This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Interpolation of Env variables.
- To: Donn Terry <donn at interix dot com>
- Subject: Re: Interpolation of Env variables.
- From: Zack Weinberg <zack at rabi dot columbia dot edu>
- Date: Thu, 08 Apr 1999 11:57:54 -0400
- cc: "egcs-patches at cygnus dot com" <egcs-patches at cygnus dot com>, Mumit Khan <khan at xraylith dot wisc dot edu>, Zack Weinberg <zack at rabi dot columbia dot edu>
On Thu, 08 Apr 1999 09:50:07 -0600, Donn Terry wrote:
>In this case, it's not just the tool chain, but everything
>that's anchored to a single environment variable: the system
>include directory (/usr/include, in effect), the system library
>directory (/usr/lib, in effect), as well as the tools themselves.
>(Yes... EVERYTHING: what you'd expect to be at / can be
>in any one location in the filesystem.) In this case, it's
>a system-provided environment variable that everything uses.
VMS works vaguely like this as well, and I think quite a few
VMS-specific ifdefs could be removed from gcc.c if the specs could
interpolate environment variables ("logical names").
[...]
>As I noted in another part of my cover note, Zack's proposal
>for a new specs language has the same feature (completely
>independently invented). It would seem to be a useful tool
>to add to the toolbox, independent of how it is used.
Actually, I saw your proposal and thought "that would be useful" so I
added the facility to mine.
My project is _very_ long term and shouldn't affect this decision.
zw