Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 13 Sep 2012 02:31:50 +0200
From:      Daniel Nebdal <dnebdal@gmail.com>
To:        Doug Barton <dougb@freebsd.org>
Cc:        Jerry <jerry@seibercom.net>, freebsd-ports@freebsd.org
Subject:   Re: Clang as default compiler November 4th
Message-ID:  <CA%2Bt49P%2Bvve_%2BGoQS9D7=K8S0bqT=dFGjrK9eugO1fLvrsBzSQQ@mail.gmail.com>
In-Reply-To: <5050F7CF.4070204@FreeBSD.org>
References:  <20120910211207.GC64920@lor.one-eyed-alien.net> <20120911104518.GF37286@deviant.kiev.zoral.com.ua> <20120911120649.GA52235@freebsd.org> <20120911122122.GJ37286@deviant.kiev.zoral.com.ua> <20120911123833.GA54483@freebsd.org> <848C813E-E6EC-4FAF-9374-B5583A077404@cederstrand.dk> <505055F7.9020809@FreeBSD.org> <20120912072209.65bc5e3d@scorpio> <5050F7CF.4070204@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Sep 12, 2012 at 10:59 PM, Doug Barton <dougb@freebsd.org> wrote:
> On 9/12/2012 1:22 AM, Jerry wrote:
>> On Tue, 11 Sep 2012 23:29:27 -1000
>> Doug Barton articulated:
>>
>>> What we need to do is what I and others have been asking to do for
>>> years. We need to designate a modern version of gcc (no less than 4.6)
>>> as the official default ports compiler, and rework whatever is needed
>>> to support this. Fortunately, that goal is much more easily achieved
>>> than fixing ports to build and run with clang. (It's harder than it
>>> sounds because there are certain key libs that define some paths
>>> depending on what compiler they were built with, but still easier
>>> than dealing with clang in the short term.)
>>
>> That is a well thought out, highly intuitive and completely doable
>> idea. Therefore it will be ignored.
>
> No, it'll be ignored because I suggested it. :)
>
>> It seems that the FreeBSD authors are more concerned with the
>> licensing language of GCC than in getting a fully functioning port's
>> compiler into the FreeBSD base system.
>
> Just to be clear, I'm not suggesting putting the "ports compiler" into
> the base. I'm suggesting that it be managed as a port, just like pkg is.
> This works fine for the ports that are already hard-coding compiler
> dependencies, and mostly worked for me back when I get it a test run
> when I made the suggestion years ago. The few glitches I (and others who
> have done it since) ran into just need some elbow grease applied.
>
> By keeping ports-related things in the ports tree we gain a huge amount
> of agility, and lose the concerns about licensing in the base. It's a
> win/win.
>
> Doug
>


Three-ish things:
a) Doesn't that remove all incentives for eventually converging on
just one compiler (bar some specific exceptions)?
a.1) Isn't that bad?
b) Doesn't that mean that at some future point, we'll have to jump the
ports compiler to a newer (probably much newer) version, with all the
maintenance fun of that?
c) I guess this still lets me use clang for most ports if I really
wish to? (I've compiled most ports with clang for a while, and the
speed + useful error messages would be hard to give up...)

I'm not sure if those are big or small issues - probably a matter of taste. :)

-- 
Daniel Nebdal



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CA%2Bt49P%2Bvve_%2BGoQS9D7=K8S0bqT=dFGjrK9eugO1fLvrsBzSQQ>