Debian source and i386 binary packages are available at the Alioth pkg-gambc project. This renews the gambc package in Debian which was orphaned some time ago. We are working towards an upload into Unstable.
https://alioth.debian.org/projects/pkg-gambc/
Please try these packages. If you identify bugs, please post them to the pkg-gambc Tracker, and not to Debian Bug 283299.
Afficher les réponses par date
"Kurt B. Kaiser" kbk@shore.net wrote/schrieb 871wx7147g.fsf@hydra.bayview.thirdcreek.com:
https://alioth.debian.org/projects/pkg-gambc/
Please try these packages. If you identify bugs, please post them to the pkg-gambc Tracker, and not to Debian Bug 283299.
I'm lucky with the source packages so far.
I needed to build because there are no amd64 binary packages (no 32bit emulator running).
After hafing downloaded *.dsc, *.diff.gz, *.tar.gz, I've unpacked (dpkg-source -x gambc_4.0.b17-1.dsc) built (dpkg-buildpackage) and installed it. The REPL starts and behaves normal as far as I can see.
Option --enable-single-host is not set this way.
Regards Thomas
Thomas Hafner thomas@hafner.NL.EU.ORG writes:
Option --enable-single-host is not set this way.
Right, I was a little concerned about the build time/space on the buildd. I figured that once we get it into Unstable we can change the option and see how it goes.
Feel free to modify the config options in debian/rules after you unpack and before you build.
On Mar 13, 2006, at 8:14 PM, Kurt B. Kaiser wrote:
Thomas Hafner thomas@hafner.NL.EU.ORG writes:
Option --enable-single-host is not set this way.
Right, I was a little concerned about the build time/space on the buildd. I figured that once we get it into Unstable we can change the option and see how it goes.
With later versions of gcc, --enable-single-host is not so bad. (I did a *lot* of benchmarking with Gambit-generated C files and the gcc developers were responsive.) And leaving it off absolutely *kills* performance for many programs.
If you wanted to reduce the resources needed to compile Gambit, you might be better off setting
[lindv2:~/Desktop] lucier% diff -p /usr/local/Gambit-C/include/ gambit.h gambit.h *** /usr/local/Gambit-C/include/gambit.h Thu Jan 19 15:17:50 2006 --- gambit.h Wed Jan 25 13:03:48 2006 *************** *** 1192,1197 **** --- 1192,1198 ---- * Determine optimization goal. */
+ #define ___OPTIMIZE_SPACE #ifdef ___OPTIMIZE_TIME #ifdef ___OPTIMIZE_SPACE @error "Define either ___OPTIMIZE_TIME or ___OPTIMIZE_SPACE"
and seeing how that goes.
Brad
Bradley Lucier lucier@math.purdue.edu writes:
With later versions of gcc, --enable-single-host is not so bad. (I did a *lot* of benchmarking with Gambit-generated C files and the gcc developers were responsive.) And leaving it off absolutely *kills* performance for many programs.
I ran the package build w/o it:
12 min 37 sec Max memory 187M during compile.
And with it:
19 min 6 sec Max memory 396M during compile of _io.c
gcc (GCC) 4.0.3 20060304 (prerelease) (Debian 4.0.2-10)
The overall average memory use during compile is a lot higher.
But it doesn't seem ridiculous, so I will add this config option.
If you wanted to reduce the resources needed to compile Gambit, you might be better off setting
[lindv2:~/Desktop] lucier% diff -p /usr/local/Gambit-C/include/ gambit.h gambit.h *** /usr/local/Gambit-C/include/gambit.h Thu Jan 19 15:17:50 2006 --- gambit.h Wed Jan 25 13:03:48 2006
*** 1192,1197 **** --- 1192,1198 ----
- Determine optimization goal.
*/
- #define ___OPTIMIZE_SPACE
Can I assume that this macro is effective for all gsc compiles, not just for the build of Gambit-C itself? If the former, I'd be reluctant to add it since most people have quite a bit of memory these days.