[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[IGNORABLE] Scruffies vs Neats

This page is part of the web mail archives of SRFI 48 from before July 7th, 2015. The new archives for SRFI 48 contain all messages, not just those from before July 7th, 2015.


I have been out of town and see that the historical CL Scruffies vs Neats 
discussions are still vigorous.

Some observations..

As noted, having terse, readable templates is helpful.  Backquote saves typing 
and enhances readability.  Define-syntax adds error checks and avoids most 
problems with simple macros--even if not a complete solution where 
name-capture is required.
It has been proposed that CASE be removed from the language.  If this had been 
done and variant CASE definitions used in libraries, I would argue that this 
would cause useless work in going from one Scheme to another.

While there are many, many object systems for Scheme--I have written at least 
four myself--there is no portable, standard one.  I will argue that the lack 
of windowed & GUI development environments is probably due more to lack of 
libraries of inheritable/OO widgets than with lack of a standard FFI.  
Contrast this with CLIM (CommonLisp Interface Manager) which is supported by 
all commercial CL implementations.

Getting records with subtyping into Scheme is taking decades, while record 
implementations have been around for decades.  This indicates lack of 
agreement on a process for convergence.  The SRFI process is non-binding and 
has at least engendered agreement and some implementation computability.

Two decades ago, one frequently wrote in C and some of the first code written 
in going to a new system were routines like block_move, string_compare, and 
sometimes printf.  This is not true today.  I do, however, continue to 
port/augment FORMAT, which is both silly and disheartening.

There are implementations of Format which allow for additions (Scheme48 comes 
to mind, but I know there are others).

The reason for Intermediate format is precisely to maximize agreement while 
moving toward better/more complete solutions.  It is obvious from the 
discussion so far how difficult it is to gain agreement.

Predictor/correcter baby steps at least leads somewhere.  Can we take this 
step or are you all satisfied with SRFI-28?

I'm not.

Format is here.  It exists.  It is not going away.  How about a higher minimal 

[I still don't see how people find <foo> more readable than (foo)   8^]