Re: gems is a language change, not a pkging system

On 9/28/05, Sean E. Russell <[e-mail]> wrote:
> On Monday 26 September 2005 22:41, Austin Ziegler wrote:
>> The *real* harm is that without RubyGems, there would be fewer
>> libraries easily installed.
> This is a straw-man argument.

Not in the least, Sean, and I'm *really* getting tired of repeating why
it's not.

> The easiest way for users to install software -- ANY software,
> including libraries -- is by using the package manager that they use
> for installing all other software on their system.

Wrong. If others aren't actively repackaging software -- including
micropackages -- for your platform, you're screwed. Period. I said it on
my other response, but where are the native Rails packages for HP-UX,
Solaris, and AIX? What about for MacOS X and Windows .exe installers?

If these aren't being packaged for users in native formats -- and Ruby,
fortunately, runs on more platforms than just Linux -- then your
statement is utter bunkus. RubyGems increases what can be easily
installed on RUBY platforms.

>> You don't like RubyGems? MAKE SOMETHING YOURSELF AND PROMOTE IT.
>> Otherwise you're simply being obstructionist and *no one* has time
>> for that.
> Another straw man. Users may oppose RubyGems for other reasons not
> having to do with any specific implementation of a library manager.

Again, not a straw man. If you don't like RubyGems, create something
else and promote it.

> I only fear getting locked into using a specific library manager.  I
> (and you) expect the following process to occur:
>
> 1 RubyGems is bundled with Ruby.
> 2 Gems then becomes viewed as the standard library distribution
>   mechanism.
> 3 Library developers, especially newcomers who've never used setup.rb,
>   will only distribute their packages as Gems.
> 4 The perception of the standard becomes a fact, resulting in a
>   feedback loop at 3.

Yes, this is more or less what I expect to happen. However, if the
proposals that I've made are accepted, that won't be a big deal at all.

> None of this bothers me as a developer, because I won't be
> distributing my libraries as Gems. However, it bothers me greatly as a
> user, because now I won't be able to install Ruby libraries without
> gems, which forces me to use it. It also means that I won't be able to
> use Ruby at work, because of RubyGems' aforementioned inability to
> deal with firewalls properly.

Um. That inability is a bug. I'm sure it can be fixed. I don't know how
to fix it, but I'm sure it can be fixed.

>> API version numbering in the name is one of the *worst* things you
>> can do. I much prefer supporting AIX and HP-UX to supporting Linux
>> because of this nonsense. (Solaris isn't much better than Linux for
>> this.)
> "Carthago delenda est!"  --  The versioning mechanism must be separate
> from the library manager.

Sorry, but as I said in the other response to you, I disagree.
Soversions suck, and something needs to be done properly in the language
to support versioned APIs. RubyGems isn't perfect for this, but it's a
hell of a lot better than not having anything at all.

-austin
--
Austin Ziegler * [e-mail]
               * Alternate: [e-mail]