Re: [boost] [cmake] Minimum viable cmakeification for Boost

Subject: Re: [boost] [cmake] Minimum viable cmakeification for Boost
From: Robert Ramey (ramey_at_[hidden])
Date: 2017-06-21 18:47:21

On 6/21/17 10:01 AM, Edward Diener via Boost wrote:
>> Boost build has been hampered by a number of design decisions which 
>> have never been revisited.
>>
>> a) It is "too smart".  It tries to infer the toolset by looking around 
>> the file system.  If one has more than one compiler set there is 
>> really know way to know which one is going to pick.  So then one is 
>> forced to go into the workings of boost build to figure out how to 
>> over ride this behavior.
> 
> The solution, of course, is always to pass toolset=xxx to the b2 command 
> line. In my own use of b2 I always do this.

Right and so do I.

I also pass the other switches.  This re-enforces my point that bjam 
could my made much better and easier to use with huge changes by 
altering it's user interface philosophy.  If I were required to pass all 
the switches either on the command line or in the local jamfile itself, 
then I would only get helpful error messages rather than inexplicable 
behavior.  I believe that evolving the bjam user interface to his 
phisophy wouldn't be a huge project.

Robert Ramey

> 
> snipped...
>>
>> Robert Ramey
> 
> 
> _______________________________________________
> Unsubscribe & other changes: 
> http://lists.boost.org/mailman/listinfo.cgi/boost
>