Re: [Sbcl-devel] Building from ECL fails

[Christophe R. <cs...@ca...>]

Gabriel Dos Reis <gd...@in...> writes:

> On Sun, Jul 5, 2009 at 11:41 AM, Christophe Rhodes<cs...@ca...> wrote:
>> Gabriel Dos Reis <gd...@in...> writes:
>>
>>> Note that I had to convince ECL maintainers that the change was
>>> necessary in order to get SBCL working with ECL.  Others -- as I've
>>> seen through private emails -- are interested as well.
>>
>> OK, fair enough.  (I'm slightly intrigued as to why, but I suppose I'm
>> the last person to question people's motivation in this respect).
>
> There many reasons (apparently each individual with his/her own reasons.)
> In my case, it is that CLISP is no longer a reliable to bootstrap SBCL.
> ECL, on the other hand has proven to be more portable.

Um, it should be.  I did a signficant amount of work in April to make
it reliable (see
<http://www.advogato.org/person/crhodes/diary.html?start=129>) and I
consider clisp build failures to be regressions.  (I don't know
whether ECL is more portable or not, but since cross-compiling SBCL is
so easy I'm not sure I see why that's relevant).

Best,

Christophe