[r6rs-discuss] [Formal] #;<datum> comments useless

From: Per Bothner <per>
Date: Sat Sep 30 15:00:24 2006

Anton van Straaten wrote:
> Also, the objection that #; makes it harder to see where the comment
> ends applies equally to the Lisp/Scheme abbreviation for QUOTE, but this
> doesn't seem to be a problem in practice -- in fact, that abbreviation
> is one of the more useful syntactic features of Lisp & Scheme.

Well, those of us with bad eyes aren't that thrilled by having
to distinguish between "`", "'" and "," but they traditional.

> So the alleged danger of this feature seems overstated. It seems much
> closer to harmless than harmful to me.

I won't say it's very dangerous. I think on the whole the usefulness
of the feature is outweighed by (I think) making code harder to read
and (slightly easier) to make mistakes, so on the whole I think it's
a better language without it than with it. But not not much.

> The remaining objections seem to boil down to the fact that it's not an
> essential feature. That's true. Neither is the abbreviation for QUOTE.
> But #; has been considered useful enough that multiple Schemes
> implement it.

None is listed in http://srfi.schemers.org/srfi-implementers.html
but of course it is difficult keeping it up to date.

> One minor argument for specifying it in R6RS is that it
> should reduce the temptation for other implementations to use an
> alternative character sequence for expression comments (we all know what
> mavericks the Scheme implementors are...)

That is certainly a good reason with having it in R6RS. After all.
the goal isn't to design the world's best programming language, but
to design as good a language as we can within the existing Scheme
tradition.

I think we've exhausted the topic. I think it's a dubious language
feature, but if enough people like it in spite of my arguments against,
and (more importantly) enough implementations already provide it, then
there is a case for including it.
-- 
	--Per Bothner
per_at_bothner.com   http://per.bothner.com/
Received on Sat Sep 30 2006 - 15:00:00 UTC

This archive was generated by hypermail 2.3.0 : Wed Oct 23 2024 - 09:15:01 UTC