Proposal to reduce anti-bundling requirements

[Vít Ondruch]

Dne 11.9.2015 v 18:08 Vít Ondruch napsal(a):
> Dne 11.9.2015 v 17:21 Matthew Miller napsal(a):
>> On Fri, Sep 11, 2015 at 03:51:33PM +0200, Vít Ondruch wrote:
>>>>  * Increases the available pool of software that can be packaged
>>>> substantially (many modern languages such as Ruby and Go are
>>>> realistically only functional with allowable bundling)
>>> Not sure why you put Ruby into this group. There is no evidence that
>>> bundling is more prominent in Ruby then in other languages. If your
>>> judgment is based on existence of rubygem-bundler, then you completely
>>> misunderstand purpose of the project.
>> Really? I think it's completely fair to say that "vendor everything!"
>> has at least historically been a cultural norm for Ruby development.
>>
>>

And actually, this "vendor everything" is nothing different from what
are doing containers now. This is typically very easy to unbundle.

Vít

> So what are your chances to install for example Redmine on Fedora?
> Bundle everything using Bundler, because Fedora does not provide
> multiple versions of Rails.
>
> And I repeat it again and again, rails32-3.2-1, rails40-4.0-1,
> rails41-4.1-1 and rails42-4.2-1 packages with all their dependencies are
> not way forward. The tooling should support rails-3.2-1 and rails-4.0-1
> simultaneous installation, but even the minimal step towards this, i.e.
> treet the packages as install only packages and allow the upgrades when
> only release is changed  is actively denied by all parties involved in
> tooling.
>
> And not going far for example, this [1] ticket was opened just today.
> This does not harm anybody, but yet if seems to be problem to have to
> versions of rubygem-json.
>
>
>
> Vít
>
>
> [1] https://bugzilla.redhat.com/show_bug.cgi?id=1261931
>