This is the mail archive of the guile@sourceware.cygnus.com mailing list for the Guile project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Syntatic sugar and identifier permissivity


On Wed, Apr 05, 2000 at 12:16:52AM +0200, Mikael Djurfeldt wrote:
> 
> What are you aiming at?  That we should introduce class-centered
> techniques into GOOPS?

No; that we make GOOPS easier to learn and quicker to use.

> (Your suggested syntax is reasonably consistent also with GOOPS,
> however.  The main argument against it is that it is unnecessary.)

Yes, actually I tend to consider myself a competent designer,
so I proposed something that I thought was consistent :-)

Of course it is unnecessary. Otherwise I'd have called it
"meat" or something instead of "sugar" ;-) the goal here is to
make things easier when you're learning and/or writing code.
Think of it as an UI question.


> We have very good reasons for choosing a CLOS-style object system as
> the basic OO system in Guile.  It has, or can be made to have, all of
> the capabilities of a class-centered system, but, in addition, has the
> important power of the MOP.  In GOOPS, you can customize the language
> itself.  This is an extremely valuable property of an extension
> language.  I've already seen this many times in my own projects.
> 
(...)
> 
> Also, as Jost has pointed out, it is possible to use the MOP to
> emulate a class-centered OO system on top of GOOPS.  (The other way
> around is simply not realistic.)

I realize that and have already said so in this thread. As I
said, this isn't about implementation, it's about "UI". You can
emulate the implementation, but with an UI like that it just
won't make too much of a difference ;-)

Of course I can just forget this and "wait" for the
translators, then people can code in whatever they like. But
I'd really prefer to make it easier for them to do it in Scheme.


=== META ===

> I think you're putting blame on the wrong reasons.  If you are
> concerned about getting more people using Guile, you should contribute
> to the reference manual.

No, I don't think documentation is the real problem. Look at
sawmill (and actually a lot of GNOME projects) to learn why
people don't use Guile. Yes the documentation sucks but it
isn't all about documentation. It is the fact that:

- it is a moving target (I'm speaking for myself here; I
completely shelved FEAST untill we have a good C API for the
module system and GOOPS)

- the documentation sucks, allright

- if I have to lock my users to one language, sorry but I'll
lock them to Python. Most other programmers will say the same
for Perl or TCL.


Please for a moment forget the context in which this discussion
originated and think: is Guile lacking focus or not? (Or see my
other post for the cartoon analogy)

[]s,
                                               |alo
                                               +----
--
          Hack and Roll  ( http://www.hackandroll.org )
            News for, uh, whatever it is that we are.


http://www.webcom.com/lalo           mailto:lalo@hackandroll.org
                 pgp key in the personal page

Brazil of Darkness (RPG)    ---     http://zope.gf.com.br/BroDar

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]