[chef] Re: Re: Ruby version support in cookbook metadata

[Anthony Goddard < >]

Hey Bryan,
That sounds good to me. I guess my thought with being more explicit was to 
emphasize the difference between unsupported and untested, but  '>=1.9.3' etc 
definitely seems more elegant.

Ant



On Apr 5, 2012, at 9:13 PM, Bryan McLellan wrote:

> On Thu, Apr 5, 2012 at 8:13 PM, Anthony Goddard 
> <
 >
>  wrote:
>> I'm sure much of the community is split between ruby 1.8.7 and 1.9.* still,
>> and as time goes on the chance of cookbooks appearing with 1.9 only syntax
>> only increases. I guess the current method of declaring this rests either 
>> in
>> checks within the cookbooks themselves, or in cookbook comments on the
>> community site (http://community.opscode.com/cookbooks/metachef/comments).
>> 
>> What do you guys think of adding a "supported ruby" attribute in cookbook
>> metadata? Perhaps with 'supported' and 'unsupported' values for version
>> (then anything undeclared would be 'untested')? Something like that anyway,
>> and then it could be displayed on the community site.
>> 
>> http://community.opscode.com/ideas/26
> 
> From IRC the other day:
> 
> | Is there any guidance on recommended/required versions of ruby to
> run Chef with?
> |  I've been working just fine with ruby 1.8.7, but I've hit some
> cookbooks from the
> | community site that are apparently coded in ruby 1.9 and cause
> syntax errors with
> | my current installation. I'm hesitant/reluctant to possibly break
> other cookbooks if I
> | jump up to 1.9.x, and I've not seen any docs on opscode that gives
> guidance to
> | cookbook authors on what they should
> 
> It does seem to be a problem for folks, particularly when Ruby is new to 
> them.
> 
> Do we need more than a 'ruby' field and using the same operators we
> use for cookbook versions? Granted we have more control over our
> cookbook versions than we do of Ruby, but I suspect the kind of
> Control we're really talking about is something like '>= 1.9.3'
> 
> Bryan