[33798] trunk/dports/mail

Anders F Björklund afb at macports.org
Tue Feb 5 23:32:28 PST 2008


Ryan Schmidt wrote:

>> As this software is just a perl script and requires no extra steps 
>> for other architectures shouldn't it considered to be already 
>> universal?
>>
>> Which would be:
>>   default_variants +universal
>>   variant universal {}
>
> I would say no. I would say "universal_variant no" is currently 
> appropriate for software that is architecture-agnostic, like this perl 
> script.

Technically +universal would be 2 (or 4) architectures, while this is 
for 0 (zero) architectures.
That's why it is called "noarch" in some other packaging systems* (that 
call the universal "fat")

Having this "don't need compile" metadata in the Portfile is good, just 
that "universal" isn't it:
http://trac.macports.org/projects/macports/ticket/12206

But it could be joined to a single variable, like Ryan is suggesting, 
as they do affect eachother ?
(if it doesn't have a machine architecture then it doesn't need to have 
a +universal variant either)

--anders

* that would be RPM (.noarch.rpm), while another name is used in DEB 
(.all.deb)
   difference being how many it is built for, versus how many that it 
applies to



More information about the macports-dev mailing list