Boost logo

Boost-Build :

From: Rene Rivera (grafik666_at_[hidden])
Date: 2002-11-27 10:45:02


[2002-11-27] Vladimir Prus wrote:

>David Abrahams wrote:
>> Rene Rivera <grafik666_at_[hidden]> writes:
>>>Or even:
>>>
>>>source test-src-common : source1.cpp source2.cpp ;
>>>source test-src : main-gcc.cpp test-src-common : <toolset>gcc ;
>>>source test-src : main-msvc.cpp test-src-common : <toolset>msvc ;
>>>source test-src : main-darwin.cpp test-src-common : <toolset>darwin ;
>>>exe test : test-src ;
>>
>>
>> Ding!!
>>
>> Beautiful. I vote yes.
>>
>
>Ding as well! This even will be simple to implement -- much
>in the same way as 'symlink' target that Rene did.
>
>I've filed
>
> http://zigzag.cs.msu.su:7814/scarab/servlet/scarab/id/BB12
>
>on it, so that we do it eventually. I'd only prefer
>to use more general name, something like "group", because there's
>no reason why you can't use main targets together with source files.
>Something like
>
> group os-abstraction : win.cpp my_i18_lib : <os>NT ;

I'm not sure about using "group" instead of "source", for various reasons...

1) "source" seems more directly intuitive than "group". Seeing as it maps
directly to the intended use in "<target-rule> <target-name> : <**sources**>
: <requirements> ;"

2) I don't see a problem in using "source" with main targets as you point
out. This seems perfectly obvious to me also:

lib my_i18_lib : source.cpp ;
source os-abstraction : win.cpp my_i18_lib : <os>NT ;

3) We might want to use "group" for the categorization of targets that is
needed for the stage rule. For example:

lib some-lib : source.cpp : <group>one ;
lib another-lib : source2.cpp : <group>two ;
stage release : : <group>two ;

Don't hold me to that grammar ;-) It looks kinda of klunky, but using
"group" for this make more sense to me than using it for grouping sources.

-- grafik - Don't Assume Anything
-- rrivera_at_[hidden] - grafik_at_[hidden]
-- 102708583_at_icq - Grafik666_at_AIM - Grafik_at_[hidden]

 


Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk