Rack 1.0.1 maintenance release
Joshua Peek
test your apps and frameworks on that branch. It should be a fully
compatible bug fix release.
I went though most of the Lighthouse tickets and only found one that
really concerned 1.0.
"Rack::Utils.unescape problems in Ruby 1.9.1"
http://rack.lighthouseapp.com/projects/22435/tickets/48-rackutilsunescape-problems-in-ruby-191
http://rack.lighthouseapp.com/projects/22435-rack/milestones/current
Please reassign any 1.1 milestones tickets to 1.0 or let me know of
anything else master that you'd liked cherry-picked into rack-1.0.
Looking forward to Rack 1.1.
Here's what changed on rack-1.0 since the 1.0 tag:
commit 04a5e732a1aa708c0953c55984c17e01252d0aa9
Author: Julien Sanchez <julien....@gmail.com>
Date: Mon Aug 3 10:49:11 2009 -0500
Mongrel handler should use rack_input var
Signed-off-by: Joshua Peek <jo...@joshpeek.com>
:100644 100644 3a5ef32... 3e2cd8b... M lib/rack/handler/mongrel.rb
commit 488cf8091bb5d0262c784da40fa2b60d874e7303
Author: Joshua Peek <jo...@joshpeek.com>
Date: Mon Aug 3 10:33:18 2009 -0500
Return an empty string for Request#host if HTTP_HOST and
SERVER_NAME are both missing
:100644 100644 be534a8... 2f64bd6... M lib/rack/request.rb
:100644 100644 7e4be77... 74e2f00... M test/spec_rack_request.rb
commit 9a1e52b240be1e8cb62989739fba4d7e46681a1d
Author: Joshua Peek <jo...@joshpeek.com>
Date: Mon Aug 3 12:25:45 2009 -0500
Kill request instance memoization
Conflicts:
test/spec_rack_mock.rb
:100644 100644 0bff7af... be534a8... M lib/rack/request.rb
:100644 100644 b16306b... 9c392a2... M test/spec_rack_mock.rb
:100644 100644 dbfd5b1... 7e4be77... M test/spec_rack_request.rb
commit 663b8ec47eff66f75abd037736213232823838f6
Author: Joshua Peek <jo...@joshpeek.com>
Date: Thu Apr 30 19:02:57 2009 -0500
Bump remainder of rack.versions
:100644 100644 e38156c... f45f3d7... M lib/rack/handler/cgi.rb
:100644 100644 c32c186... 11e1fca... M lib/rack/handler/fastcgi.rb
:100644 100644 c65ba3e... 7231336... M lib/rack/handler/lsws.rb
:100644 100644 f0c0d58... 3a5ef32... M lib/rack/handler/mongrel.rb
:100644 100644 9495c66... 6c4932d... M lib/rack/handler/scgi.rb
:100644 100644 829e7d6... 2bdc83a... M lib/rack/handler/webrick.rb
:100644 100644 331f988... 818fabd... M test/spec_rack_cgi.rb
:100644 100644 66ab4f7... 69478de... M test/spec_rack_fastcgi.rb
:100644 100644 1da0af4... d73e884... M test/spec_rack_mongrel.rb
:100644 100644 a7393e9... 3e63ea6... M test/spec_rack_webrick.rb
commit 3c4fad7529074b1ac2b2de0fedea2726729f9107
Author: Ryan Tomayko <rtom...@gmail.com>
Date: Sat Apr 25 07:35:10 2009 -0700
fix rack.version in MockRequest
:100644 100644 70852da... 847763c... M lib/rack/mock.rb
commit 851782648fbbb141b1b56b73d8efc99efcac1b27
Author: Hongli Lai (Phusion) <hon...@phusion.nl>
Date: Wed Jun 17 17:33:21 2009 -0500
Document more clearly that rack.input must be opened in binary
mode, and enforce it in Rack::Lint.
Signed-off-by: Joshua Peek <jo...@joshpeek.com>
:100644 100644 bf2e978... 796807a... M lib/rack/lint.rb
:100644 100644 8c6419d... a227b56... M test/spec_rack_lint.rb
commit dfab30daae0f1748c0e93009583dc33dbc838a7c
Author: Hongli Lai (Phusion) <hon...@phusion.nl>
Date: Wed Jun 17 17:19:19 2009 -0500
Add unit tests for Rack::Util::HeaderHash#delete. [#54
state:resolved]
Signed-off-by: Joshua Peek <jo...@joshpeek.com>
:100644 100644 7dd5880... d1823cd... M test/spec_rack_utils.rb
commit f61f21c22caf2d784d2a0e00e4866cf528c21ced
Author: tlrobinson <t...@280north.com>
Date: Fri May 22 15:46:01 2009 -0700
Fix for form names containing "=": split first then unescape
components
Signed-off-by: Christian Neukirchen <chneuk...@gmail.com>
:100644 100644 40f9b49... 2d1ddc5... M lib/rack/utils.rb
:100644 100644 8f4fd75... 7dd5880... M test/spec_rack_utils.rb
commit 480eb783477175d5c409d38258a65ce6cf3175e4
Author: Simon Chiang <simon.a...@gmail.com>
Date: Fri May 15 22:53:04 2009 +0800
optimizations of parse/build query
Signed-off-by: Joshua Peek <jo...@joshpeek.com>
:100644 100644 447e89c... 40f9b49... M lib/rack/utils.rb
commit fb01269645cddfaff95eefa91b93c894666ea96b
Author: raggi <jftu...@gmail.com>
Date: Wed Apr 29 00:28:12 2009 +0100
Switch to emacs style encoding comment
:100644 100644 a9ab0d9... 447e89c... M lib/rack/utils.rb
commit 9abab4a36b10a0ce3f667ef1099e3361a648f425
Author: raggi <jftu...@gmail.com>
Date: Wed Apr 29 00:15:53 2009 +0100
Fix multipart uploads on 1.9
:100644 100644 ea24fec... a9ab0d9... M lib/rack/utils.rb
commit e7ab02a770756d3df7a838d50d04ec6e0b593046
Author: Joshua Peek <jo...@joshpeek.com>
Date: Tue Apr 28 11:43:22 2009 -0500
Add anchor to nested params parsing regexp to prevent stack
overflows
:100644 100644 6aaf478... ea24fec... M lib/rack/utils.rb
Christian Neukirchen
Joshua Peek <jo...@joshpeek.com> writes:
> I'd like to do a 1.0.1 maintenance soon of the rack-1.0 branch. Please
> test your apps and frameworks on that branch. It should be a fully
> compatible bug fix release.
This is a good idea. When you consider it ready, I can write an
announcement and do a formal release on Rubyforge and my servers.
> I went though most of the Lighthouse tickets and only found one that
> really concerned 1.0.
>
> "Rack::Utils.unescape problems in Ruby 1.9.1"
> http://rack.lighthouseapp.com/projects/22435/tickets/48-rackutilsunescape-problems-in-ruby-191
>
> http://rack.lighthouseapp.com/projects/22435-rack/milestones/current
>
> Please reassign any 1.1 milestones tickets to 1.0 or let me know of
> anything else master that you'd liked cherry-picked into rack-1.0.
>
> Looking forward to Rack 1.1.
I'm open for Rack 1.1 suggestions. 1.1-discussed things can be merged
into master now.
I have more time to do a release cycle in the next few weeks, so if we
merge quickly, we can push a 1.1 pretty soon.
blink, do you still want to split the openid parts? 1.1 would be a
good juncture.
Thanks to everyone,
--
Christian Neukirchen <chneuk...@gmail.com> http://chneukirchen.org
Joshua Peek
Neukirchen<chneuk...@gmail.com> wrote:
> I'm open for Rack 1.1 suggestions. 1.1-discussed things can be merged
> into master now.
>
> I have more time to do a release cycle in the next few weeks, so if we
> merge quickly, we can push a 1.1 pretty soon.
For 1.1, I want to pull parts of Bryan Helmkamp's Rack::Test lib into
core. Its been proven and widely accepted by many frameworks so far. I
want to start a separate thread for this and we can discuss whats
important, what belongs in core, etc. 1.1 will give us a chance to
improve Rack's testing support.
http://github.com/brynary/rack-test/
--
Joshua Peek
Joshua Peek
>> test your apps and frameworks on that branch. It should be a fully
>> compatible bug fix release.
>
> This is a good idea. When you consider it ready, I can write an
> announcement and do a formal release on Rubyforge and my servers.
Can anyone report any failures on 1.9? I have some that are related to
a ruby regression (in 1.9.1-p152).
http://groups.google.com/group/rack-devel/browse_thread/thread/a2aab3a4720f34c4
Also, I'm trying to contact some people atm to setup a rack CI server for us.
--
Joshua Peek
James
> I went though most of the Lighthouse tickets and only found one that
> really concerned 1.0.
>
> "Rack::Utils.unescape problems in Ruby 1.9.1"
To correctly solve this issue Rack needs to make some decisions about
how to mark the encoding of incoming data.
Jeremy Kemper and I started a discussion on that ticket that really
needs to be promoted to a mailing list.
Ignoring the issue of discerning the encoding of the request query
string, there's still the request body on POST requests. According to
RFC2616, Rack can use the charset segment of the requests Content-Type
header to determine the request body encoding and assume ISO-8859-1 if
a charset isn't provided.
However FF 3.5 doesn't appear to specify the charset of its UTF-8
encoded requests, which means Rack *should* (incorrectly) mark the
incoming data as ISO-8859-1.
Can anyone shed some light on what the correct behaviour for Rack
should be?
James
Joshua Peek
>> test your apps and frameworks on that branch. It should be a fully
>> compatible bug fix release.
>
> This is a good idea. When you consider it ready, I can write an
> announcement and do a formal release on Rubyforge and my servers.
Rails 2.3 is all systems go w/ rack 1.0.1.
Should wait on Ryan to test w/ Sinatra.
--
Joshua Peek
Christian Neukirchen
Ryan Tomayko
Sorry. I'm way behind on ML/lighthouse duties.
Sinatra looks good under the rack-1.0 branch. +1 for release.
The Sinatra test suite fails like crazy under 1.9 due to this
rack.input / external encoding Rack::Lint check:
http://github.com/rack/rack/commit/851782648fbbb141b1b56b73d8efc99efcac1b27#L0R240
The Sinatra test suite puts Rack::Lint in the middleware pipeline for
every test, though. I don't think that's a common practice. It passes
100% when I remove that check. The issue seems to be with the way
Rack::MockRequest wraps the :input string in an StringIO without any
encoding work. Or, maybe we should be setting the encoding on the
input string before passing it in. I should be able to work around
this in Sinatra somehow.
Thanks,
Ryan
Joshua Peek
On Aug 9, 2:50 am, Ryan Tomayko <r...@tomayko.com> wrote:
>
> > On Mon, Aug 3, 2009 at 5:25 PM, Christian
> > Neukirchen<chneukirc...@gmail.com> wrote:
> >>> I'd like to do a 1.0.1 maintenance soon of the rack-1.0 branch. Please
> >>> test your apps and frameworks on that branch. It should be a fully
> >>> compatible bug fix release.
>
> >> This is a good idea. When you consider it ready, I can write an
> >> announcement and do a formal release on Rubyforge and my servers.
>
> > Rails 2.3 is all systems go w/ rack 1.0.1.
>
> > Should wait on Ryan to test w/ Sinatra.
>
> Sorry. I'm way behind on ML/lighthouse duties.
>
> Sinatra looks good under the rack-1.0 branch. +1 for release.
>
> The Sinatra test suite fails like crazy under 1.9 due to this
> rack.input / external encoding Rack::Lint check:
>
Scytrin dai Kinthra
Moving the openid stuff into rack-contrib for 1.1 sounds good. I am on a vacation till the end of next week so as soon as I finish some personal things I can do the migrations.
Additionally, as memcached ruby gem seems to be more stable with 0.17, I plan on evaluating using that rather the memcache-client, unless there are any objections.
--
stadik.net
On Aug 3, 2009 4:00 PM, "Christian Neukirchen" <chneuk...@gmail.com> wrote:
[Sorry for being rather inactive lately, my real life is too busy. ;)]
Joshua Peek <jo...@joshpeek.com> writes: > I'd like to do a 1.0.1 maintenance soon of the rack-1.0 ...
This is a good idea. When you consider it ready, I can write an
announcement and do a formal release on Rubyforge and my servers.
> I went though most of the Lighthouse tickets and only found one that > really concerned 1.0. > > ...
Christian Neukirchen
> Moving the openid stuff into rack-contrib for 1.1 sounds good. I am on a
> vacation till the end of next week so as soon as I finish some personal
> things I can do the migrations.
>
> Additionally, as memcached ruby gem seems to be more stable with 0.17, I
> plan on evaluating using that rather the memcache-client, unless there are
> any objections.
Alright!
Ok, everyone cherry-pick fixes for 1.0.1, and I can make a release
this week (Thu or Fri).
James Tucker
>
> On Mon, Oct 5, 2009 at 10:14 PM, Scytrin dai Kinthra <scy...@gmail.com
> > wrote:
>> Moving the openid stuff into rack-contrib for 1.1 sounds good. I am
>> on a
>> vacation till the end of next week so as soon as I finish some
>> personal
>> things I can do the migrations.
>>
>> Additionally, as memcached ruby gem seems to be more stable with
>> 0.17, I
>> plan on evaluating using that rather the memcache-client, unless
>> there are
>> any objections.
>
> Alright!
>
> Ok, everyone cherry-pick fixes for 1.0.1, and I can make a release
> this week (Thu or Fri).
to rack-1.0?
Christian Neukirchen
Quin Hoxie
On Oct 6, 7:06 am, Christian Neukirchen <chneukirc...@gmail.com>
wrote:
>
> > Moving the openid stuff into rack-contrib for 1.1 sounds good. I am on a
> > vacation till the end of next week so as soon as I finish some personal
> > things I can do the migrations.
>
> > Additionally, as memcached ruby gem seems to be more stable with 0.17, I
> > plan on evaluating using that rather the memcache-client, unless there are
> > any objections.
>
> Alright!
>
> Ok, everyone cherry-pick fixes for 1.0.1, and I can make a release
> this week (Thu or Fri).
>
> --
Christian Neukirchen
Daniel Rodríguez Troitiño
I sent a patch to the list a couple of days ago. I have been looking
in the github page and it haven't been included. Does this mean that
it will not be included in 1.0.1?
Thanks.
Daniel Rodríguez Troitiño.
El 16/10/2009, a las 22:30, Christian Neukirchen
<chneuk...@gmail.com> escribió:
Christian Neukirchen
> Does this mean that
> it will not be included in 1.0.1?
I'll merge it.