[asdf-devel] new asdf does not like matlisp.asd

[Robert Goldman]

On 4/11/10 Apr 11 -11:42 AM, Faré wrote:
> 1- Would making unix-dso a subclass of component but not of module
>    somehow help solve the problem?

<blush>I didn't even carefully read the system definition --- I just dug
in to trying to make it load.  You are right --- I don't see any reason
why this should be a subclass of module instead of component.</blush>

> 
> 2- I admit I have little familiarity with TRAVERSE. Looking at it
>    makes me shudder. I really don't want to try understanding it enough
>    to be able to modify it in ways that I can say for sure will be
>    backwards compatible with all previous reasonable uses.
>    It's one of the reasons I'd rather spend time on XCVB.

Right.  I /have/ tried to understand and modify it, and there are still
a few corners that I truly don't understand.
> 
> 3- To debug with a 32-bit implementation on a 64-bit architecture,
>    give the -m32 flag to gcc. Dunno exactly about ld; same flag or different
>    flag might be needed. Otherwise, you can install a 32-bit chroot
>    installation and run stuff inside it.

Thanks.

> 
> 4- Is there an ILC this year? Where, when? I could talk about ASDF 2,
>    but I wouldn't be able to talk competently about TRAVERSE. If some
>    kind soul would co-present a paper with me, we could do the whole thing.

I would be delighted to coauthor a paper with you about ASDF.  I thought
the ILC was going to be in Japan.  If so, I'm sorry, but my back and
height put Japan out of range by coach flight, so no co-presenting!  But
I've been dying for an excuse to go to Europe for a lisp symposium! ;-)

Best,
r