[26400] trunk/dports/math/GiNaC/Portfile

[Juan Manuel Palacios]

x
On Jun 21, 2007, at 2:20 AM, Chris Pickel wrote:

> On 21 Jun, 2007, at 01:58, source_changes at macosforge.org wrote:
>>  depends_lib	port:readline	\
>>  		port:cln
>>
>> +platform darwin 6 {
>> +		ui_msg "GiNaC is not supported on Jaguar (OS X 10.2.x)."
>> +		exit 1
>> +}
>> +
>>  platform darwin 7 {
>> -		depends_build	port:gcc40
>> -
>> -		configure.env	CXX=g++-dp-4.0			\
>> -				CPPFLAGS=-I${prefix}/include	\
>> -				LDFLAGS=-L${prefix}/lib
>> +		ui_msg "GiNaC is not supported on Panther (OS X 10.3.x)."
>> +		exit 1
>>  }
>>
>>  platform darwin 8 {
>
> Hi, shouldn't these sorts of messages be in pre-fetch? Placing them  
> in the platform statement will still cause non-installing  
> statements (e.g. `port info GiNaC`, `port echo  
> maintainer:gwright at macports.org`) to fail on Panther.
>
> As an example, take a look at fusefs [1]. In this portfile, the  
> platform check is deferred until pre-fetch, thereby ensuring that  
> users will only get complaints if they're actually trying to  
> install the port.
>
>
> Chris
>
> [1] http://trac.macosforge.org/projects/macports/browser/trunk/ 
> dports/fuse/fusefs/Portfile


	Portfiles should definitely not use exit, it can break many things  
(for instance, if the index were being regen'd on a Panther box,  
GiNaC would break it). Also, the GiNaC port still uses deprecated  
configure.env instructions to set what are now default CPPFLAGS and  
LDFLAGS, they should all be removed.


-jmpp