Re: Do the guile test scripts work correctly for anyone ?

[Richard Frith-Macdonald]

On 1 Sep 2004, at 14:57, Pete French wrote:

> > The quick answer to the subject line is ... yes, they work on my apple
> > laptop (ppc, macosx) and my desktop machine (intel, debian).
>
> Good, that is helpful to know.
>
> > Of course, the most likely problem is that there is bug in the 64-bit
> > support in gstep-guile, gstep-base, or guile,
>
> Ahhh... sorry, I didnt make this clear I guess. I am not
> running on 64 bit - this is on bog standard 32 bit i386 Unix.
> Which is where it should work fine. But if it works fine on
> your machine then that indicates a problem with the FreeBSD
> port of something or other I guess. Which helps a lot as it means
> I am not trying to get something running which is simply broken.
>
> > I'm afraid that, without gdb, you are pretty much reduced to adding
> > instructions to print stuff at various places
>
> Debugging by friction :-) I might have a go at that later - am finding
> this one of the few occasions where I am actually hindered by being
> an electronic systems engineer rather than a computer scientist. tring
> to teach myself the lambda calculus in half an hour is a lot harder 
> than
> I had imagined! Knowing this seems to be a pre-requisite to learning
> Scheme, and that seems to be a pre-requisite to learning guile; which 
> is
> a pre-requisite to understanding the test framework. :-)

Yeah, I wanted to avoid the horrors of DejaGNU and TcL ... and my
choice of language to implement a test framework was guided by the
fact that a guile-based test framework was on the FSF projects list.
Bad move ... while guile is undoubtedly better than TcL (robust and
consistent etc), it's not a language I'll ever find intuitive or program
in for fun.

That being said, if you look at the existing testcases you will see that
the oddities of the language are largely ignored ... and tests mostly 
execute
a simple sequence of operations.