>>>> "Hrvoje" == Hrvoje Niksic
<hniksic(a)srce.hr> writes:
Hrvoje> "Stephen J. Turnbull" <turnbull(a)sk.tsukuba.ac.jp> writes:
> However, this is a matter of taste. For me it is in fact a
> valid argument against Common Lisp.
Hrvoje> OK, Steven, you are entitled to your opinion. But I *am*
Hrvoje> astonished that out of all the valid argument against of
Hrvoje> Common Lisp, you choose the `when' macro.
I don't think it's a big deal, as I already said; you chose to attack
it as invalid. I'm defending it as a valid criticism; that's all.
What I _do_ think is a big deal is stuff like control structures.
Defining coroutines in 6 lines of code (or whatever it was) is pretty
impressive.
Coroutines aren't threads, but if Scheme's call/cc semantics allow
that economical definition of new control structures, I'm definitely
interested.
I understand your arguments against "poor man's CL," but Michael's
twin claims that (1) you can have a rich man's CL based on a Scheme
substrate (at great expense in implementation, I assume) and that (2)
Scheme provides idioms and semantics that cannot be efficiently
provided by the CL substrates under discussion both sound plausible.
If they are both true, then the Scheme engine provides both more
immediate power and more long-run flexibility. The costs are, I guess
(1) not getting the very rich set of idioms (and their documentation
and immediate access to CL hackers familiar with them as a development
resource) provided by CL at low development cost, (2) not getting the
benefit of using standardized packages, etc.
At the moment, I'm leaning towards Scheme, FWIW, but reading this
thread with both eyes (and both minds ;-) open.
> And whether having to learn zillions of redundant (in a formal
> sense) idioms as well as huge libraries of editing functions
> just to read other people's code will repel potential new
> developers.
Hrvoje> ??????
This is just one of Martin's less important arguments against always
requiring ChangeLogs paraphrased. You've made similar comments about
Scheme idiom. In my case, and maybe that of others, if I must learn a
new language for some casual hacking I'd rather learn a parsimonious
one.
For serious developers, I don't think it's that much of an issue.
Aside from the "Naggum effect," from what I've read, I don't think any
of the contributors to this thread would actually quit XEmacs over a
choice between Scheme and CL, although there may be a lot of moaning.
(Actually, I don't even expect much of that.)
--
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences Tel/fax: +1 (298) 53-5091