missing 'Content-Type' header in 'send_data' + zena advice

rubyonrails-core
1

Dear list,

This is one email with two questions:

Question 1:

There is a bug in send_file_headers! (3.0.pre: metal/streaming.rb): missing ‘Content-Type’ => self.content_type

The current test does not test the actual headers for this parameter.

  • Should I open a ticket with a patch for this ?

Question 2:

I intended to porting zena from rails 1.2.3 to 2.3 but need the new “inverse_of” patch which is only commited to 3.0.pre (AFAIK).

  • What is the status of the “master” branch ? Is it reasonable to work the transition there directly or is it broken more often then not ?

Thanks for any advice,

Gaspard

2

I am asking the question again since the subject of the original email was not very clear:

I am porting zena from rails 1.2.3 to 2.3 but need the new "inverse_of" patch which is only commited to 3.0.pre (AFAIK).

* What is the status of the "master" branch ? Is it reasonable to work the transition there directly or is it broken more often then not ?

Thanks for any insight,

Gaspard

3

I am asking the question again since the subject of the original email was not very clear:

I am porting zena from rails 1.2.3 to 2.3 but need the new "inverse_of" patch which is only commited to 3.0.pre (AFAIK).

* What is the status of the "master" branch ? Is it reasonable to work the transition there directly or is it broken more often then not ?

The master branch is *generally* stable, however it's not recommended as a development target unless you're comfortable hitting a few bumps every now and then. While we pretty much never commit code we *know* is broken, occasionally a refactoring we do could accidentally break your app or a plugin you use. Also, lots of plugins today won't work on master.

So yeah, you could do it, but if you're still on 1.2.3 you've got a fair way to go first :).

Some advice, when you're doing your upgrade go from 1.2.3->1.2.6->2.0.x->2.1.x->2.2.x->2.3.x. Each step along the way will give you warnings which you can fix, rather than upgrading and having nothing work.

4

Thanks a lot for the reply Koz.

I'm glad to learn the *master* branch is stable enough: I don't mind finding quirks (I'll propose patches along the way if I can) and I don't need to go in production right away.

As for the transition: the application launches on 3.0.pre already and I have 1500 tests to show me what's going wrong so I should not miss something important. I have also refactored some code with some of the niceties from recent versions of rails (nested_attributes, inverse_of, I18n, etc) so I think the direct jump to 3.0.pre will be actually easier.

Thanks again,

Gaspard