Re: Emacs vista build failures

[Gilaras Drakeson]

Hi,

>> I've set up a new list, address@hidden, and subscribed these
>> addresses (you should have received welcome msgs):

> I don't mind using that address although I understand the
> topic to be larger in scope than just emacs.


[Since I am not sure if the name of that list will change, I post my
comment here. I apologize if this is not the right mailing-list.]

I really hope to see a portable build system in a reasonable lisp. To
me, "the complexity of configure scripts" reads as "the accumulated
outcome of a not so great language for the job"[*]. I hope the emergence
of ad hoc build systems around various scripting languages and also ones
using XML will inspire people to pick a (or rather, the) proper language
for a build system of the future.

A POSIX compatibility layer [**] and a large library of build/packaging
related functions/macros will make a good head start. One ought to take
a look at dh_* scripts in Debian [***] for packaging related utilities,
which is an aspect sometimes overlooked in build systems.

I shouldn't be saying this, since you all know way better than me, that
the build system is a vital component in the software distribution (and
it is even more crucial for free software). Yet it is in the sad state
that we have today.

The `bootstrapping' problem for a new build system won't be solved
unless there is a concerted effort to push such a new system until it
becomes a [de facto] standard.

Such day-dreaming can really happen sometime in the future, if at least
some of us are convinced that it is something important to have.

[*]: I am not saying that at the time the choice of the language was
made it was a mistake. I am arguing that lisp will make a much better
choice for a future system.

[**]: A Windows compatibility layer is also important to have.

[***]: Those scripts come with the package `debhelper'

-- 
Gilaras Drakeson