From: jason marshall Date: 2012-09-12T05:38:28+09:00 Subject: Re: Looking for advice on how to report performance issues --20cf3079c0bc6a4b7204c9730e63 Content-Type: text/plain; charset=ISO-8859-1 I used that for a couple months (then lost the URL, thanks for reposting it). It took about 100-110s versus the 60 my coworkers got. Unfortunately it would also randomnly crash with a 10 digit error code while I was in the middle of testing. I was looking for something more stable. Currently I'm using https://raw.github.com/gist/3242245/5cdb932c0a544cfbb51fdee5759942d4d2c263f5/1-complete.diff which Luis pointed me at in a stackoverflow comment. But again I'm kind of a hard numbers guy, and I'd like to find a way to offer something more constructive than some anecdotal numbers. On Mon, Sep 10, 2012 at 2:52 PM, Jon Forums wrote: > jason marshall wrote in post #1075372: > > Hello, > > > > We have a Rails app running on Ruby 1.9.3, and a under Windows we're > > seeing > > a 20x time penalty in our RSpec tests. > > > > I have looked at some big patches from both Xavier and Luis and while > > they > > were very helpful, we're still seeing it take 7-8x as long to run the > > tests. That's the difference between a minute (very usable) and 7:30 > > (nearly unusable). > > Are your 7-8x as long numbers using the > `tcs-ruby193_require_fenix_gc_hash_20120527.7z` ready-to-use binary for > Windows from https://github.com/thecodeshop/ruby/wiki/Downloads > > This download includes performance improvements from Luis and others and > I'm curious to see how it runs in your Windows environment. > > https://groups.google.com/forum/?fromgroups=#!topic/thecodeshop/WVA4N2lJoM4 > > Jon > > --- > Fail fast. Fail often. Fail publicly. Learn. Adapt. Repeat. > http://thecodeshop.github.com | http://jonforums.github.com/ > twitter: @jonforums > > -- > Posted via http://www.ruby-forum.com/. > > -- - Jason --20cf3079c0bc6a4b7204c9730e63 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable I used that for a couple months (then lost the URL, thanks for reposting it= ). =A0It took about 100-110s versus the 60 my coworkers got. =A0

Unfortunately it would also randomnly crash with a 10 digit error = code while I was in the middle of testing. =A0 I was looking for something = more stable. =A0Currently I'm using
=A0
https:/= /raw.github.com/gist/3242245/5cdb932c0a544cfbb51fdee5759942d4d2c263f5/1-com= plete.diff

which Luis pointed me at in a stackoverflow comment.

But again I'm kind of a hard numbers guy, and I&= #39;d like to find a way to offer something more constructive than some ane= cdotal numbers.

On Mon, Sep 10, 2012 at 2:52 PM, Jon Forums = <lists@ruby-forum.com> wrote:
jason marshall wrote in post #1075372:
> Hello,
>
> We have a Rails app running on Ruby 1.9.3, and a under Windows we'= re
> seeing
> a 20x time penalty in our RSpec tests.
>
> I have looked at some big patches from both Xavier and Luis and while<= br> > they
> were very helpful, we're still seeing it take 7-8x as long to run = the
> tests. =A0That's the difference between a minute (very usable) and= 7:30
> (nearly unusable).

Are your 7-8x as long numbers using the
`tcs-ruby193_require_fenix_gc_hash_20120527.7z` ready-to-use binary for
Windows from https://github.com/thecodeshop/ruby/wiki/Downloads
This download includes performance improvements from Luis and others and I'm curious to see how it runs in your Windows environment.

https://groups.google.com/forum/?fromgroup= s=3D#!topic/thecodeshop/WVA4N2lJoM4

Jon

---
Fail fast. Fail often. Fail publicly. Learn. Adapt. Repeat.
http://thecodes= hop.github.com | http://jonforums.github.com/
twitter: @jonforums

--
Posted via http://= www.ruby-forum.com/.




-- - Jason
--20cf3079c0bc6a4b7204c9730e63--