Re: clisp - general interest in code cleanup

Hi Sam,

2016-03-21 23:46 GMT+01:00 Sam Steingold <[email protected]>:
>
> Hi Daniel,
>
> > * Daniel Jour <[email protected]> [2016-03-21 15:07:02 +0000]:
> >
> > the CLISP project seems to be "not that active", judging from the
mailing
> > list / bug tracker / repository activities. The latest release is 6
years
> > old.
>
> Alas, this is, indeed, the case.
>
> > Looking at the code base, I think this low activity stems at least in
part
> > from the IMO relatively bad condition it is in. Most of it seems to be
> > quite old, there are (now talking of the C parts) parts that invoke
> > undefined behavior (general pointer comparisons, integer overflows, out
of
> > bounds indexing) and some other odd things (the whole additional
> > preprocessing, pre C99 code, indexing from 1 instead of 0, apparently no
> > internal interfaces / abstraction, probably more).
> >
> > I'd like to change that. In order to do so I'd like to "modernize" the
code
> > base, starting with the C code. That means among other things
converting to
> > standard compliant code (C99 probably), splitting code into smaller
units,
> > establishing internal interfaces.
> >
> > Before putting (a lot) more work into this I'd like to know whether
> > a) there's is a general interest to make such - probably huge - changes
to
> > the code base, and whether
>
> Yes, there is at least some interest in modernizing CLISP code base.
> Note, however, that this is a fairly large task, and you are not the
> first one to volunteer to do this (previous attempts failed because
> people bit more than they could chew).
> This is why I am wary of this being the first CLISP project you attempt.

That's great. I agree, modernizing the complete code base is a task that's
too huge to attempt as a first CLISP project. What I hoped for was that it's
ok if, picking a random example, readtable related code got extracted from
src/io.d, i.e. it's ok when runtime code is refactored to some degree.

> > b) this could be seen as a project (with further details I'd work out
> > then) suitable for GSoC.
>
> Let us start small - or, at least, smaller.
>
> Specifically, make a release.

Sounds good.

> This will require you to learn the current CLISP infrastructure which
> will be absolutely critical when you move on to the next step of
> cleaning up the code.
> This will also allow us, the CLISP maintainers, to pass on our expertise
> to you and to monitor and mentor your progress.
> It is also a finite task with a clear and visible deliverable.

I think I have already a rough understanding of the infrastructure, in the
sense of "where is what" and how it fits together. (Apart from a lot of
code I read the complete implementation notes and was searching
through the clisp-devel list for info).
Though there's still a lot I don't understand :)

> Currently, CLISP hg tip is in a bad state: an import of gnulib packages
> resulted in a code base which probably works only on Linux.
>
> Specifically, what you would need to do is
>
> 1. update all gnulib imports

To what extend is CLISP using gnulib atm? I've found the
gnulib-imported target in Makefile.devel (as well as related stuff), but am
unsure how far the actual src/* files make use of it. As far as I can tell
only
src/lispbibl.d, src/_clisp.c and src/glmalloc.c seem to use it (only taking
the
runtime into approach, not any modules).

I didn't know about the patches section of sf.net until Ken Brown brought
it up (https://sf.net/p/clisp/mailman/message/34957653/).

> 2. fix gnulib/Windows interaction (gnulib assumes FD to be an int,
>    Windows uses Handles)

Can you provide a bit more background on that? Is this related to
http://lists.gnu.org/archive/html/bug-gnulib/2016-03/msg00033.html and
thus a general gnulib issue? Or does this only affect CLISP?

> 3. make sure that CLISP builds and passes self-tests on as many
>    platforms as possible, but at least
>    A. Linux (fedora, ubuntu &c)
>    B. Windows (whatever the most common version is now, with and without
>       cygwin)
>    C. Mac OS X
>    D. *BSD (FreeBSD, OpenBSD, NetBSD)

I should be able to get access to all of these:
I'm working on a Linux machine, NixOS distribution (if it compiles here
it'll
probably compile one every Linux, given correct dependencies) and have
access to Debian machines at my university. I have NetBSD and FreeBSD
qemu images available, and can probably get access to both Windows VMs
as well as "real machines" through my university. I can probably also get
access to Mac computers (there's a lab / pool at the university).

How "extensive" are these self-tests?

> 4. Make an official source and binary release.
>
> I think the above is a good GSoC project.

I guess I should get a proposal ready ASAP then.

> Bruno, WDYT?
>
> > I'm asking these - probably too detailed - questions on this list here
> > instead of on one of the CLISP mailing lists because they seem, as
already
> > pointed out, to be dead. I was unsure about whether to CC one of the
> > maintainers, because it's discouraged in some communities, and so I
didn't.
> > I didn't CC one of the maintainers because such CCing is discouraged in
> > some communities. What's the "correct" way to get in contact here? And:
Are
> > there any possible mentors for the CLISP project?
>
> Yes, I am prepared to mentor you.
> I hope that if you will show some progress, Bruno will chip in too.

Thank you a lot. I'll give my best to get a suitable proposal ready
until the deadline.

Daniel

> > Finally, a very brief info about me: My name is Daniel Jour nee Oertwig,
> > I'm currently pursuing my computer science master's degree at University
> > Wuerzburg, Germany.
> > I'm a self taught programmer, started learning about 14 years ago and
have
> > been doing all kinds of different things, like some hobby OS kernels,
> > programming language ideas, an orbit trajectory calculation tool (I did
my
> > bachelor's degree in aerospace computer science :) ), C++ code
> > transformation tools and a lot more. Thus I've been exposed to a lot of
> > technologies, but I'm most proficient with C, C++ and (Common) Lisp and
> > operating system concepts.
> > I've been working as a teaching assistant for my university's operating
> > systems course, teaching mostly stuff related to multithreading
(examples
> > in Java) and GNU/Linux usage. I'm an active user on stackoverflow.
> > I'm 24 years old, married, two kids (and I should note that my family
takes
> > precedence over anything else, thus I'm probably an a bit "riskier"
> > potential GSoC student)
> >
> > Best regards,
> > Daniel Jour
>
> --
> Sam Steingold (http://sds.podval.org/) on Ubuntu 15.10 (wily) X
11.0.11702000
> http://www.childpsy.net/ http://think-israel.org http://iris.org.il
> http://palestinefacts.org http://americancensorship.org http://dhimmi.org
> My plan for 'after work' is to retire.

------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785351&iu=/4140

_______________________________________________
clisp-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/clisp-devel