Re: Bignum modular types in Ada95

["Nick Roberts" <Nick.Roberts@dial.pipex.com>]

Point taken, but I have another question.  Traditional bignums (e.g. in
LISP) are dynamically extensible: they can change size during run-time to
fit the actual value stored.  For some applications, this approach would
make sense, but for others it may not.  So what approach to take:
dynamically resizing, statically sized, or both?  And if both, how to
choose (depending on the range, by a representaion clause, by a pragma, or
something else)?

-- 

== Nick Roberts ================================================
== Croydon, UK                       ===========================
==                                              ================
== Proprietor, ThoughtWing Software                   ==========
== Independent Software Development Consultant            ======
== Nick.Roberts@dial.pipex.com                              ====
== Voicemail & Fax +44 181-405 1124                          ===
==                                                            ==
==           I live not in myself, but I become               ==
===          Portion of that around me; and to me             ==
====         High mountains are a feeling, but the hum        ==
=======      Of human cities torture.
===========                             -- Byron [Childe Harold]


Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk> wrote in article
<34D04FFD.41C6@cl.cam.ac.uk>...
> Robert A Duff wrote:
> > >Handling 1024-bit integer arithmetic in the Ada compiler and not in
> > >some library package has the advantage that the compiler will later
> > >be able to do much better optimization (e.g. automatic register
> > >allocation), once we get CPUs with 1024-bit integer registers and
> > >ALUs, which I expect to happen in the next three years.
> > 
> > Are you talking about special-purpose hardware?  I doubt if 1024-bit
> > registers will exist in general-purpose computers any time soon.  (I
> > reserve the right to redefine "soon" at will.)
> 
> No, I am talking about the standard off-the-shelf Pentium
> successor in a few years, not about any exotic special hardware.
> IPv6 and electronic commerce will make it necessary that normal
> workstations can to thousands of 1024-bit modexp operations per
> second for authentication protocols. This is commonly expected to
> be the next major functional extention after MMX.
> 
> 20-dollar smartcard microcontrollers have such 1024-bit registers/ALUs
> already available today. It is just a matter of time until we
> see them in workstation processors.
> 
> > Anyway, having the feature "built in" gives other advantages: literals,
> > range checking, case_statements, etc.  None of that works with some
> > library package (unfortunately).
> 
> Agree. Dear Ada compiler developers, please have a look again at
> builtin bignum support!