Message boards : Number crunching : Not HAPPY people.
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · Next

AuthorMessage
Jerry Camden

Send message
Joined: 13 Jul 05
Posts: 4
Credit: 19,555
RAC: 0
Message 13165 - Posted: 29 Mar 2006, 0:57:49 UTC

And I think LHC is such a worthwhile project, that I don't mind there being a pause once in a while. My hosts always have some other project they can run. But they also will run LHC when WU are ready. And I think that is due to the state of the Boinc! Still not perfect, but pretty smooth.
ID: 13165 · Report as offensive     Reply Quote
uioped1

Send message
Joined: 26 Aug 05
Posts: 18
Credit: 37,965
RAC: 0
Message 13168 - Posted: 29 Mar 2006, 6:16:48 UTC - in response to Message 13165.  

Now, what really bothers me about this project are the days with 100, or even 50 workunits still in progress. I don't know if the scientists actually have to wait for those (I doubt it) but it irks me just the possibility...
ID: 13168 · Report as offensive     Reply Quote
m.mitch

Send message
Joined: 4 Sep 05
Posts: 112
Credit: 1,864,470
RAC: 0
Message 13178 - Posted: 29 Mar 2006, 17:09:00 UTC - in response to Message 13151.  

With Climate, you are running an "adapted" 64bit climate model, with millions of Fortran code lines, that has originally been coded to run on supercomputers. Since the model started being coded in the 70s (I think I read it somewhere) I guess we now all have supercomputers at home ;-)


Ain't that the truth!!! I remember my first DEC mainframe (not midrange). It had the brand spanking new eight bit words, WOW. AND, we had FIVE, yes FIVE 100 kilobyte platters!! It was a powerful monster.
I'm not sure, but I think we had a whopping 1024 kilobytes of RAM (more probably 16 or 32 ;-). HUGE tape backup, three boxes. And this magic thing that hooked up to a telephone at 300 baud, so if our system was down, the failsafe system could take over all the work over the phone. True! We called it "Distributed Computing". No just kidding, but that's what it was in effect. It could also take up extra load. We ran Fortran and/or BASIC (No COBOL). We also had chrome toggle switches with LEDs above them. To get it started, we'd get out a book and set the toggle switches just as described. Push a button, then do the next setting. The buttons and LEDs were arranged in groups of eight with a slight space between the first four and the last four. I think that was what was called a "boot strap".

It was a huge improvement on the punch cards :-)





Click here to join the #1 Aussie Alliance on LHC.
ID: 13178 · Report as offensive     Reply Quote
m.mitch

Send message
Joined: 4 Sep 05
Posts: 112
Credit: 1,864,470
RAC: 0
Message 13179 - Posted: 29 Mar 2006, 17:16:41 UTC - in response to Message 13152.  
Last modified: 29 Mar 2006, 18:15:15 UTC

1. They have the best communication with users of any BOINC project
... [all else snipped]...


Nah, Einstein's the best, then µFluids, then Rosetta. It's still a bit subjective. We don't need much communication with project staff at LHC. We already know were not going to get any from SETI project staff, so we don't ask and we're not disappointed.




Click here to join the #1 Aussie Alliance on LHC.
ID: 13179 · Report as offensive     Reply Quote
lpoorman

Send message
Joined: 24 Jul 05
Posts: 18
Credit: 4,236
RAC: 0
Message 13180 - Posted: 29 Mar 2006, 17:17:24 UTC - in response to Message 13178.  


Ain't that the truth!!! I remember my first DEC mainframe (not midrange). It had the brand spanking new eight bit words, WOW. AND, we had FIVE, yes FIVE 100 kilobyte platters!! It was a powerful monster.

[/quote]
Was that a PDP-11? Did you put the first boot strap in with the switches and the
second one with a paper tape?
ID: 13180 · Report as offensive     Reply Quote
m.mitch

Send message
Joined: 4 Sep 05
Posts: 112
Credit: 1,864,470
RAC: 0
Message 13181 - Posted: 29 Mar 2006, 17:55:43 UTC - in response to Message 13163.  

Meanwhile more LHC work showed up! (28 Mar 2006 20:47:09 UTC)


Bugger!! Mine didn't show up until:
30/03/2006 2:42:51 AM|LHC@home|No schedulers responded

Then I got:
30/03/2006 3:14:14 AM|LHC@home|Started download of
sixtrack_4.67_windows_intelx86.exe
30/03/2006 3:14:14 AM|LHC@home|Started download of bottomOverlay_1.02_.tga
30/03/2006 3:14:19 AM|LHC@home|Finished download of bottomOverlay_1.02_.tga

Wonder what I did to delay it about 30 hours :-0




Click here to join the #1 Aussie Alliance on LHC.
ID: 13181 · Report as offensive     Reply Quote
m.mitch

Send message
Joined: 4 Sep 05
Posts: 112
Credit: 1,864,470
RAC: 0
Message 13182 - Posted: 29 Mar 2006, 18:12:40 UTC - in response to Message 13180.  
Last modified: 29 Mar 2006, 18:18:05 UTC

Ain't that the truth!!! I remember my first DEC mainframe (not midrange). It had the brand spanking new eight bit words, WOW. AND, we had FIVE, yes FIVE 100 kilobyte platters!! It was a powerful monster.


Was that a PDP-11? Did you put the first boot strap in with the switches and the second one with a paper tape?


Oh no. The second one was putting all the toggle switches in to there second positions after all the LEDs had gone from yellow to green. That was a three person exercise. This was a big mother, it took up two rooms. It predated the beautiful PDP11/780 that I worked with latter (that was back in my cobbling days 8-).

The PDP11 had 16 bit words didn't it, or were they Long Words? Then I went on to a MicroVAX Cluster :-) Hold your breath, I did Alpha and Beta testing on both the VAX Alpha and OpenVMS 8-D.

Those were the days, I really liked working with computers then. I did bench marking on VAX's too!!

Did you ever see a VAX PDP11/730? They were so cute and tiny.




Click here to join the #1 Aussie Alliance on LHC.
ID: 13182 · Report as offensive     Reply Quote
River~~

Send message
Joined: 13 Jul 05
Posts: 456
Credit: 75,142
RAC: 0
Message 13190 - Posted: 29 Mar 2006, 20:00:41 UTC - in response to Message 13181.  

Meanwhile more LHC work showed up! (28 Mar 2006 20:47:09 UTC)


Bugger!! Mine didn't show up until [...]30/03/2006
[...]
Wonder what I did to delay it about 30 hours :-0



Nothing you did, you were just unlucky. They took their time adding the new work - and for the first 24hrs or so the queue kept emptying. 12 hours after I got my work the front page was down to double figures of work available for example. I have around a dozen boxes crunching LHC and I had a spread in the start times.

Then at a 3hour standoff you only had to miss a couple of attempts due to congestion (the "no schedulers responded" message) takes you to the 30 hours.

River~~
ID: 13190 · Report as offensive     Reply Quote
m.mitch

Send message
Joined: 4 Sep 05
Posts: 112
Credit: 1,864,470
RAC: 0
Message 13192 - Posted: 29 Mar 2006, 20:08:14 UTC - in response to Message 13190.  

Nothing you did, you were just unlucky. They took their time adding the new work - and for the first 24hrs or so the queue kept emptying. 12 hours after I got my work the front page was down to double figures of work available for example. I have around a dozen boxes crunching LHC and I had a spread in the start times.

Then at a 3hour standoff you only had to miss a couple of attempts due to congestion (the "no schedulers responded" message) takes you to the 30 hours.

River~~


That's a relief, thanks River. It also explains why LHC was giving me nothing found messages for sometime. Usually I pick the WU's up straightaway. Oh well, I have a heap now. So on with the crunching ;-)




Click here to join the #1 Aussie Alliance on LHC.
ID: 13192 · Report as offensive     Reply Quote
Profile Yeti
Volunteer moderator
Avatar

Send message
Joined: 2 Sep 04
Posts: 453
Credit: 193,369,412
RAC: 10,065
Message 13193 - Posted: 29 Mar 2006, 20:45:40 UTC - in response to Message 13178.  
Last modified: 29 Mar 2006, 20:46:40 UTC

We also had chrome toggle switches with LEDs above them. To get it started, we'd get out a book and set the toggle switches just as described. Push a button, then do the next setting. The buttons and LEDs were arranged in groups of eight with a slight space between the first four and the last four. I think that was what was called a "boot strap".

It was a huge improvement on the punch cards :-)

Yeah, my first computer-contacts have been with a pdp-11/10, and I remember these boot-straps ...

We had terminals and on each terminal, we could use 2K Memory. What a big progress, when this was blown up to 8K !

There I wrote a program (in BASIC !), that would today be named "Virus" or malicious-code: It was hidden in the machine, slept for a given time and then started to first throw messages on sessions of other users and later on, sent commands to their sessions. They punished me with keeping me off the machine for 4 weeks :-((

But it was a great time with big platters ;-)



Supporting BOINC, a great concept !
ID: 13193 · Report as offensive     Reply Quote
m.mitch

Send message
Joined: 4 Sep 05
Posts: 112
Credit: 1,864,470
RAC: 0
Message 13195 - Posted: 30 Mar 2006, 0:30:58 UTC - in response to Message 13193.  

Yeah, my first computer-contacts have been with a pdp-11/10, and I remember these boot-straps ...

We had terminals and on each terminal, we could use 2K Memory. What a big progress, when this was blown up to 8K !

There I wrote a program (in BASIC !), that would today be named "Virus" or malicious-code: It was hidden in the machine, slept for a given time and then started to first throw messages on sessions of other users and later on, sent commands to their sessions. They punished me with keeping me off the machine for 4 weeks :-((

But it was a great time with big platters ;-)


Do you remember the first time someone could afford a second lot of 16K of RAM! Hell, the entire EDP department shut down, while we all watched through the glass in to the forbidden computer room while they put it in. Were only our Sysmgr and his two sysops could go 8-) and of course the DEC people with 16K of RAM. It might have ben 64K by then, it's all a bit long ago 8-)

You actually jogged my memory then. I first started on punch cards and we programmed in some deviant of Fortran ;-)




Click here to join the #1 Aussie Alliance on LHC.
ID: 13195 · Report as offensive     Reply Quote
Profile Yeti
Volunteer moderator
Avatar

Send message
Joined: 2 Sep 04
Posts: 453
Credit: 193,369,412
RAC: 10,065
Message 13200 - Posted: 30 Mar 2006, 17:20:36 UTC

Our box must have been a little bit later than yours; we had a card-reader, that could read cards, that had not been punshed, but signed with a pen (pencil ?)



Supporting BOINC, a great concept !
ID: 13200 · Report as offensive     Reply Quote
River~~

Send message
Joined: 13 Jul 05
Posts: 456
Credit: 75,142
RAC: 0
Message 13202 - Posted: 30 Mar 2006, 20:14:39 UTC - in response to Message 13200.  

Our box must have been a little bit later than yours; we had a card-reader, that could read cards, that had not been punshed, but signed with a pen (pencil ?)


We had a Fortran (Fortran I, only it was just Fortran then) compiler that used paper tape for its intermediate storage. It was designed to do the second pass of the code in reverse order, so you had to put the intermediate tape in backwards. It was actually very convenient, as the end that was still hanging out of the punch went straight into the reader. Then the compiler produced a second tape that was the object code - because it was being punched on the backward pass, you put that in rear end first as well.

It got so it felt wrong to be putting a tape into the reader arrow head first...

River~~

ID: 13202 · Report as offensive     Reply Quote
Gaspode the UnDressed

Send message
Joined: 1 Sep 04
Posts: 506
Credit: 118,619
RAC: 0
Message 13204 - Posted: 30 Mar 2006, 21:06:48 UTC

Sorry - aren't you all a bit off topic here?

Or are you saying that you aren't (weren't) happy with your PDP10, Fortran, punch tape?

Nice as all this wallowing in nostalgia may be for you, those of us who majored on slide rules and log tables find it all a bit too hi-tech!

;-)



Gaspode the UnDressed
http://www.littlevale.co.uk
ID: 13204 · Report as offensive     Reply Quote
R/B
Avatar

Send message
Joined: 30 Jan 06
Posts: 10
Credit: 1,947
RAC: 0
Message 13205 - Posted: 30 Mar 2006, 21:39:02 UTC

Punch cards? I'm currently running LHC on punch cards. Perhaps noone told me I couldn't find a way to get that done?

Seriously though, In the last day or two Rosetta has made great progress toward solving their 1% bug with the help of Rom Watkins. For those here that I saw complain about it they might need to go and read their Science forums.....
Founder of BOINC group Objectivists: Rational people crunching data for science.

ID: 13205 · Report as offensive     Reply Quote
m.mitch

Send message
Joined: 4 Sep 05
Posts: 112
Credit: 1,864,470
RAC: 0
Message 13208 - Posted: 31 Mar 2006, 5:54:13 UTC - in response to Message 13204.  

Sorry - aren't you all a bit off topic here?

Or are you saying that you aren't (weren't) happy with your PDP10, Fortran, punch tape?

Nice as all this wallowing in nostalgia may be for you, those of us who majored on slide rules and log tables find it all a bit too hi-tech!

;-)




"slide rules and log tables". Hey, you must be my vintage. That's what we did at school. As soon as I got out, I was shown these humungus things that looked an awful lot like Star Treck hardware that did all that for me. I'd been robbed at school, in real life we could spend all our time fixing the problem not worry how accurate our slide rule was or how we should interprolate a few lines from a log book. All though, in the early days I was know to check the output against the "trusty" slide rules and log tables 8-)

Oh, the other thing computers introduced in to the work force was "waiting". One couldn't leave the area until the job run finished, it might fail!



Click here to join the #1 Aussie Alliance on LHC.
ID: 13208 · Report as offensive     Reply Quote
Profile The Gas Giant

Send message
Joined: 2 Sep 04
Posts: 309
Credit: 715,258
RAC: 0
Message 13221 - Posted: 31 Mar 2006, 13:52:40 UTC - in response to Message 13205.  

Punch cards? I'm currently running LHC on punch cards. Perhaps noone told me I couldn't find a way to get that done?

Seriously though, In the last day or two Rosetta has made great progress toward solving their 1% bug with the help of Rom Watkins. For those here that I saw complain about it they might need to go and read their Science forums.....

Could have sworn that was Rom Walton.....
ID: 13221 · Report as offensive     Reply Quote
Profile UBT - Halifax--lad
Avatar

Send message
Joined: 30 Sep 04
Posts: 23
Credit: 1,372
RAC: 0
Message 13228 - Posted: 2 Apr 2006, 8:15:37 UTC - in response to Message 13221.  

Could have sworn that was Rom Walton.....


Unless he has changed his name to hide from us all, ROM has done some good work over @Rosetta helping to get rid of loads of bugs in Windows, still trying to track the 1% bug but it seems to have run away to hide at the moment
Join us in Chat (see the forum) Click the Sig


Join UBT
ID: 13228 · Report as offensive     Reply Quote
R/B
Avatar

Send message
Joined: 30 Jan 06
Posts: 10
Credit: 1,947
RAC: 0
Message 13230 - Posted: 3 Apr 2006, 13:18:39 UTC - in response to Message 13228.  

Could have sworn that was Rom Walton.....


Unless he has changed his name to hide from us all, ROM has done some good work over @Rosetta helping to get rid of loads of bugs in Windows, still trying to track the 1% bug but it seems to have run away to hide at the moment

Yes, Rom Walton. I apologize for quoting his last name wrong.
Founder of BOINC group Objectivists: Rational people crunching data for science.

ID: 13230 · Report as offensive     Reply Quote
Nuadormrac

Send message
Joined: 26 Sep 05
Posts: 85
Credit: 421,130
RAC: 0
Message 13258 - Posted: 7 Apr 2006, 15:11:15 UTC - in response to Message 13072.  

It is my understanding that they are evaluating the results of the last lot of work we did so they can develop the next study. I tend to suspend LHC@home when there is no work and just check this site regularly and reactivate once work appears rather than have my system trying to call for new work all the time. My CPU is crunching either Predictor or Fighting Aids as I write. I am really pleased to think that I have made a contribution, no matter how small, to the development of the LHC.


This is what I've done also (well perhaps not the same project mix as you, but with the suspend). I don't see the big deal here. It's not like he can't suspend LHC during long outages and check the site. When it looks like work might be close to comming (aka when the number of WUs out there that haven't been returned come close to 0), one can just check the site a bit more, and when work is announced, resume it... This was one of the features allowed in the newer CCs, and can work well. It isn't like the resources have to be taken during these outages.

And peterthomas, may I suggest you never sign up for an alpha or beta project. You'll likely pop a blood vessel if periods of no work bothers you here. I'll have you know that work here is more available then at some projects like RALPH, or pirates for that matter... But that's also the nature of testing, and tbh things really need to be that way, as we're not always needed... It is also why most of us have other projects too ;) Honestly, sign up at CPDN or some other such project if you're that worried about running out of work :D
ID: 13258 · Report as offensive     Reply Quote
Previous · 1 · 2 · 3 · 4 · 5 · Next

Message boards : Number crunching : Not HAPPY people.


©2024 CERN