FYI, the last item in his thesis is computing multi-precision factorial on
the Cray.
All digits of 1000! can be computed in 28ms on the Cray Y_MP:
1000! = 40238726007709377354370243392300398571937486421071463254379991
04299385123986290205920442084869694048004799886101971960586316668729
94808558901323829669944590997424504087073759918823627727188732519779
50595099527612087497546249704360141827809464649629105639388743788648
73371191810458257836478499770124766328898359557354325131853239584630
75557409114262417474349347553428646576611667797396668820291... (L = 428)
time = 0.0279850799440299s clocks = 4664180
Here is the same calculation on today's laptop using Go and Robert's go
library big integer code:
celeste:fact mtj$ fact
1000! =
402387260077093773543702433923003985719374864210714632543799910429938512398629020592044208486969404800479988610197196058631666872994808558901323829669944590997424504087073759918823627727188732519779505950995276120874975462497043601418278094646496291056393887437886487337119181045825783647849977012476632889835955735432513185323958463075557409114262417474349347553428646576611667797396668820291207379143853719588249808126867838374559731746136085379534524221586593201928090878297308431392844403281231558611036976801357304216168747609675871348312025478589320767169132448426236131412508780208000261683151027341827977704784635868170164365024153691398281264810213092761244896359928705114964975419909342221566832572080821333186116811553615836546984046708975602900950537616475847728421889679646244945160765353408198901385442487984959953319101723355556602139450399736280750137837615307127761926849034352625200015888535147331611702103968175921510907788019393178114194545257223865541461062892187960223838971476088506276862967146674697562911234082439208160153780889893964518263243671616762179168909779911903754031274622289988005195444414282012187361745992642956581746628302955570299024324153181617210465832036786906117260158783520751516284225540265170483304226143974286933061690897968482590125458327168226458066526769958652682272807075781391858178889652208164348344825993266043367660176999612831860788386150279465955131156552036093988180612138558600301435694527224206344631797460594682573103790084024432438465657245014402821885252470935190620929023136493273497565513958720559654228749774011413346962715422845862377387538230483865688976461927383814900140767310446640259899490222221765904339901886018566526485061799702356193897017860040811889729918311021171229845901641921068884387121855646124960798722908519296819372388642614839657382291123125024186649353143970137428531926649875337218940694281434118520158014123344828015051399694290153483077644569099073152433278288269864602789864321139083506217095002597389863554277196742822248757586765752344220207573630569498825087968928162753848863396909959826280956121450994871701244516461260379029309120889086942028510640182154399457156805941872748998094254742173582401063677404595741785160829230135358081840096996372524230560855903700624271243416909004153690105933983835777939410970027753472000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000
time = 0.000501814s
code:
package main
import (
"fmt"
"math/big"
"time"
)
func main() {
const n = 1000
var f big.Int
t := time.Now()
fmt.Printf("%d! = %v\n", n, f.MulRange(1, n))
fmt.Printf("time = %vs\n", time.Now().Sub(t).Seconds())
}
On Tue, Feb 5, 2019 at 3:06 PM Michael Jones <michael.jo...@gmail.com>
wrote:
> ha ha! yes indeed. i was counting intermediaries, so off by one. wirth was
> on robert's thesis committee. (A Programming Language for Vector Computers,
> his auto vectorizing Oberon subset for the Cray-YMP.)
>
> On Tue, Feb 5, 2019 at 10:55 AM Jan Mercl <0xj...@gmail.com> wrote:
>
>> Just an of-by-one error. After all, Michael is also a programmer ;-)
>>
>> On Tue, Feb 5, 2019, 19:51 Dan Kortschak <d...@kortschak.io> wrote:
>>
>>> Robert Griesamer *is* Niklaus Wirth?
>>>
>>> On Tue, 2019-02-05 at 09:19 -0800, Michael Jones wrote:
>>> > Go learns from Oberon via Go and Oberon insider Robert Griesemer,
>>> > whose
>>> > Wirth-number is zero.
>>> >
>>> > On Tue, Feb 5, 2019 at 3:47 AM Gerard <gvdsch...@gmail.com> wrote:
>>> >
>>> > >
>>> > > Hello everyone. There has been one issue in Go that has never
>>> > > gotten out
>>> > > of my head, so it went on and on. The problem is modules.
>>> > >
>>> > > In the beginning there was Oberon. Let's just face it. Oberon was a
>>> > > brilliant designed piece of engineering. Oberon did have some
>>> > > marvelous
>>> > > features that just don't exist today, such as GC for everything,
>>> > > including
>>> > > closing files (anyone ever seen that), and their memory system was
>>> > > also GC,
>>> > > including modules entirely. So they made a counter for each module
>>> > > that was
>>> > > being used. You add one, the counter increase. You lose one, the
>>> > > counter
>>> > > went down and when that counter went zero that module got erased
>>> > > from
>>> > > memory. Pretty clever.
>>> > >
>>> > > Oberon also got very small compiled modules. They were compiled and
>>> > > the
>>> > > API was being check-summed. And they didn't got generics ;-) There
>>> > > was no
>>> > > need for that since the basic types were so simple, a lot simpler
>>> > > than in
>>> > > Go.
>>> > >
>>> > > Why were they using this compiled API file? That was only used for
>>> > > identification. Everything that has public code goes into the
>>> > > compiled
>>> > > header file. There is AFAIK no no linking. The only thing that is
>>> > > being
>>> > > used are the compiled modules and compiled header files.
>>> > >
>>> > > Now I have explained everything that I know that I know about
>>> > > modules.
>>> > > What are the areas of interest? The only answer that I can find out
>>> > > is OS
>>> > > design. But for that the benefits are huge, but only if you have
>>> > > the guts
>>> > > to really gutter the whole thing down.
>>> > >
>>> > > What compiles:
>>> > > That is pretty easy. Everything that is public will be part of a
>>> > > header
>>> > > file, the rest stays inside the module itself.
>>> > >
>>> > > The benefits:
>>> > >
>>> > > 1. This maps a lot better for OS development.
>>> > > 2. No linking involved.
>>> > > 3. updates could have been "on the fly", with just a couple of
>>> > > LOC you
>>> > > can download and compile an entire module, as long as the API
>>> > > hasn't
>>> > > changed.
>>> > > 4. Fit well with systems such as apt-get, GNU GUIX, but also go
>>> > > get.
>>> > >
>>> > >
>>> > > The downsides:
>>> > >
>>> > > 1. This could confuse people who tend to use it. How can you use
>>> > > it?
>>> > > That is why I think that this could probably only work for OS
>>> > > design. You
>>> > > just don't want to download a half baked module.
>>> > > 2. It could have been used proprietary. Personally I have a lot
>>> > > against proprietary code.
>>> > >
>>> > >
>>> > > Questions:
>>> > >
>>> > >
>>> > >
>>> > > Links:
>>> > >
>>> > > 1. http://members.home.nl/jmr272/Oberon/ModToOberon.pdf
>>> > > 2. https://en.wikipedia.org/wiki/Oberon_(programming_language)
>>> > > 3. http://www.ethoberon.ethz.ch/WirthPubl/ProgInOberon.pdf
>>> > >
>>> > >
>>> > > --
>>> > > You received this message because you are subscribed to the Google
>>> > > Groups
>>> > > "golang-nuts" group.
>>> > > To unsubscribe from this group and stop receiving emails from it,
>>> > > send an
>>> > > email to golang-nuts+unsubscr...@googlegroups.com.
>>> > > For more options, visit https://groups.google.com/d/optout.
>>> > >
>>> >
>>> > --
>>> >
>>> > *Michael T. jonesmichael.jo...@gmail.com <michael.jo...@gmail.com>*
>>> >
>>>
>>> --
>>> You received this message because you are subscribed to the Google
>>> Groups "golang-nuts" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to golang-nuts+unsubscr...@googlegroups.com.
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>> --
>>
>> -j
>>
>
>
> --
>
> *Michael T. jonesmichael.jo...@gmail.com <michael.jo...@gmail.com>*
>
--
*Michael T. jonesmichael.jo...@gmail.com <michael.jo...@gmail.com>*
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to golang-nuts+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
