Path: csiph.com!v102.xanadu-bbs.net!xanadu-bbs.net!feeder.erje.net!us.feeder.erje.net!newsfeed.fsmpi.rwth-aachen.de!newsfeed.straub-nv.de!eternal-september.org!feeder.eternal-september.org!mx02.eternal-september.org!.POSTED!not-for-mail
From: Ben Bacarisse <ben.usenet@bsb.me.uk>
Newsgroups: comp.lang.javascript
Subject: Re: Overflow, integers and strings
Date: Tue, 05 May 2015 01:10:59 +0100
Organization: A noiseless patient Spider
Lines: 50
Message-ID: <87y4l3dgto.fsf@bsb.me.uk>
References: <e7aac5c2-b226-44f4-a77f-fa284c3877ba@googlegroups.com> <6ca07e69-1694-49c1-93eb-748c44d8ab68@googlegroups.com> <XnsA48CD1D512C3Beejj99@194.109.133.246> <b378aebc-5acc-4b88-9050-6b812221cba9@googlegroups.com> <XnsA48D6DDF66B5Feejj99@194.109.133.246> <871tj08vrb.fsf@bsb.me.uk> <XnsA48D942F1A7FFeejj99@194.109.133.246> <mhvtku$rh2$1@solani.org> <1997740.h7VUlrh2Em@PointedEars.de> <mi0rm4$d4a$1@solani.org> <XnsA48E35CB52B7eejj99@194.109.133.246> <72e44d34-7f4d-4d32-8304-e73688132274@googlegroups.com> <3770518e-e91f-4368-9a69-cbd2f2461343@googlegroups.com> <a446b186-9ac4-410d-846d-3b2406a0bfd2@googlegroups.com> <87lhh5zmuy.fsf@bsb.me.uk> <9e56ec52-91b5-4e3e-9206-fd970fa69254@googlegroups.com> <2a1bdab8-bc0e-4a82-bce5-37c151ba4b47@googlegroups.com> <87zj5kebta.fsf@bsb.me.uk> <c2f87900-aeec-4d43-9e53-6968285f9f76@googlegroups.com> <87oam0e7t7.fsf@bsb.me.uk> <308b20cc-7c68-4828-badd-5213ef2107b3@googlegroups.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
Injection-Info: mx02.eternal-september.org; posting-host="017616aa25f81ec581c44d76d61ba2f3"; logging-data="20081"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX19sMHyjYzV4JqP45QgLFUfglAuLpZZheu0="
Cancel-Lock: sha1:kjOw6SvNZD83/zR+Qk7o2H0PVqQ= sha1:6KNt2jqh05ozyq/EGDi5czsm7a4=
X-BSB-Auth: 1.0f9b8c79b11ba0eeaf22.20150505011059BST.87y4l3dgto.fsf@bsb.me.uk
Xref: csiph.com comp.lang.javascript:27528

jonas.thornvall@gmail.com writes:

> Den måndag 4 maj 2015 kl. 16:28:09 UTC+2 skrev Ben Bacarisse:
>> jonas.thornvall@gmail.com writes:
>> 
>> > Den måndag 4 maj 2015 kl. 15:01:44 UTC+2 skrev Ben Bacarisse:
>> >> jonas.thornvall@gmail.com writes:
>> >> <snip>
>> >> > This is only an outline / pseudocoe for how the base conversion would
>> >> > look with bignumb implementation will look. You have to check the
>> >> > arrays so they contain intergers and not chars, that is probably what
>> >> > fail in first loop.
>> >> 
>> >> There is no value in using a bugnum for the base in a bignum
>> >> implementation.  This whole part of the project is a waste of time
>> >> (unless you just want learn ECMAScript programming but that seems
>> >> unlikely).
>> >
>> > I can see that the implementation faster multiplying in base 2048
>> > rather than base 2, although it right now cause overflows if number
>> > outside integer scope for all bases other than 10.
>> 
>> I can't follow this.
>
> There is nothing to follow i say the higher base my implementation use
> the faster arithmetic. Of course it could be due to my implementation,
> but it isn't
> it is due to the weight of the digiplaces.

What is the optimum base for a calculation like AxB?  Presumably it
depends of the sizes of A and B.  According to you, as the base
increases, the calculation gets faster.  But once the base is so large
that A, B and AxB are all single digits, making it bigger won't help.
This means there is an optimum size.  You current argument implies that
this optimum base is AxB+1 (or there abouts).  The calculation is then
just a single digit multiplication, but those digits are bignums
themselves.  What is the optimum base for *these* bignums?  Do you see
the problem now?

>> > I will fix that though, and that it will becom apparent when
>> > multiplying big numbers which approach is faster.
>> >
>> > Binary base or bignumb base.

Does this mean you don't know that a bignum base will be faster?  You
seemed to be saying that you did know.

<snip>
-- 
Ben.