Re: [Chicken-hackers] bootstrapping (again)

2011-12-22 Thread Sven Hartrumpf
Hi Christian. Thu, 22 Dec 2011 10:29:20 +0100, ckeen wrote: > So the problem will occur if you install these extension with the > installed chicken? Yes. > Is this from a clean slate or are there old extensions hanging around? Old extensions were present. But I reran this test as follows: c

Re: [Chicken-hackers] bootstrapping (again)

2011-12-22 Thread Christian Kellermann
* Sven Hartrumpf [111222 10:19]: > Thu, 22 Dec 2011 10:04:28 +0100, ckeen wrote: > > Sorry if I missed something, which commands do I need to run to reproduce > > the problem? > > Just one: the chicken-install command from my mail. I saw that. I am still unsure about the correct circumstances t

Re: [Chicken-hackers] bootstrapping (again)

2011-12-22 Thread Sven Hartrumpf
Thu, 22 Dec 2011 10:04:28 +0100, ckeen wrote: > Sorry if I missed something, which commands do I need to run to reproduce the > problem? Just one: the chicken-install command from my mail. ___ Chicken-hackers mailing list Chicken-hackers@nongnu.org htt

Re: [Chicken-hackers] bootstrapping (again)

2011-12-22 Thread Christian Kellermann
* Sven Hartrumpf [111222 10:01]: > If any one can reproduce this, good. > If not, I known that I have to run such commands several times - no problem > here :-) Sorry if I missed something, which commands do I need to run to reproduce the problem? Kind regards, Christian -- Who can (make) t

Re: [Chicken-hackers] bootstrapping (again)

2011-12-22 Thread Sven Hartrumpf
Hi again. Thanks Felix for helping me with bootstrapping more easily. The following random crash type remains: Tue, 13 Dec 2011 08:54:13 +0100 (CET), hartrumpf wrote: > [panic] `##sys#error-hook' is not defined - the `library' unit was probably > not linked with this executable - execution term

Re: [Chicken-hackers] [PATCH] remove unboxing

2011-12-22 Thread Christian Kellermann
* felix winkelmann [111221 21:39]: > The attached patch removes the unboxing pass from the compiler. It > never worked reliably and has evolved into an unmaintainable > mess. This will result in a reduction of runtime performance for > unsafe floating-point intensive code. Does this obsolete the