Hello all.
Noticed today that this code:
class CreateCache < ActiveRecord::Migration
def self.up
create_table :cache do |t|
t.bigint :name
end
end
end
in rails 2.3.x *silently* ignores the bigint setting [!]
Might be good to at least warn or raise an exception even better
Cheer
> What are the the other ways to reduce RAM requirement ?
You probably need to figure out what is eating all that RAM.
memprof might help.
-r
--
You received this message because you are subscribed to the Google Groups "Ruby
on Rails: Core" group.
To post to this group, send email to rubyonrai
>> require 'active_support'
NameError: uninitialized constant ActiveSupport::Dependencies::Mutex
from C:/installs/ruby-1.8.6-p388-i386-mingw32/lib/ruby/gems/
1.8/gems/activesupport-2.3.8/lib/active_support/dependencies.rb:55
from C:/installs/ruby-1.8.6-p388-i386-mingw32/lib/ruby/
si
> > Currently rails 2.3.8 and 3.0 all force a full loading of full
> > rubygems.
>
> Could you explain how?
If I remember correctly, bundler does something like
Gem::Dependency # force loading of full rubygems
Though I should look more closely at that one, since I was also under
the assumption t
Situation:
Currently rails 2.3.8 and 3.0 all force a full loading of full
rubygems.
Unfortunately this adds an extra 10MB of RSS to the application,
making servers slower because of GC. Typically one wants the
convenience of full rubygems, but not always.
Suggestion: only lazy load full rubygem
On Jul 3, 2:11 pm, Yehuda Katz wrote:
> Unfortunately, simply requiring the files in a mutex won't work. Consider
> the following simplified example:
>
> Thread.new do
> # Person is already created, so the const_missing will not fire (or
> autoload, etc.)
> # but say_hello is not yet defined
> I don't think this is necessary. Currently the reloader relies on the usage
> of require_or_load. Wouldn't it be possible to check at this point whether
> the file changed or not? That way watching the tree would not be needed.
>
> Also, this is somewhat part of my RSoC project.
Glad to hear t
I am under the impression that if config.threadsafe! was set that it
preloaded all files in the require paths, instead of using autoload.
Suggestion: use autoload when threadsafe!, but with a wrapper, like
autoload :Command,'thin/command'
becomes
autoload :Command, 'wrappers/
file_t
Request (unless rails already does this...)
Currently with active_support it reloads the class chain at the
beginning of each incoming HTTP request.
Suggestion: reload the class chain only if a file in a lib path *has
changed*
Thoughts?
-r
--
You received this message because you are subscribed
> Looking at the profile trees, the only one difference that stands out
> is in 2.3.5, 3.11% of the time is spent in
> ActiveRecord::ConnectionAdapters::ConnectionHandler#retrieve_connection_poo l
> which doesn't appear in 2.1.2.
that's not 40% worth...what else is different speed-wise/profile wis
Situation:
At least on my box, rails has the outside potential of making gems
load in the wrong order on 1.9
ex:
[first install test-unit gem v 2.0.6, 1.2.3]
gem 'rails', '= 2.3.5'
$LOAD_PATH.uniq! # active_support (?) does this internally
# now load a gem:
gem 'test-unit', '= 1.2.3'
As a note, currently in windows with 1.8.6
D:\dev\test2>ruby script\dbconsole
d:/installs/ruby_msvc_rc2/lib/ruby/gems/1.8/gems/rails-2.3.3/lib/
commands/dbconsole.rb:61:in `exec': No such file or directory -
mysql.exe (Errno::ENOENT)
This is caused by a bug in ruby 1.8.x on doze where
exec("comm
Currently if I do a gem fetch rails, gem unpack xxx.gem
cd xxx
rake rdoc
I get this:
Generating HTML...
Could not find HTML template '../doc/template/horo'
and the doc file remains empty. Is this expected?
Thank you.
=r
--~--~-~--~~~---~--~~
You received this me
> This is the first patch I've submitted to the Rails core and I would
> love to hear any feedback.
Appears it was accepted.
=r
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups "Ruby
on Rails: Core" group.
To post to
Thought I'd point out this gem
http://allgems.faithpromotingstories.org/gems/enumerable-extra/0.1.0
that I ran into the other day. It has some interesting ideas that
might be useful to incorporate into rails core since they would
benefit the whole community.
Just thinking out loud.
Thanks!
=r
-
> It looks like a domain squatter grabbed rubyonrails.org -- anybody
> know what's going on there?
>
> http://reports.internic.net/cgi/whois?whois_nic=rubyonrails.org&type=...
Interesting. For me I get the spam page on wiki.rubyonrails.org but
not for the main site--guess it's a DNS propagation
> > 3) don't regenerate query plans every time [cache them if possible].
>
> > 5) Avoid re-generating sql escaped copies of each column name for
> > tables. Cache the sql escaped column names.
>
> > 6) LRU cache all sql-escaped data [?]
>
> I attempted this 2-3 years ago with my work on prepared
Here are a few thoughts I had a year or so backy for some potential
low hanging fruit for speeding rails up [if anybody ever wants to
implement them :)
Forgive the naivety of these, they're mostly just suggestions from
someone not all that familiar with the core code.
1) wrap tests in transactio
k I'm going to punt on this and offer a $50 bounty if anybody can come
up with the following:
A better colored rails error output message that has code folding for
the code lines surrounding each line of backtrace, and, if it was run
with --debugger, has full parameters for each call [possible--s
> > Reloading isn'tthreadsafeand *can't* be due to the way constant
> > definition works in ruby. Why are you doing it in production mode?
> > allow_concurrency should be false for development mode? Is itnot?
I'm curious about the "constant definition not being threadsafe" -- is
this a thread sa
Here is an example of what the current patch does.
Any feedback?
-=R
http://wilkboardonline.com/roger/latest.html
> A nicer error page would certainly be great to have, and this is a
> good start. Perhaps the easiest way for people to give feedback would
> be to stick up some static HTML somewhe
Hi. I submitted a patch [1] which adds some coloration to
actionpack-2.1.0/lib/action_controller/templates/rescues/layout.erb
and was wondering if there are any other places besides it that could
also use improved error reporting/coloration.
Thoughts?
Thanks!
-=Roger Pack
[1]
http://rails.li
It seems that by default the Mysql adapter uses
instance.query_with_result = true [the default] for its queries.
Is there any special reason for that?
Thanks!
-=R
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups "Ruby
23 matches
Mail list logo