Forum: Ruby-core [ruby-trunk - Bug #6019][Open] Revision r34582 significantly slows down rails app

Issue #6019 has been reported by Yura Sokolov.

----------------------------------------
Bug #6019: Revision r34582  significantly slows down rails app
https://bugs.ruby-lang.org/issues/6019

Author: Yura Sokolov
Status: Open
Priority: Normal
Assignee:
Category: core
Target version: 2.0.0
ruby -v: ruby 2.0.0dev (2012-02-14 trunk 34592) [i686-linux]


Before r34582 simple rails32 app (query sqlite and disply 20 rows of 
simple data)
benchmarked as 55req/second , after its performance is 33req/second.

I could agree that RubyOnRails should be in sync with ruby, but maybe 
there is other way
to fix `respond_to?` ?

Issue #6019 has been updated by Yura Sokolov.


Fix were relatively easy. Close this issue, please. Excuse me for 
disturbing.
----------------------------------------
Bug #6019: Revision r34582  significantly slows down rails app
https://bugs.ruby-lang.org/issues/6019

Author: Yura Sokolov
Status: Open
Priority: Normal
Assignee:
Category: core
Target version: 2.0.0
ruby -v: ruby 2.0.0dev (2012-02-14 trunk 34592) [i686-linux]


Before r34582 simple rails32 app (query sqlite and disply 20 rows of 
simple data)
benchmarked as 55req/second , after its performance is 33req/second.

I could agree that RubyOnRails should be in sync with ruby, but maybe 
there is other way
to fix `respond_to?` ?

Issue #6019 has been updated by Yui NARUSE.

Status changed from Open to Assigned
Assignee set to Akinori MUSHA


----------------------------------------
Bug #6019: Revision r34582  significantly slows down rails app
https://bugs.ruby-lang.org/issues/6019

Author: Yura Sokolov
Status: Assigned
Priority: Normal
Assignee: Akinori MUSHA
Category: core
Target version: 2.0.0
ruby -v: ruby 2.0.0dev (2012-02-14 trunk 34592) [i686-linux]


Before r34582 simple rails32 app (query sqlite and disply 20 rows of 
simple data)
benchmarked as 55req/second , after its performance is 33req/second.

I could agree that RubyOnRails should be in sync with ruby, but maybe 
there is other way
to fix `respond_to?` ?

Issue #6019 has been updated by Yura Sokolov.


Rails already include fix against this issue into its trunk: 
https://github.com/rails/rails/pull/5039

Any way, could second parameter have more variants? So that:
- `respond_to?(:meth)` and `respond_to?(:meth, nil)` will test for 
public methods,
- `respond_to?(:meth, true)` will test all methods
- `respond_to?(:meth, false)` (or `respond_to?(:meth, :with_protected)`) 
will test for public and protected methods

It is just a suggestion and I could not prove it will be very useful.
----------------------------------------
Bug #6019: Revision r34582  significantly slows down rails app
https://bugs.ruby-lang.org/issues/6019

Author: Yura Sokolov
Status: Assigned
Priority: Normal
Assignee: Akinori MUSHA
Category: core
Target version: 2.0.0
ruby -v: ruby 2.0.0dev (2012-02-14 trunk 34592) [i686-linux]


Before r34582 simple rails32 app (query sqlite and disply 20 rows of 
simple data)
benchmarked as 55req/second , after its performance is 33req/second.

I could agree that RubyOnRails should be in sync with ruby, but maybe 
there is other way
to fix `respond_to?` ?

Issue #6019 has been updated by knu (Akinori MUSHA).

Status changed from Assigned to Closed
Target version deleted (2.0.0)

Thanks for the followup.  This change was meant to be a fix for a 
misfeature, and it's great Rails followed it quickly.

As for a feature request, I think we need a little more consideration 
and details.  Please submit one again when you work out a concrete 
specification with some confidence.
----------------------------------------
Feature #6019: Revision r34582  significantly slows down rails app
https://bugs.ruby-lang.org/issues/6019#change-36492

Author: funny_falcon (Yura Sokolov)
Status: Closed
Priority: Normal
Assignee: knu (Akinori MUSHA)
Category: core
Target version:


Before r34582 simple rails32 app (query sqlite and disply 20 rows of 
simple data)
benchmarked as 55req/second , after its performance is 33req/second.

I could agree that RubyOnRails should be in sync with ruby, but maybe 
there is other way
to fix `respond_to?` ?