Re: [gentoo-dev] portage-ng concurse entry Was: Updated Portage project page

[David Holm]

On Fri, 5 Dec 2003 01:58:17 -0800
George Shapovalov <george@g.o> wrote:

[BIG SNIP]

> However that's not all. I have produced some basic prototyping code to
> illustrate what could be expected. The prototype is quite crude, as I did
> this during relatively rare breaks from writing an article (completely
> unrelated to CS :)), but it should serve the purpose. Did I say the code
> shoul be readable? So, even though I do not expect many people to be familiar
> with that language I would still suggest trying to look at the code. You are
> in for a one nice surprise ;).
> (I am not revealing the name of language in this posting deliberately, because
> I want people to read through arguments first).
>
> The code is available here:
> http://dev.gentoo.org/~george/proto_portage-0.7.5.tar.bz2
> but you will probably want to read the text before that.

It is an interesting idea to implement this in Ada and I only see benefits in doing so,
especially since I'm no big fan of prolog myself. The problem at the moment however is that
GNAT-3.15p has only been ported to x86 and ppc in portage so far. I have a few problems with
porting to other architectures, the main one being having access to them and another big issue
is that threading is only implemented on a few architectures and so far only supported on x86 in
our portage version. Since you need an already working copy of GNAT to port it and the fact that
GCC 2.8.1 is becoming increasingly difficult to compile due to it's age I have not been looking
forward to the day when someone asks me to port it to another arch.
GNAT-5.x is based on GCC 3.2 which makes it much easier to port but you still need an existing
GNAT or cross-compiling GNAT for the arch and we still have the same issues with threading.
There was talk about reimplementing threading under Linux using nptl now that it has become
more widely acknowledged (GNAT requires a more advanced threading model than the standard Linux
threading model can provide it with). So far I have not seen anyone attempt to do this though
and it seems ACT (the company developing GNAT) are the only ones doing active work on it. ACT
also recently took their CVS version of GNAT-5.x offline (5.x has not been publicly released
yet) and stated that they were moving things to GCC's CVS. This is both good and bad, good
because maybe now more people feel they can submit code, bad because the reason they haven't
done this earlier is because the GCC team does not acknowledge the extremely high reliability
standards ACT have which means we might see a split where you will only be able to get a
"stable" version of GNAT by purchasing it from ACT (the commercial version of GNAT currently
costs something like $10,000).

Still, if we can overcome the portability issue of GNAT then I find this an interesting idea.
GNAT is the only freely available Ada compiler for Linux btw.

//David Holm, Ada maintainer etc