Re: [Chicken-users] depreciating CHICKEN_HOME

[Goetz Isenmann]

On Sat, Jun 10, 2006 at 12:09:29AM +0200, felix winkelmann wrote:
> On 6/10/06, Brandon J. Van Every <address@hidden> wrote:
> >>
> >> Sorry, I can't follow this at all: vcbuild.bat doesn't use CHICKEN_HOME,
> >> also makefile.vc doesn't. For Windows platforms built via makefile.vc,
> >> csc.scm
> >> uses CHICKEN_HOME, as this seems to be the only decent way of
> >> specifiying an
> >> installation destination.
> >
> >What I mean is, csc built from vcbuild.bat or makefile.vc relies on a
> >hardwired behavior for CHICKEN_HOME.  ./configure builds and
> >installations do not appear to depend on CHICKEN_HOME for anything,
> >either at build time or runtime, although I could be mistaken.
> 
> No, you're perfectly right.

I had hoped, that I

 - could compile chicken with any arbitrary installation path (say
   /var/tmp/chicken or %TEMP%\chicken),

 - build a binary package from that,

 - install this package into another directory on many machines or on
   a network share (if neccessary different versions side by side),

 - and use these installations after setting CHICKEN_HOME and
   LD_LIBRARY_PATH (or equiv).

Currently, that does not seem to work. Is that something, that could
be fixed, or are there other reasons, why I should not try to do this
with the interpreter, the compiler, the runtime, or (static/dynamic
linked) applications?
-- 
Goetz Isenmann