[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Chicken-users] README.darcs bootstrap claim probably false
From: |
felix winkelmann |
Subject: |
Re: [Chicken-users] README.darcs bootstrap claim probably false |
Date: |
Fri, 9 Jun 2006 22:21:03 +0200 |
On 6/7/06, Brandon J. Van Every <address@hidden> wrote:
README.darcs says at the very beginning:
0. You will need:
* A system Chicken has been ported to. See the homepage for details.
To bootstrap the system Chicken 2.315 or newer is needed.
I bet this is false. I'm building on CMake using a two-stage bootstrap,
and plain Chicken 2.3 works just fine. Now granted, I don't test the
canonical Darcs bootstrap, only my own CMake bootstrap. Maybe my build
can do something that yours can't, ha ha ha ha ha! My secret sauce is I
don't bother with .exports when generating my bootstrap compiler. Only
the final compiler needs .exports. Granted further, I haven't done much
testing of the results, so maybe they'll keel over and die. But
anyways, my suspicion is that requiring 2.315 is excessively conservative.
It is rather conservative, but your cmake-based build is 2-stage and such
somewhat more forgiving regarding the (1st stage) bootstrapping compiler
version.
If only 2.3 is required, I'd like to see that statement made in
README.darcs. I'm redrafting my INSTALL-CMake.txt docs and if we can
say the same thing, we should say the same thing. Originally I also
thought I was going to refer to the README.darcs instructions, but now I
realize it's got lotsa stuff that CMake doesn't care about. So I'll
write up my own instructions on how to obtain Darcs and so forth.
Again, for a 1-stage bootstrap, 2.315 will generate a better chicken, so for
the autotools-based build, 2.315 is recommended. But since the cmake-based
build is somewhat more sophisticated, I guess you're right.
(felix)