Re: [boost] Review: lambda library

From: David Abrahams (david.abrahams_at_[hidden])
Date: 2002-03-22 03:09:21

----- Original Message -----
From: "Jeremy Siek" <jsiek_at_[hidden]>

> On Thu, 21 Mar 2002, Asger Alstrup Nielsen wrote:
>
> alstru> I agree with Dave that it is premature to incorporate the
current Lambda
> alstru> Library into Boost given the recent developments and
discussions.
>
> Dave did not vote against the Lambda Library, he abstained. However,
his
> message was confusing to read. Dave, would you like to clarify?

Not sure what to say; I'm ambivalent. I think the LL would be a valuable
addition to boost, but I think others working in the same area have had
insights which I'd like to see incorporated into any boost FP library we
accept. Some of the syntactic innovations in Phoenix resonate for me
immediately in ways that the LL syntax does not. The impact of this is
not to be underestimated (i.e. don't say "it's just syntactic sugar" -
if sugar didn't count we would program in machine code). Some of the
design decisions in the boost bind library (and Phoenix) seem like
better matches for the needs of real users. For my personal work,
support for MSVC6 is still important. If I need to use an FP library
with Boost.Python, I guess I'll be working with Phoenix. The
conversations that have come up during this review period have been
fascinating; it seems as though ideas new even to the LL authors have
been raised. Perhaps the process of defending a library through the
formal review process doesn't allow for the kind of deep consideration
of alternatives that I'm hoping for.

I hope this doesn't just leave everyone more confused,

Dave