On 15.04.2013 15:04, Klaus Kaempf wrote:

* Stephan Kulow [Apr 15. 2013 10:27]:

...

We might want to discuss a strategy - I would like to replace ruby1.9 with ruby2.0 in 13.1.

Given that 13.1 is the base for SLE12, this upgrade is a no-brainer.

http://s.kulow.org/3B shows the difference between factory failures in d:l:r:e and ruby 2.0 default repo. Have a lot of fun. Greetings, Stephan -- To unsubscribe, e-mail: opensuse-ruby+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-ruby+owner@opensuse.org

On 29.04.2013 15:47, Klaus Kaempf wrote:

* Stephan Kulow [Apr 29. 2013 10:17]:

...

On 15.04.2013 15:04, Klaus Kaempf wrote:

...

* Stephan Kulow [Apr 15. 2013 10:27]:

...

We might want to discuss a strategy - I would like to replace ruby1.9 with ruby2.0 in 13.1.

Given that 13.1 is the base for SLE12, this upgrade is a no-brainer.

http://s.kulow.org/3B shows the difference between factory failures in d:l:r:e and ruby 2.0 default repo. Have a lot of fun.

So there are 20-30 gems not building properly, doesn't sound too bad.

And which of those failing gems are actually relevant ?

What is your (as the openSUSE release guard) acceptance criteria for Ruby 2.0 in factory ? All packages in d:l:r:e building ?

Nah. I think of those in factory only activesupport-3_2 and passenger are broken. activesupport will hopefully fix itself with the next 3.2 update (without the broken i18n dependency) and passenger someone will hopefully be able to fix. But I also noticed that xmlhash is not found when I tried to port build service to rails 4.0 on top of ruby 2.0, because for some reason ruby 2.0 only looks at gems/xmlhash-1.3.5/lib/xmlhash/xmlhash.so while there is only gems/xmlhash-1.3.5/ext/xmlhash/xmlhash.so in our packages. Problems like that you will face with webyast too - but I can easily setup a real factory staging project for ruby 2.0 if people are interested in fixing webyast Greetings, Stephan -- To unsubscribe, e-mail: opensuse-ruby+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-ruby+owner@opensuse.org

On 30.04.2013 08:28, Klaus Kaempf wrote:

* Stephan Kulow [Apr 29. 2013 16:15]:

...

Problems like that you will face with webyast too - but I can easily setup a real factory staging project for ruby 2.0 if people are interested in fixing webyast

Please do the necessary steps to get Ruby 2.0 into Factory. I do trust the Rubyists to fix whatever is needed.

If WebYaST does not get fixed in time, then there is no interest and it can be dropped.

Taking that WebYaST already fails for 44 days (and failed for 120 days before it was fixed once for 12.3), I wonder how to define "fixed in time". Greetings, Stephan -- To unsubscribe, e-mail: opensuse-ruby+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-ruby+owner@opensuse.org

Dne 3.5.2013 11:44, Ladislav Slezak napsal(a):

Dne 3.5.2013 09:56, Stephan Kulow napsal(a):

...

Taking that WebYaST already fails for 44 days (and failed for 120 days before it was fixed once for 12.3), I wonder how to define "fixed in time".

Yeah, I'll look at it...

And as for Ruby 2.0 I guess there should not be many problems, adding support for 1.9 was relatively easy (Webyast currently supports both 1.8 and 1.9).

And AFAIK 2.0 does not have many changes compared to 1.9.

Um, I just have looked at it and the tests fail for some strange reason, all "_path" helpers are undefined, e.g. ActionView::Template::Error: undefined local variable or method `destroy_account_session_path' for #<#Class:0x00000002e675d8:0x00000003d44448> See https://build.opensuse.org/package/live_build_log?arch=x86_64&package=webyast-base&project=YaST%3AWeb&repository=openSUSE_Factory It does not work even for "root_path" helper. But in the rails console it works:

rails c Loading development environment (Rails 3.2.13) irb(main):001:0> app.destroy_account_session_path => "/accounts/sign_out" irb(main):002:0> app.root_path => "/"

'rake routes' displays the routing with correct names, e.g. root / controlpanel#index And if I run the tests in Rails 3.2.12 with Ruby 1.9.1 (from 12.3) it is OK. Any idea what might be wrong? Google did not find anything relevant for me. Anyone tried the latest Rails with Ruby 2.0? -- Ladislav Slezák Appliance department / YaST Developer Lihovarská 1060/12 190 00 Prague 9 / Czech Republic tel: +420 284 028 960 lslezak@suse.com SUSE -- To unsubscribe, e-mail: opensuse-ruby+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-ruby+owner@opensuse.org

On 13.05.2013 17:39, Ladislav Slezak wrote:

Dne 3.5.2013 11:44, Ladislav Slezak napsal(a):

...

Dne 3.5.2013 09:56, Stephan Kulow napsal(a):

...

Taking that WebYaST already fails for 44 days (and failed for 120 days before it was fixed once for 12.3), I wonder how to define "fixed in time".

Yeah, I'll look at it...

And as for Ruby 2.0 I guess there should not be many problems, adding support for 1.9 was relatively easy (Webyast currently supports both 1.8 and 1.9).

And AFAIK 2.0 does not have many changes compared to 1.9.

Um, I just have looked at it and the tests fail for some strange reason, all "_path" helpers are undefined, e.g.

ActionView::Template::Error: undefined local variable or method `destroy_account_session_path' for #<#Class:0x00000002e675d8:0x00000003d44448>

Did you find an answer to that? I found it: test/functional/raising_controller_test.rb rewrites the routes - all tests after that have not your standard routes. Greetings, Stephan -- To unsubscribe, e-mail: opensuse-ruby+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-ruby+owner@opensuse.org

On 07.06.2013 12:06, Ladislav Slezak wrote:

Dne 6.6.2013 17:44, Stephan Kulow napsal(a):

...

Did you find an answer to that? I found it:

Not yet, currently I'm killing YCP in another project...

Then let's kill webyast. It just needs more care than we can invest ;( Greetings, Stephan -- To unsubscribe, e-mail: opensuse-ruby+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-ruby+owner@opensuse.org