama




From nathan at teamjunkshow.com  Tue Sep 25 21:32:16 2007
From: nathan at teamjunkshow.com (Nathan Witmer)
Date: Tue, 25 Sep 2007 19:32:16 -0600
Subject: [Boulder Ruby Group] RSpec presentation slides & notes
Message-ID: <29377135-5DE5-4FAC-9AC0-6DAC1E1B5C50@teamjunkshow.com>

Here's the slides and notes from my rspec presentation last week.  
http://notebook.otherward.net/assets/2007/9/26/rspec.pdf

Nathan

From kevwil at gmail.com  Wed Sep 26 11:02:43 2007
From: kevwil at gmail.com (Kevin Williams)
Date: Wed, 26 Sep 2007 09:02:43 -0600
Subject: [Boulder Ruby Group] RSpec presentation slides & notes
In-Reply-To: <29377135-5DE5-4FAC-9AC0-6DAC1E1B5C50@teamjunkshow.com>
References: <29377135-5DE5-4FAC-9AC0-6DAC1E1B5C50@teamjunkshow.com>
Message-ID: <683a886f0709260802udc72266s4832756fbdf08747@mail.gmail.com>

Does the group web site have a "post presentations here" page?

A common place to revisit these presentations would be cool.


On 9/25/07, Nathan Witmer <nathan at teamjunkshow.com> wrote:
> Here's the slides and notes from my rspec presentation last week.
> http://notebook.otherward.net/assets/2007/9/26/rspec.pdf
>
> Nathan
> _______________________________________________
> Bdrg-members mailing list
> Bdrg-members at rubyforge.org
> http://rubyforge.org/mailman/listinfo/bdrg-members
>


-- 
Cheers,

Kevin Williams
http://kevwil.com/
http://www.almostserio.us/
http://kevwil.jaiku.com/

From ara.t.howard at gmail.com  Wed Sep 26 11:37:11 2007
From: ara.t.howard at gmail.com (ara.t.howard)
Date: Wed, 26 Sep 2007 09:37:11 -0600
Subject: [Boulder Ruby Group] RSpec presentation slides & notes
In-Reply-To: <683a886f0709260802udc72266s4832756fbdf08747@mail.gmail.com>
References: <29377135-5DE5-4FAC-9AC0-6DAC1E1B5C50@teamjunkshow.com>
	<683a886f0709260802udc72266s4832756fbdf08747@mail.gmail.com>
Message-ID: <191DF4C9-272A-4DDC-A5E2-A2D50A3ED937@gmail.com>


On Sep 26, 2007, at 9:02 AM, Kevin Williams wrote:

> Does the group web site have a "post presentations here" page?
>
> A common place to revisit these presentations would be cool.

all mine are here - i'd be happy to host them somewhere central.   
maybe we should just put them on s3 in a bucket we all have the  
password too?

a @ http://drawohara.com/
--
we can deny everything, except that we have the possibility of being  
better. simply reflect on that.
h.h. the 14th dalai lama




From ara.t.howard at gmail.com  Sun Sep 30 12:02:43 2007
From: ara.t.howard at gmail.com (ara.t.howard)
Date: Sun, 30 Sep 2007 10:02:43 -0600
Subject: [Boulder Ruby Group] rails/ruby memory leak
Message-ID: <87C8ABE9-2EE5-4B3D-891E-CF7E3E30B80A@gmail.com>


hi all

i've been chasing my tail on a tiny memory leak for a few days.  i
isolated the cause to what amounts to a rails project with database
sessions enabled.  now i'm not sure that the leak is in rails, in fact
i think it's in ruby's cgi/session.rb but i want to gather a few data
points to confirm the leak i'm seeing.  if you have a little time,
check out

   http://drawohara.tumblr.com/post/13655303

or just grab the project from

   http://s3.amazonaws.com/drawohara.com.downloads/leak.tgz

and get it running on your box.  it should be *very* quick to do if
you have sqlite3 installed.  running it is also easy, but seeing the
leak can take a bit of time - 30 minutes to 6 hours depending on how
fast your box can hammer ./script/server.  anyhow, people can report
here or to me directly.  i'll summarize my findings and, if there is
in fact a leak in either ruby or rails - follow up with a bug report.

kind regards.

a @ http://drawohara.com/
--
we can deny everything, except that we have the possibility of being  
better. simply reflect on that.
h.h. the 14th dalai lama




From kevwil at gmail.com  Sun Sep 30 16:12:17 2007
From: kevwil at gmail.com (Kevin Williams)
Date: Sun, 30 Sep 2007 14:12:17 -0600
Subject: [Boulder Ruby Group] rails/ruby memory leak
In-Reply-To: <87C8ABE9-2EE5-4B3D-891E-CF7E3E30B80A@gmail.com>
References: <87C8ABE9-2EE5-4B3D-891E-CF7E3E30B80A@gmail.com>
Message-ID: <683a886f0709301312o7cd0fc31vce6387c5757c90a@mail.gmail.com>

I couldn't run it long enough to get what I would call 'conclusive
results' because my MBP seemed about to melt. :( There did seem to be
a slight increase, but not as much as in your tumblr post.

(same OS, same ruby version, same rails version)


On 9/30/07, ara.t.howard <ara.t.howard at gmail.com> wrote:
>
> hi all
>
> i've been chasing my tail on a tiny memory leak for a few days.  i
> isolated the cause to what amounts to a rails project with database
> sessions enabled.  now i'm not sure that the leak is in rails, in fact
> i think it's in ruby's cgi/session.rb but i want to gather a few data
> points to confirm the leak i'm seeing.  if you have a little time,
> check out
>
>    http://drawohara.tumblr.com/post/13655303
>
> or just grab the project from
>
>    http://s3.amazonaws.com/drawohara.com.downloads/leak.tgz
>
> and get it running on your box.  it should be *very* quick to do if
> you have sqlite3 installed.  running it is also easy, but seeing the
> leak can take a bit of time - 30 minutes to 6 hours depending on how
> fast your box can hammer ./script/server.  anyhow, people can report
> here or to me directly.  i'll summarize my findings and, if there is
> in fact a leak in either ruby or rails - follow up with a bug report.
>
> kind regards.
>
> a @ http://drawohara.com/
> --
> we can deny everything, except that we have the possibility of being
> better. simply reflect on that.
> h.h. the 14th dalai lama
>
>
>
> _______________________________________________
> Bdrg-members mailing list
> Bdrg-members at rubyforge.org
> http://rubyforge.org/mailman/listinfo/bdrg-members
>


-- 
Cheers,

Kevin Williams
http://kevwil.com/
http://www.almostserio.us/
http://kevwil.jaiku.com/

From tony at clickcaster.com  Sun Sep 30 17:56:06 2007
From: tony at clickcaster.com (Tony Arcieri)
Date: Sun, 30 Sep 2007 15:56:06 -0600
Subject: [Boulder Ruby Group] rails/ruby memory leak
In-Reply-To: <87C8ABE9-2EE5-4B3D-891E-CF7E3E30B80A@gmail.com>
References: <87C8ABE9-2EE5-4B3D-891E-CF7E3E30B80A@gmail.com>
Message-ID: <c7e6b2b00709301456k4cc38773w4af90f9c99b587f6@mail.gmail.com>

Are you using ObjectSpace to track leaks?

At one point Evan Weaver had a graph of an ObjectSpace profiler versus one
he wrote using direct heap instrumentation in C, and the ObjectSpace version
gave the appearance of phantom leaks which really weren't there...

Have you tried Bleak House?

http://blog.evanweaver.com/articles/2007/09/25/bleak_house-3-tells-you-your-leaks/

On 9/30/07, ara.t.howard <ara.t.howard at gmail.com> wrote:
>
>
> hi all
>
> i've been chasing my tail on a tiny memory leak for a few days.  i
> isolated the cause to what amounts to a rails project with database
> sessions enabled.  now i'm not sure that the leak is in rails, in fact
> i think it's in ruby's cgi/session.rb but i want to gather a few data
> points to confirm the leak i'm seeing.  if you have a little time,
> check out
>
>    http://drawohara.tumblr.com/post/13655303
>
> or just grab the project from
>
>    http://s3.amazonaws.com/drawohara.com.downloads/leak.tgz
>
> and get it running on your box.  it should be *very* quick to do if
> you have sqlite3 installed.  running it is also easy, but seeing the
> leak can take a bit of time - 30 minutes to 6 hours depending on how
> fast your box can hammer ./script/server.  anyhow, people can report
> here or to me directly.  i'll summarize my findings and, if there is
> in fact a leak in either ruby or rails - follow up with a bug report.
>
> kind regards.
>
> a @ http://drawohara.com/
> --
> we can deny everything, except that we have the possibility of being
> better. simply reflect on that.
> h.h. the 14th dalai lama
>
>
>
> _______________________________________________
> Bdrg-members mailing list
> Bdrg-members at rubyforge.org
> http://rubyforge.org/mailman/listinfo/bdrg-members
>



-- 
Tony Arcieri
ClickCaster, Inc.
tony at clickcaster.com
720-227-0129 ext. 202
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/bdrg-members/attachments/20070930/25037677/attachment.html 

From ara.t.howard at gmail.com  Sun Sep 30 21:29:01 2007
From: ara.t.howard at gmail.com (ara.t.howard)
Date: Sun, 30 Sep 2007 19:29:01 -0600
Subject: [Boulder Ruby Group] rails/ruby memory leak
In-Reply-To: <683a886f0709301312o7cd0fc31vce6387c5757c90a@mail.gmail.com>
References: <87C8ABE9-2EE5-4B3D-891E-CF7E3E30B80A@gmail.com>
	<683a886f0709301312o7cd0fc31vce6387c5757c90a@mail.gmail.com>
Message-ID: <C88810F4-E018-4828-B307-50F8DFE78402@gmail.com>


On Sep 30, 2007, at 2:12 PM, Kevin Williams wrote:

> I couldn't run it long enough to get what I would call 'conclusive
> results' because my MBP seemed about to melt. :( There did seem to be
> a slight increase, but not as much as in your tumblr post.
>
> (same OS, same ruby version, same rails version)

the results i posted was after around 24 of run time - it really is  
slow but, afaikt, a real leak.

cheers.

a @ http://drawohara.com/
--
we can deny everything, except that we have the possibility of being  
better. simply reflect on that.
h.h. the 14th dalai lama