how to add two no. of 100 digits or more?

[Richard Heathfield]

Tak said:

I confess you are right.
but should we say "sorry , we can't help you, please goto comp.lang.c+
+ where may help you"?

Firstly, the OP posted in comp.lang.c, so we may presume in the absence
of further information that he is programming in C. It was not he, but
you, who introduced C++ into the thread.

Secondly, you /were/ told that there is a group appropriate for C++
discussions. See message <[email protected]> where
Flash told you:

"Please read the name of the group again. This group is for C, there is
a separate group for C++."

It is true that he didn't actually mention comp.lang.c++ by name. I
expect he thought you could figure that out for yourself.

 

[Tak]

Tak said:



Firstly, the OP posted in comp.lang.c, so we may presume in the absence
of further information that he is programming in C. It was not he, but
you, who introduced C++ into the thread.

Secondly, you /were/ told that there is a group appropriate for C++
discussions. See message <[email protected]> where
Flash told you:

"Please read the name of the group again. This group is for C, there is
a separate group for C++."

It is true that he didn't actually mention comp.lang.c++ by name. I
expect he thought you could figure that out for yourself.

Richard,forgive my coming unpolite question, I find that you are very
familiar with the google group,how long have you been here? and Why
are meteorologists so lousy at their job?

 

[Richard Heathfield]

Tak said:

Richard,forgive my coming unpolite question, I find that you are very
familiar with the google group,how long have you been here?

This isn't a Google Group. This is Usenet. Calling it a Google Group is
like calling the sky "Heathrow Airport" just because you happened to
board a plane there.

But to answer your question, my first post to comp.lang.c was in 1998,
which is before Google was even incorporated. "Google Groups" came into
existence some years later, in 2001. The comp.lang.c newsgroup was here
long before Google Groups existed, then - and no doubt it will still be
here long after Google Groups has died.

and Why
are meteorologists so lousy at their job?

<grin> That one's easy - the weather is a chaotic system. Weather
prediction is basically impossible, so they're onto a loser. (To be
fair, they do of course know this, and still do the best they can.)

 

[Tak]

But to answer your question, my first post to comp.lang.c was in 1998,
which is before Google was even incorporated. "Google Groups" came into
existence some years later, in 2001. The comp.lang.c newsgroup was here
long before Google Groups existed, then - and no doubt it will still be
here long after Google Groups has died.

Oh,I even don't know what is Internet in 1998.

 

[Richard Heathfield]

Tak said:

Oh,I even don't know what is Internet in 1998.

In fact, comp.lang.c dates back to at least 1983, although it was called
net.lang.c at first (or quite possibly net!lang!c). The Internet kind
of evolved, so it's hard to pin a date on its creation, although that
doesn't stop lots of people insisting on particular dates.

 

[Jean-Marc Bourguet]

Tak said:


In fact, comp.lang.c dates back to at least 1983, although it was called
net.lang.c at first (or quite possibly net!lang!c).

I don't think so (but I wasn't there). I think that ! where used for
explicit routing of email (so you gave all the computers via which it had
to be transfered) with uucp. news could also be propagated via uucp, but
there is no interest in specifying a routing. You still have trace of this
usage in the Path: header, which give you the route by which the post has
reached you.

The Internet kind of evolved, so it's hard to pin a date on its creation,
although that doesn't stop lots of people insisting on particular dates.

http://www.computerhistory.org/exhibits/internet_history/

is probably a good start.

Yours,

 

[Robbie Hatley]

Here's a program at Robert Munafo's site:
http://home.earthlink.net/~mrob/pub/perl/hypercalc.txt
which operates on *much* bigger numbers than the
tiny ones you mention It's written in perl;
sorry if that makes it off-topic here.

Much, much larger than 10^2000000000? Sheesh, that's one
damn big number!

::: looks at site :::

The biggest number that calculator can handle is...
10^(10^(... ^(10^300) ...)) (with 10000000000 10s).
Exponential recursion 10000000000 levels deep. Wow.
I can't even comprehend how big that is.

And yet, I wonder how many of those digits are significant?
Probably about 10 digits of actual precision.

My system, on the other hand, offers 2147483647 significant
digits of precision. (Or it would if I had at least three
1-gibibyte RAM modules in my machine, instead of just one,
as I currently have.)

I was able to calculate the first 10,000 digits of the
fibonacci sequence -- out to 2092 sig figs -- in about
1 second last night.

A very brief sample:

7654090467756936378415884538348976340768064993978954512095813
12384578529797304192493293627316781267732493780359086838016392
20038668997554240570909178165665757608500558774338041350112205
32423247527351544763402471792982538876233052554697128188128597
52461916524905785334311649958648296484733611329035169538240802
84885164052257330097714121751630835360966663883732297726369399
137347080577163115432025771710279131845700275212767467264610201
222232244629420445529739893461909967206666939096499764990979600
359579325206583560961765665172189099052367214309267232255589801

For a good time, browse around at Mr. Munafo's
site. He has some interesting pages, including
mention of some integers which are too
large even for his hypercalc program.

Like I say, the precision of that calculator doesn't impress me.
The SIZE of the numbers, yes; but not the precision (number of
SIGNIFICANT digits). I doubt the calculator could handle
even the nine integers above.

Now, as for Mr. Munafo's web site...

http://home.earthlink.net/~mrob/pub/math/largenum.html

::: looks at site :::

Hmmm... He looks a lot like my coworker, Ron. Is receding
hairlines genetically linked to computer-programming ability? ;-)

Nice chart combining both the SI and IEC-binary prefixes.
I think I want a 1-ZettaHertz computer with 37 YobiBytes of RAM.

Cool site altogether. I bookmarked it for future reference.
Thanks for the link.

--
Cheers,
Robbie Hatley
East Tustin, CA, USA
lonewolf aatt well dott com
triple-dubya dott tustinfreezone dott org

 

[Richard Tobin]

I think that ! where used for
explicit routing of email (so you gave all the computers via which it had
to be transfered) with uucp. news could also be propagated via uucp,

Usenet news was almost entirely uucp until the late 80s, though I
think some sites received magnetic tapes each day.

but
there is no interest in specifying a routing. You still have trace of this
usage in the Path: header, which give you the route by which the post has
reached you.

And you will notice that the path usually ends "not-for-mail", to stop
you trying to send mail back along the path, which used to be
possible.

-- Richard

 

[Richard Heathfield]

Jean-Marc Bourguet said:

I don't think so (but I wasn't there). I think that ! where used for
explicit routing of email

Oh yes, of course you're right.

http://www.computerhistory.org/exhibits/internet_history/

is probably a good start.

Well, it certainly illustrates my point. Even assuming that that history
is canonical, it would still be hard to give a particular date and say
"on this day" (or even in this *year*), "the Internet did not exist,
but by the next, it did". Somewhere between 1977 and 1983 would be my
guess.

 

[Charlton Wilbur]

RH> Jean-Marc Bourguet said:

RH> Well, it certainly illustrates my point. Even assuming that
RH> that history is canonical, it would still be hard to give a
RH> particular date and say "on this day" (or even in this
RH> *year*), "the Internet did not exist, but by the next, it
RH> did". Somewhere between 1977 and 1983 would be my guess.

It depends on what you mean by "Internet," and what definition you use
to distingush what-preceded-the-Internet from what-is-the-Internet,
and whether it was sufficient for the system to be specified, working,
or in broad use.

Once you agree on those criteria, pinning down a date to within a few
weeks should be a simple matter of research.

Charlton

 

[Keith Thompson]

In fact, comp.lang.c dates back to at least 1983, although it was called
net.lang.c at first (or quite possibly net!lang!c). The Internet kind
of evolved, so it's hard to pin a date on its creation, although that
doesn't stop lots of people insisting on particular dates.

As has already been noted, foo!bar!baz style paths were used for
e-mail routing, not newsgroup names.

The net.lang.c newsgroup was created October 21, 1982. Google has the
announcement, which apparently was the very first message on the
newsgroup, in its archives.

The group was renamed from net.lang.c to comp.lang.c as part of the
"Great Renaming". The name change was announced November 7, 1986.

Note that Usenet was *not* originally part of the Internet. The
Internet is, by definition, a network that communicates via the
Internet Protocol (IP, often TCP/IP, but there are other
sub-protocols). In the early days, Usenet was transmitted mostly via
UUCP, and in some cases, as Richard Tobin mentioned, via magnetic
tapes.

(I never posted to net.lang.c. My first posting to comp.lang.c,
according to Google, was on August 5, 1989. My first posting to
comp.lang.c that I would now consider topical was on April 23, 1996
(and that was part of a flame war cross-posted to 4 different
newsgroups).

 

[user923005]

I confess you are right.
but should we say "sorry , we can't help you, please goto comp.lang.c+
+ where may help you

Send him your answer by email, or refer him to a group where your
answer is topical.

 

[Dik T. Winter]

> On 6=D4=C21=C8=D5, =CF=C2=CE=E711=CA=B134=B7=D6, Richard Heathfield <r...@s=
> ee.sig.invalid> wrote:
>
>
> Oh,I even don't know what is Internet in 1998.

It is even much older. And Usenet is not Internet. In about 1983 we
received Usenet but had not yet access to Internet because the US
NSF would not allow non-US and non-Canadian sites on the Internet at
that point of time. The oldest article by me I can find in net.lang.c
dates from 1985.

 

[Ben Bacarisse]

It is even much older. And Usenet is not Internet. In about 1983 we
received Usenet but had not yet access to Internet because the US
NSF would not allow non-US and non-Canadian sites on the Internet at
that point of time.

UCL[1] was linked to Arpanet sometime in the late seventies[2]. By 1980
the connection was real service rather than a research toy. By the
time I started there in 1983 it was just part of the infrastructure.

You many be thinking of the fact that, as I remember it, the NSF
started to get involved in the funding in around 1983 (it was all
DARPA and UK sources such at BT up until then).

[1] University College London (i.e. in Europe).
[2] Initially the connection was part-time.

 

[Dik T. Winter]

> In article <[email protected]>,

>
> Usenet news was almost entirely uucp until the late 80s, though I
> think some sites received magnetic tapes each day.

Australia received its news and mail feed from UCB by magnetic tape.

>
> And you will notice that the path usually ends "not-for-mail", to stop
> you trying to send mail back along the path, which used to be
> possible.

Not necessarily. There always have been paths that were usable for
news only (or for mail only). Also there were edges in the paths that
were not bidirectional. So using the path never made sure that your
mail message would arrive. That is why you see in old Usenet messages
nearly always in the signature some mention of a mail-path. My first
one was:
decvax!mcvax!turing!dik
followed in December 1984 by:
{decvax|seismo|philabs}!mcvax!turing!dik
because by that time we had three possible feeds from the US (all still
dial-up). The need of the local machine name was removed in 1985 (NFS
came in use). April 1986 we got a first fixed link, but NSF did not yet
allow us on Internet, so it became:
dik%[email protected]
and finally in early 1987
(e-mail address removed)
which it still is (the first domain outside the US and Canada).

(There is a lot of history in my signatures, and also in mcvax, a
VAX-780 with serial number 38. See <www.godfatherof.nl>.)

 

[Dik T. Winter]

> > It is even much older. And Usenet is not Internet. In about 1983 we
> > received Usenet but had not yet access to Internet because the US
> > NSF would not allow non-US and non-Canadian sites on the Internet at
> > that point of time.

>
> UCL[1] was linked to Arpanet sometime in the late seventies[2]. By 1980
> the connection was real service rather than a research toy. By the
> time I started there in 1983 it was just part of the infrastructure.
>
> You many be thinking of the fact that, as I remember it, the NSF
> started to get involved in the funding in around 1983 (it was all
> DARPA and UK sources such at BT up until then).

The DARPA and UK-sites were mainly defense oriented. So even many
research institutes in the UK could not get Internet access. There
were good reasons that when the European network started with cwi.nl,
there were many links from UK Universities through our US link. When
NSF took it over it escaped from the defense orientation. If I
remember correctly we had feeding links to Scandinavia, the UK, Germany,
France, Greece, and a few more. And before the Internet link they were
UUCP links for mail and usenet.

 

[Ben Bacarisse]

It is even much older. And Usenet is not Internet. In about 1983 we
received Usenet but had not yet access to Internet because the US
NSF would not allow non-US and non-Canadian sites on the Internet at
that point of time.

UCL[1] was linked to Arpanet sometime in the late seventies[2]. By 1980
the connection was real service rather than a research toy. By the
time I started there in 1983 it was just part of the infrastructure.

You many be thinking of the fact that, as I remember it, the NSF
started to get involved in the funding in around 1983 (it was all
DARPA and UK sources such at BT up until then).

The DARPA and UK-sites were mainly defense oriented.

I don't recall any defence oriented sites in the UK at that time. In
around 1973-74 when UCL got its IMP[1] there certainly were none shown
on the map.

So even many
research institutes in the UK could not get Internet access. There
were good reasons that when the European network started with cwi.nl,
there were many links from UK Universities through our US link. When
NSF took it over it escaped from the defense orientation. If I
remember correctly we had feeding links to Scandinavia, the UK, Germany,
France, Greece, and a few more. And before the Internet link they were
UUCP links for mail and usenet.

I don't want to detract from the excellent work that CWI has done. It
is a fine institution (I have been a visitor there on more than one
occasion). I just wanted to correct the idea there were no "Internet"
sites outside the USA and Canada before about 1983. To that end, I
should add that Norway (at the NUCC in Kjeller) also had a node in
about 1973 or '74.

[1] Interface Message Processor (for those under 45 years old!).
Originally this did not process IP packets as we know them today, so
you could, quite legitimately, say that this was not the "Internet"
(although it was an internet). The ARPA network made a transition to
TCP/IP gradually sometime around 1981-82. I don't know when the first
transatlantic IP tests were done, but I'd guess it was sometime in
1981 or maybe 82. Of course, CWI may have been doing IP tests at
about the same time.

 

[Tim Smith]

The net.lang.c newsgroup was created October 21, 1982. Google has the
announcement, which apparently was the very first message on the
newsgroup, in its archives.

Interesting. Anyone still here that goes back that far? The earliest
Google has for me in net.lang.c is from late 1984:

<http://groups.google.com/group/net.lang.c/msg/ba3197a79f1cc20f?dmode=sou
rce>

 

[Mark McIntyre]

Interesting. Anyone still here that goes back that far? The earliest
Google has for me in net.lang.c is from late 1984:

<http://groups.google.com/group/net.lang.c/msg/ba3197a79f1cc20f?dmode=sou
rce>

sadly. my first identifiable posting was in bit.listserv.infonets back
in 1992. In CLC I can't go back before 2001 as I can't remember the
posting personal pseudonym I used when I was an undergrad in the late
eighties... :-(
--
Mark McIntyre

"Debugging is twice as hard as writing the code in the first place.
Therefore, if you write the code as cleverly as possible, you are,
by definition, not smart enough to debug it."
--Brian Kernighan

 

[Dik T. Winter]

> > The DARPA and UK-sites were mainly defense oriented.

>
> I don't recall any defence oriented sites in the UK at that time. In
> around 1973-74 when UCL got its IMP[1] there certainly were none shown
> on the map.

Well, that is as I did understand it. I think they got links because
DARPA thought they had their uses. It was all mainly DARPA controlled.
Anyhow, we were refused access to their sattelite link, needed to link
to their network. It was certainly *not* an open network. And I know
for certain that from some UK universities mail sent to UCL took a
path through us to the US, where they arrived on ARPAnet and hence
came back to UCL. If my memory is good enough, the University of Kent
was the main contact between the UK and the remainder of the world.

> To that end, I
> should add that Norway (at the NUCC in Kjeller) also had a node in
> about 1973 or '74.

And there were some additional sites in Scandinavia linked with
BITNET/EARN (IBM's network), and presumably also DECNET links. All
very private. We *had* an EARN connection (HAMCWI6 or something
like that), but did not really use it. And from that time I remember
the tale how a mail message sent from one computer to another in the
same room somewhere in Scandinavia went through the US where there was
a link between the two different networks.

> [1] Interface Message Processor (for those under 45 years old!).

Ah, yes, Bolt-Beranek-Newman. Without one you could not connect.

> The ARPA network made a transition to
> TCP/IP gradually sometime around 1981-82. I don't know when the first
> transatlantic IP tests were done, but I'd guess it was sometime in
> 1981 or maybe 82. Of course, CWI may have been doing IP tests at
> about the same time.

No, we were definitely later. The main reason was the link to the US.
Initially we used 300 Bd auto-diallers (that were illegal in the Netherlands),
to reduce cost. It was better to connect at night, but we did not have
the resources to have someone sitting at night to establish connections.
Later we got a leased line from AT&T, but even that had its problems
(do not ask about AT&T billing). But as soon as that was established,
TCP/IP was used. And even the use of TCP/IP was problematical as the EU
had decreed that OSI networking was the way to go. We now all know what
the result of that is.