tech-toolchain archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Initial LLVM/clang patch



On Fri, Feb 04, 2011 at 07:01:35AM +0000, YAMAMOTO Takashi wrote:
> IMO, there's no problem to have work-in-progress stuff in any
> (possibly ugly or even wrong) shape which is easy for the interested
> developer to work on, unless it hurts others.

Yes, I think you're right - I've been talking about one thing, and
everyone else has been talking about something else completely.  I'd
been looking at this from a "we need to have a good framework for
external toolchains", and I think everyone else has been looking at
this from a "we need to be able to work on clang/llvm even before it's
ready".

To that end:

+ Perfect may well be the enemy of the good here.

+ I'd still like to see pkgsrc grow fetch methods for "svn co" and
"git clone"

+ I'd also like to see a decent framework at the time the files for clang
hit the tree

+ Until such time as that happens, I think a "svn download" makefile is
probably workable, as long as it's understood that this is not the end
state, that it's a temporary measure, and that the clang/llvm files will
be in-tree when we and they are ready

...and that I withdraw my objections if we can all agree on the above. I
know that others have sent me private email supporting my stance, though,
so you may have to convince them, too.

And, if it's not apparent - thanks for working on this, Joerg - getting
more and better support for external toolchains is great. Thanks also to
the pcc folks in this regard.

Best,
Alistair


Home | Main Index | Thread Index | Old Index