Re: [Chicken-hackers] plain lambdas as syntax transformers

[Felix]

From: Peter Bex <address@hidden>
Subject: Re: [Chicken-hackers] plain lambdas as syntax transformers
Date: Mon, 13 May 2013 12:34:16 +0200

> On Mon, May 13, 2013 at 12:09:09PM +0200, Felix wrote:
>> Peter, Joerg:
>> > But the subtle damage it does to the newbies brain when reading
>> > existing code to learn… that's bad.
>> 
>> Sorry - but that's just bullshit. We are talking about a simple
>> default here: if no explicit transformer-constructor is used, then
>> default to explicit-renaming. What's so confusing about this?
> 
> It makes it harder to explain what exactly er-macro-transformer does
> (ie, nothing?).

I'm unable to see the problem with implicitly wrapping an
er-macro-transformer around a given lambda. That's fairly easy to
understand.

> 
>> Explicit renaming is a simple model, it is older and IMHO easier to
>> understand.
> 
> It's simple, but easy to do things wrong.  In that sense it's too
> low-level.

That's just your opinion.

> 
>> Scheme48 and syntax-case both use plain lambdas instead of
>> transformer-procedures and I haven't heard anybody complaining about
>> them being the cause of brain-damage to newbies.
> 
> It's typical of the thinking that produced syntax-case to pre-empt the
> whole system as if nothing else could co-exist with it.  In fact, unless
> I'm misunderstanding, we can't support syntax-case as-is with our
> system, exactly because of this!  I don't think this should be used as
> a good example.

Isn't that what normally module-systems are used for?

> 
> That Scheme48 does it is probably due to it being so old.

Probably. Probably not. That's not a convincing argument.


cheers,
felix