Mersenne Digest Sunday, May 16 1999 Volume 01 : Number 559
----------------------------------------------------------------------
Date: Sun, 16 May 1999 04:51:57 -0800 (AKDT)
From: "Christopher E. Brown" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Getting maximum speed out of a Linux machine
On Sun, 16 May 1999 [EMAIL PROTECTED] wrote:
> With Win 9x, yes. I believe NT tends to allocate from the bottom up.
Nope, NT is top down as well. This caused an issue with early PII
chips as well, not only could they only cache 512M, but there was a bug in
the cache controller if you added more than 512 it would mirror the writes
below 512. A PII with 768 ram, and the L2 enabled would crash before it
was able to blue screen.
> Are you _sure_ this is the case - most of my development was done under NT
> but I did check the effects were similar under 9x - also I thought that
> deficient chipsets, like TX, failed to cache _any_ memory once you went
> over the 64MB cacheable memory limit?
Depands on the BIOS, a few disabled all L2 caching if more than 64
was present, however most mainboards (both VX and TX chipsets) simply
cache the first 64, and don't cache whatever else there is.
- ----
First Law of System Requirements:
"Anything is possible if you don't know what you're talking about..."
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 14:37:50 -0000
From: [EMAIL PROTECTED]
Subject: Re: Mersenne: Screen saver killers?
>
>[...snip...] Now, how to incorporate that in Prime95
>is another matter, especially as I don't think George would like too many
>zeroed-out security codes due to people running own-compiled versions.
>
I _know_ it's not as efficient in the absolute sense, but, if people would
prefer to run a screensaver _and_ a version of Prime95, I wonder if it
would be worthwhile updating the old Win 3.1 screensaver version. The point
is, we want to make it attractive for people to participate.
Perhaps we should wait a few weeks, then poll a selection of users as to
whether they prefer to participate using a "Prime95 style" background
program or a "Seti@home" style screensaver. I mean the style of interaction,
not the way they percieve the project.
>>What if you periodically had each machine copy it's P/Q files to some
>>central location on a server.
Something like this may possibly be required in the long run, as run
times on the very large exponents which v19 can cope with could be more
than the lifetime of a system! However, a few megabytes per savefile
times tens of thousands of systems is _still_ a lot of filestore, and
shifting that much data across dial-up network links is going to be
painful.
Questions have also been asked in the past about what use could be made
of a number of machines which are only available temporarily. Being able
to upload partially completed assignments would be useful from this
point of view, too.
There are technical (security) and non-technical (what happens to the
prize money if more than one PrimeNet user has worked on the assignment)
problems, but I don't see these as insuperable.
>>When an exponent is checked out by your
>>pseudo proxy, have it check if there's a partially worked on P file and send
>>that out or something. At the very least, it'll be nice to back those up on
>>occassion to prevent lost work when someone formats their drive.
I take it that's your own pseudo proxy, in which case it can do more or
less what you want it to... I would have its clients check in fairly
frequently (assuming permanent network connectivity & a reasonable
bandwidth) & upload the P file for the exponent it's working on. If a
client fails to check in twice running, the next time a client wants a
new assignment, give it the assignment and the last P file from the
"broken" machine. And log the event so someone can check out what went
wrong.
>
>How? That is the question... I guess I need to get VC++ back, enter Windows
>(yuck) and do something smart (an FTP upload, for instance). If any coders
>out there are willing to help me, it would be nice.
Doesn't sound too hard. The psuedo proxy could run OS/360 or Multics for
all the clients care. I guess all you need to do at the client end is to
find a FTP client which can be run from the command line, then have the
Mersenne client simply construct the command, fork off a copy of the
FTP client to do the file transfer, leave FTP to do its thing & get on
with crunching numbers. You might want to worry about what happens to
the client if it wants to rewrite the save file before the FTP finishes,
my guess is that the best thing is to (a) abort & restart the FTP _and_
(b) double the interval between writing save files.
Actually, for local groups, you could probably just use the ordinary OS
"file copy" command across filesystems shared using Windows, NFS, Netware
or Samba (depending on the (N)OS or mix of (N)OSes).
George - suggestion for V19 - option to run an external command (supplied
by user as value in local.ini file) at fixed intervals (time again supplied
as value in local.ini file) ? Those who think it's justifiable _could_
use this option to kill screensavers ;-)
Regards
Brian Beesley
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 16:53:52 +0200
From: Petri Holopainen <[EMAIL PROTECTED]>
Subject: Re: Mersenne: SETI on ABC News last night
Herman De Wael wrote:
>
> You have the correct argument, but you draw the wrong
> conclusion, Petri.
>
> It takes only one civilization to populate the galaxy.
> Since we find no signs of them (and they should have been
> here as well !), we might conclude we are the one that will
> do it !
>
I've heard this argument before.
Firstly, why should they have been here as well? It's a mighty
big Galaxy....
Secondly, SETI has only been going on for a few decades. Maybe we
have been listening to the wrong frequencies, or maybe we have been
listening at the wrong coordinates... Who knows? The only thing we
can be sure of is that we haven't looked hard enough yet.
Of course, there is a possibility that an interstellar civilization
would not use radiowaves, or they wouldn't be interested in contacting
"primitive" civilizations like ours. In that case, SETI would be
pointless.
But it would be pretty sad if we didn't even *try* to find them...
- -- Petri
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 09:05:16 -0600
From: "Aaron Blosser" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: Screen saver killers?
> Doesn't sound too hard. The psuedo proxy could run OS/360 or Multics for
> all the clients care. I guess all you need to do at the client end is to
> find a FTP client which can be run from the command line, then have the
> Mersenne client simply construct the command, fork off a copy of the
> FTP client to do the file transfer, leave FTP to do its thing & get on
> with crunching numbers. You might want to worry about what happens to
> the client if it wants to rewrite the save file before the FTP finishes,
> my guess is that the best thing is to (a) abort & restart the FTP _and_
> (b) double the interval between writing save files.
>
> Actually, for local groups, you could probably just use the ordinary OS
> "file copy" command across filesystems shared using Windows, NFS, Netware
> or Samba (depending on the (N)OS or mix of (N)OSes).
>
> George - suggestion for V19 - option to run an external command (supplied
> by user as value in local.ini file) at fixed intervals (time
> again supplied
> as value in local.ini file) ? Those who think it's justifiable _could_
> use this option to kill screensavers ;-)
For that matter, if there were some link to spawn an external command, you
could use the plain jane FTP that comes with Win95/98/NT to FTP the file
somewhere, or just a copy command to a UNC share.
The Windows FTP command supports scripting via -S:scriptfile, so it's
possible I suppose, and it's not a bad idea to have something like that in
Prime v19 since you could do a whole bunch of stuff we haven't thought of.
I like that idea Brian!
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 08:50:24 -0700
From: Martin Lehmann <[EMAIL PROTECTED]>
Subject: RE: RE: Mersenne: SETI on ABC News last night
> Date: Sun, 16 May 1999 00:14:18 -0600
> From: "Aaron Blosser" <[EMAIL PROTECTED]>
> Subject: RE: Mersenne: SETI on ABC News last night
>
[.....snip.....]
> Then the odds of having planets suitable for life, such as earth. I'm
> certain there are only thousands of such planets out of possible billions.
That is just your opinion/belief, so it's not a basis for a scientific
discussion ;-)
> Then, whether you believe in evolution or not, there are still pretty big
> odds against life forming anyway, but I'll gloss over that and say it
> happens but it's rare (something I don't agree with anyway). And then,
> there are the odds against any such life evolving into a form intelligent
> enough to become intelligent enough to develop transmission mechanisms.
And those odds we don't know anything about which is why it would be so
interesting to find another sentient species (to investigate those odds)!
Lately, there have been many discoveries of life in very unusual places (hot
vents on the sea floor - no sunlight, no oxygen, temperature > 250F, and
still there are some critters....just one example), so I think we basically
have to say: we don't know about the odds (although to me it looks like life
is way more robust than we all thought).
> And then, barring any fictional FTL messaging, there is the incredible
> amount of time it would take for any signals to even reach our isolated
> planet, on a distant spiral arm of a remote galaxy. And any signals would
> be incredibly weak if they even did exist, and would be barely detectable,
> if at all, against the background radiation of space (big bang radiation).
Although you are correct with the time problem, the microwave background
from the big bang is not a problem, since it is in a way different part of
the frequency spectrum. The biggest interfering radiation sources are
military radar installations and the like....
> And the fact that there is just SO much area to cover, so many signals to
> analyze.
>
> And then the fact that we've only been scanning space for any kind of
> signals for only a few decades.
>
> It's like looking for a needle (that might not exist) in one of trillions
of
> haystacks, and expecting to find that needle in the first few tries.
Nobody is seriously expecting anything. The point is that the discovery of
another advanced civilisation would be a revolution, whereas the discovery
of another Mersenne prime is, while still a great accomplishment, an
evolutionary progress. Don't get me wrong (and flood me with flame mail
;-)) - I love GIMPS and have been participating in it for the last 2.5 years
(in the beginning, I had the computer in my dorm room and left it on 24/7
just for GIMPS), but SETI is just too fascinating for me personally and I
think even the slightest chance of the aforementioned revolutionary
discovery is worth the effort.
I case I insulted/offended anybody reading this, my apologies, that was not
my intention (don't want this discussion to become emotional!)
> Naturally, I'd be blown out of my socks if we found any sort of pattern
> emanating from deep space, indicative of intelligence (or signs of
> intelligence from Earth itself :-), but I feel quite confident in saying
> that the odds are 100% in favor of us never finding any such signs before
> Sol blinks out for good.
Again, there is no scientific basis for this statement. The only thing we
know is that we don't know. And my opinion is: therefore we still have to
try.
On the other hand I have to say that the practical implementation of GIMPS
is far better than the one of SETI. Apparently they chose the route of
creating something for John Doe average Win95 user, so they're catering to a
different populace than GIMPS for example. I don't think that's wise,
because your average home user will probably grow tired of leaving his
machine on 24/7 pretty fast, no matter how fancy the displayed graphics,
which truly idealistic don't care about anyway (see GIMPS). You cannot run
SETI as a service under WinNT, for example, which is a bummer, because the
majority of the time NT boxes are idle is when nobody is logged on, and SETI
does not work as the logong screen saver.
[I managed to get it running as a service using SRVANY from the WinNT
Resource Kit, but the stupid thing has standard priority, so in order to
keep my dual machine responsive, I will use one SETI and one GIMPS client
each...and I would feel a little bad abandoning GIMPS totally after all that
time ;-)]
Furthermore, they don't have anything like our "DaysOfWork" setting, i. e.
when their server goes down, so go the clients, and that effect is even
worse than it would be for GIMPS since your average work unit apparently
does not take longer than 40hours (400MHz) and also requires a download of
300+kB. I don't understand their thinking.....they also lack a descent
mailing list like this one.
Oooops, I think I should stop here, because this *is* the GIMPS list, after
all....
Martin
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 18:04:38 +0200
From: "Henk Stokhorst." <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Screen saver killers?
[EMAIL PROTECTED] wrote:
>
> Perhaps we should wait a few weeks, then poll a selection of users as to
> whether they prefer to participate using a "Prime95 style" background
> program or a "Seti@home" style screensaver.
A working Prime95 screensaver program will certainly make it a lot easier to
distribute the program on machines in offices and networks etc. The Prime95
program we now use does do something we don't understand and does not do anything
useful. I think a screensaver could be written in such a way that it also chunks
up cycles that are spilled when the machine is used normally. Furthermore the
small image floating over the black screen should be a little attractive. If
Prime95 is a screensaver, we all know what is does do (It prevents the screen from
burning in, and at the same time it tries to win $50,000 from a bunch of math
geeks)
YiS,
Henk
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 17:12:37 +0100 (BST)
From: Chris Jefferson <[EMAIL PROTECTED]>
Subject: Mersenne: Back to the maths of primes for a sec....
Sorry, just a quick question!
In various places, I have read that the generalized Riemann hypothesis is
true, then there is a very simple test for primeness, namely if n is an
a-SPRP for all integers a<2(log n)^2, then n is prime. From a computation
viewpoint, is this actually of any use, as it will show if numbers are
composite and if it is quick, then primes could be checked using it, then
double checked via another means, also giving the opportunity to disprove
a major hypothesis of maths...
- ------------------------------------
Chris Jefferson, Girton College, Cambridge, [EMAIL PROTECTED]
- ------------------------------------
I have a proof that x^n+y^n=z^n never has integer solutions for n>2.
However, it won't fit into my signature file....
- ------------------------------------
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 11:31:38 -0500
From: Stephen Whitis <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Screen saver killers?
At 02:37 PM 5/16/99 +0000, you wrote:
>I _know_ it's not as efficient in the absolute sense, but, if people would
>prefer to run a screensaver _and_ a version of Prime95, I wonder if it
>would be worthwhile updating the old Win 3.1 screensaver version. The point
>is, we want to make it attractive for people to participate.
I don't read the group regularly.
But I guarantee that if you try to tell me that I can't run a screen saver on
my machine, then I'll tell you to jump off a bridge. (Disclaimer : I probably
will be somewhat less polite than that.)
Telling me "Oh, you can run our program, *and* if you want to run a screen
saver
for some stupid reason, you can run out screensaver" isn't any better.
You'll lose more cpu cycles than you'll gain if you try to tell people "Do
it our way
or else".
I run Prime NT as a service. I run a screensaver, *just because I like
it*. IF you tell
me that I can't do that, *I* will dump PrimeNT in a heartbeat. There isn't
anything
in it for me. I do it to help out, because it sounded like a cool project.
Now you
guys think you are big brother, able to control what I can and can't do with my
machine. That plan sucks.
- ---
Stephen Whitis
Visit http://www.whitis.com for information
about Delphi, NT4.0 software, and more.
Support the anti-Spam amendment
Join at http://www.cauce.org/
I won't do business with spammers, but I will report them, ridicule
them, etc. Spam me at your own risk.
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 12:53:44 -0400
From: "Rick Pali" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: Screen saver killers?
> The Prime95 program we now use does do something
> we don't understand and does not do anything useful.
I've been using the NT service version since it was released and have *zero*
problems with it. It works as advertised...nothing more and nothing less.
Change a winning formula for no good reason and you're gonna piss people
off. Sure you could add yet another program interface, but I'd hate to see
development bog down because of so many interfaces. I'm not sure I know what
you mean by 'does not do anything useful.' I find the program greatly useful
for changing the options within the program. :-)
> I think a screensaver could be written in such a way
> that it also chunks up cycles that are spilled when
> the machine is used normally.
That would be fine, but it certainly doesn't offer any advantages over what
I have now. I've never been a big fan of screensavers and don't plan on
changing now. And unless there's definitely expressed interest in this
change, there really is no point in what amounts to a 'make-work' project.
> Furthermore the small image floating over the black
> screen should be a little attractive.
I'd much prefer to see my desktop and any open applications. Maybe that's
not so attractive, but it's actually *useful.* Why would I want to cover my
work for something that amounts to an ad for something that I already have
installed?
> If Prime95 is a screensaver, we all know what is does
> do (It prevents the screen from burning in
I've had exactly two monitors in the last eight years and neither one had
anything close to burn-in artefacts. Besides privacy, security, and novelty,
there's no need at all for screensavers these days on a machine that someone
actually uses, IMO.
> and at the same time it tries to win $50,000 from a
> bunch of math geeks)
I'm already doing that without a screensaver.
Rick.
- -----
[EMAIL PROTECTED]
http://www.alienshore.com/
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 09:56:19 -0700
From: Spike Jones <[EMAIL PROTECTED]>
Subject: Mersenne: ET phone George
Petri Holopainen wrote:
> ..."primitive" civilizations like ours. In that case, SETI would be
> pointless. But it would be pretty sad if we didn't even *try* to find
> them...
What if... SETI@home does manage to find ET. Then historians
will realize that the concept of distributed computing was pioneered
by Primenet and the GIMPS project. The history books will record
George Woltman as one of the pioneers that had an idea that indirectly
led to a discovery far more profound than the next Mersenne prime.
Way to go George! {8^D spike
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 10:45:06 -0700
From: Spike Jones <[EMAIL PROTECTED]>
Subject: Mersenne: Re: Mersenne dynamic optimism
Spike Jones wrote: ...What if... SETI@home does manage to find ET...
There is yet another way to look at this. From reading the GIMPS
posts on SETI, it is clear that many are seeing SETI as a competitor
for idle CPUs. Yes, it is that, in a sense. I see that SETI@home is
getting nearly as much CPU time as GIMPS, and it is only in its fourth
full day. (Demonstrates the power of publicity). However, if we view
SETI@home, not as a competing sibling but rather as a robust
offspring, then perhaps it is not so bitter. I still think GIMPS will
be the winner in the long run, for many new CPU idlers will sign
on to SETI, realize the power of distributed computing, and many
will perhaps land in GIMPS. Sure will will lose a few GIMPSers
to SETI@home, but perhaps we will gain many.
I see arguments for why there cannot be ETs, and other arguments
for why they *must* be there, but in fact no one knows. Is not this
the nature of scientific investigation, to find out? And GIMPS? As
we say in my business, one test is worth a thousand expert opinions.
Secondly, since I am on the topic of dynamic optimism, consider what
we are doing whenever we get the result "2^yakkityyak-1 is not prime."
So, we have found nothing, right? NO! We have discovered
*another* Mersenne composite! One that was unknown before.
This process cannot be shortcut; the only way to know for sure
if a Mersenne number is composite is to test it.
So, look on the bright side, my mathematical friends. We are mapping
the mathematical landscape every time we discover a new Mersenne
composite, even if they are as common as grains of sand on the
beach. Of course we want to find the diamonds, but to do so
requires sifting the sand. When we map this landscape, it is the
same for all time and all the universe. SETI@home is making a
map of sorts too. Let us wish them well and continue. spike
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 11:13:15 -0700
From: Paul Leyland <[EMAIL PROTECTED]>
Subject: RE: Mersenne: usefulness and 486's
In my opinion, 486's are definitely good for ECM. I'm running a bunch of
old DECstation 5000/25 at home, each of which is about as powerful as a
DX2-66. I'm using the ECMNET client/server code. One machine is the
server and all are clients. (My home net is not internet-connected so I
can't use the master server at 193.128.60.129). So far, they have found
about fifteen factors of Cullen, Woodall and N!+1 numbers. I can't see any
reason at all why your 486 boxes shouldn't be equally successful, whether
with George's ECM code or the ECMNET version.
Paul
- -----Original Message-----
From: David M. Moore [mailto:[EMAIL PROTECTED]]
Sent: 15 May 1999 07:05
To: Mersenne@Base. Com
Subject: RE: Mersenne: usefulness and 486's
On 14 May 99, at 22:31, Aaron Blosser wrote:
> You could also do something more conducive to the 486 environment...I
> don't know - is ECM good work for a 486?
I don't know. That was going to be my next question if it was decided
that 486's are wasting time :-) Although none of the other projects have
really grabbed my attention, I would put the 486's on to another project if
they were more suited to what ever that project does.
David
[EMAIL PROTECTED]
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 11:18:12 -0700
From: Paul Leyland <[EMAIL PROTECTED]>
Subject: RE: Mersenne: usefulness and 486's
George wrote:
> Prime95 (LL, ECM, and factoring), Proth, and I suspect SETI are mostly
> floating point. Conrad Curry's NFS program, ECM-NET, and the encryption
> cracking are probably mostly integer operations.
ECMNET is virtually entirely integer work. The GMP library which underlies
it is *very* heavily optimized integer code on the Intel architecture.
> Of course, you have to weigh that against which projects you find
> most worthwhile and fun.
Very true! I find ECMNET worthwhile and fun, but I'm biased.
Paul
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 20:06:09 -0500
From: [EMAIL PROTECTED]
Subject: Re: Mersenne: Screen saver killers?
Maybe it would just be more simple to use B.Beesley's idea of user
education and discretion. A half powered PC is better than none. Since
the computer actually belongs to the individual, then let them decide
what to do. Perhaps Prime95 could have a message box warning that comes
up when the screen saver is installed, telling the user how much of the
system resources are being spent on the screen saver. It could try to
convince the user to switch it off but if the user isn't interested, then
click "Ok" and nothing bad comes of it.
Besides, no one wants to get a flu from cold river water.
- -oliver
On Sun, 16 May 1999 11:31:38 -0500 Stephen Whitis <[EMAIL PROTECTED]>
writes:
>
>But I guarantee that if you try to tell me that I can't run a screen
>saver on
>my machine, then I'll tell you to jump off a bridge.
___________________________________________________________________
You don't need to buy Internet access to use free Internet e-mail.
Get completely free e-mail from Juno at http://www.juno.com/getjuno.html
or call Juno at (800) 654-JUNO [654-5866]
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 14:39:04 -0400
From: Peter Doherty <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Re: Mersenne dynamic optimism
Personally, since I'm more of a dreamer, than a mathematition (I can't even
spell it) SETI@home interests me more than GIMPS....but I'm going to stick
to GIMPs for now for a few reasons..
The Program....a constantly running idle priority program is nicer than a
screensaver IMHO...
I've already started with GIMPs...and I'm around rank 440, and proud of
that high ranking, and kind of like the friendly competition.
The simplicity of the goal...GIMPS searches for large primes...SETI
searches for radio signals which may or may not be from earth, which may or
may not be natural, which may or may not may make any sense....
However...on the other hand, if SETI finds something (by something, I mean
a clearly repeating radio transmission that can only be alien in origin),
we can know that there is other intelligent life out there. I know that
it's out there, but we haven't yet found it, but proof of my belief would
be nice... Although I don't see any immediate pratical purpose to mersenne
primes...but like I said before, my interest and knowledge in math isn't
very large.
- --Peter
At 10:45 05/16/1999 -0700, you wrote:
>Spike Jones wrote: ...What if... SETI@home does manage to find ET...
>
>There is yet another way to look at this. From reading the GIMPS
>posts on SETI, it is clear that many are seeing SETI as a competitor
>for idle CPUs. Yes, it is that, in a sense. I see that SETI@home is
>getting nearly as much CPU time as GIMPS, and it is only in its fourth
>full day. (Demonstrates the power of publicity). However, if we view
>SETI@home, not as a competing sibling but rather as a robust
>offspring, then perhaps it is not so bitter. I still think GIMPS will
>be the winner in the long run, for many new CPU idlers will sign
>on to SETI, realize the power of distributed computing, and many
>will perhaps land in GIMPS. Sure will will lose a few GIMPSers
>to SETI@home, but perhaps we will gain many.
>
>I see arguments for why there cannot be ETs, and other arguments
>for why they *must* be there, but in fact no one knows. Is not this
>the nature of scientific investigation, to find out? And GIMPS? As
>we say in my business, one test is worth a thousand expert opinions.
>
>Secondly, since I am on the topic of dynamic optimism, consider what
>we are doing whenever we get the result "2^yakkityyak-1 is not prime."
>So, we have found nothing, right? NO! We have discovered
>*another* Mersenne composite! One that was unknown before.
>This process cannot be shortcut; the only way to know for sure
>if a Mersenne number is composite is to test it.
>
>So, look on the bright side, my mathematical friends. We are mapping
>the mathematical landscape every time we discover a new Mersenne
>composite, even if they are as common as grains of sand on the
>beach. Of course we want to find the diamonds, but to do so
>requires sifting the sand. When we map this landscape, it is the
>same for all time and all the universe. SETI@home is making a
>map of sorts too. Let us wish them well and continue. spike
>
>________________________________________________________________
>Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
>
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 13:53:49 -0500
From: Ken Kriesel <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Screen saver killers?
At 09:10 PM 1999/05/14 +0200, "Steinar H. Gunderson"
<[EMAIL PROTECTED]> wrote:
>OK, second message in a row, I just thought it would be nice to separate
them.
>
>Has anybody got experience in turning off/disabling screensavers under Win95?
>We run Prime95 at 40 machines (most of them 486'es, though) at school, and
>screen savers are CPU hoggers (I suppose... at least everybody tells me so).
>Two solutions I could think of (both ideally incorporated under Win95):
>
>1. Have Prime95 reset the screen saver every 5 minutes or so (possibly when
> it's outputting) to `Blank Screen'. Would need fiddling with the Registry.
>2. Do a call to mouse_event() or keyboard_event() (possibly hitting Ctrl, or
> any not-so-important key, or moving the mouse one pixel), preventing the
> screen saver from being run at all.
Prime95 already has priority levels that are settable. The default is lower
than screensaver, but you don't have to run at default setting.
Have a look at the readme.txt for Prime95 V18.1 (or any version at least
since
V14.4 which had command line argument -P settable 0 thru 5, where
3 is equal to screensaver priority and 0 is near the level of the idle
loop).
Especially avoid the use of OpenGL screensavers; these are very cpu-intensive,
and run at higher priority than remote (network) file access, for network-
based backup, for example.
Putting both intermediate files and the prime95 executable file on a file
server is a nice way to ensure work is backed up, updating executables is
straightforward, and space used is efficient. It does have the disadvantages
of requiring someone logged in unless the security configuration is unusually
lax, and presents a single point of failure. (When the server reboots, all
40 instances of prime95 stop but may appear to still be running until a
mouse cursor crosses the prime95 icon.)
An alternative is to write a script which normally copies the pq* files from
local disk to server space, at whatever interval you prefer. They get backed
up with the server, so if ghost wipes the workstation clean, you can easily
restore an exponent that's 90% complete. As we move upward in exponents,
this will be of increasing interest. The script runs when there's a logged
in user.
In general, if you'd like to automate a task in Win95/NT, take a look at
Winbatch from Wilson Windoware; it's very handy for some things. Another
possibility is Active State perl.
Ken
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 14:57:21 -0400 (EDT)
From: lrwiman <[EMAIL PROTECTED]>
Subject: Mersenne: Repeating LL remainders
all,
A couple of weeks ago, there was discussion about repeating remainders in
the LL test. There was general agreement that this would be technically
unworkable due to the fact that the remainders are so large. Wouldn't it
be possible to store the last 1024 binary digits of the remainder (saving
1024 of these would be only 1MB, not a big deal.) Then we could check for
repeating remainders in the last 1024 iterations, without signifigantly
restraining performance. If it actually repeated, it could be confirmed in
double-checking quite easily.
As I don't know that much about searching algorithms, I don't know how much
this would hurt performance. Any thoughts? If this turns out to be workable,
is it still possible to implement in V19?
- -Lucas
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 19:58:27 +0200
From: "Henk Stokhorst." <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Screen saver killers?
Rick,
Sorry, I wrote the text from the point of view of someone in an office using his
computer on which someone else installed prime95. I don't need the version for
myself, but it's great to distribute among people who have no clue about primes.
YotN,
Henk Stokhorst.
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 16:37:51 -0400
From: "Rick Pali" <[EMAIL PROTECTED]>
Subject: RE: Mersenne: Screen saver killers?
From: Henk Stokhorst.
> Sorry, I wrote the text from the point of view of
> someone in an office using his computer on which
> someone else installed prime95. I don't need the
> version for myself, but it's great to distribute
> among people who have no clue about primes.
Don't be sorry! I think this just shows what a difficult thing it is to
implement. We both are pretty familiar with the program that exists, and
what we want it to do...and we still get lost in misunderstanding.
Rick.
- -----
[EMAIL PROTECTED]
http://www.alienshore.com/
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 17:03:12 -0400
From: "Pierre Abbat" <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Repeating LL remainders
On Sun, 16 May 1999, you wrote:
> all,
> A couple of weeks ago, there was discussion about repeating remainders in
> the LL test. There was general agreement that this would be technically
> unworkable due to the fact that the remainders are so large. Wouldn't it
> be possible to store the last 1024 binary digits of the remainder (saving
> 1024 of these would be only 1MB, not a big deal.) Then we could check for
> repeating remainders in the last 1024 iterations, without signifigantly
> restraining performance. If it actually repeated, it could be confirmed in
> double-checking quite easily.
Once a remainder repeats, does it stay in a loop? If so, you can keep the
remainder when the iteration number is a power of 2, and detect much longer
loops.
phma
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 19:29:04 -0400
From: Marc Getty <[EMAIL PROTECTED]>
Subject: Mersenne: screen savers are necessary
> > If Prime95 is a screensaver, we all know what is does
> > do (It prevents the screen from burning in
> I've had exactly two monitors in the last eight years and neither one had
> anything close to burn-in artefacts. Besides privacy, security, and novelty,
> there's no need at all for screensavers these days on a machine that someone
> actually uses, IMO.
Screen burn it is definitely a problem for some. I have 72 monitors in
one of my lab that have just a tad of the desktop wallpaper burned
into it in only ~12 weeks of 24/7 operation. Luckily I setup the blank
screen saver before it became a real problem. I myself have had two
monitors so far in my lifetime, a Nec XV15 from 1992-1994 and a Nec
MultiSync 5FG since 1994. Neither one have had even the slightest of
problems, with absolutely no screen burn in.
A screen saver if used properly can definitely extend the life of a
monitor. I prefer either a blank screen as a screen saver or even
better is to use the power save features of a "green" monitor and turn
it off completely.
On the topic of a screen saver version of Prime95 I think this may be
a good idea, but let's not waste too much time on developing one.
Marc Getty - [EMAIL PROTECTED] - [EMAIL PROTECTED] - ICQ: 12916278
http://www.getty.net http://www.vwthing.org Work: 215-204-3291
http://etc.temple.edu/ Home/Cell: 215-962-5603
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 16:57:08 -0700
From: Kevin Sexton <[EMAIL PROTECTED]>
Subject: Mersenne: cacheable memory
Anybody know which chipsets have the memory caching
problems, or who can point me in the direction to find this
information?
I am interested in finding out whether to add memory to a
VXPro board, currently has 32MB, AMD K-6 200.
I know this is not strictly on topic, but it might increase
performance of prime95, especially if another app hogs the
memory I have now.
Kevin Sexton
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
Date: Sun, 16 May 1999 17:21:55 -0700
From: Kevin Sexton <[EMAIL PROTECTED]>
Subject: Re: Mersenne: Screen saver killers?
I see posts in reply to this that seem upset at disabling screen savers, I
think the original idea was to give admins. (in an environment like a school
lab) the power to automatically remove screen savers activated by
users(students).
The simple way might be to set up removal of the screen saver at computer
startup.
I don't think anyone wanted users to be forced not to have a screen saver at
their computer in an office setting, ect.
________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
------------------------------
End of Mersenne Digest V1 #559
******************************
