IRC logs of #boinc for Tuesday, 2009-03-10

00:03 <jasong>

00:03 <Romulus> <> (at

00:04 <jasong> After the whole nuclear sub thing I see this a few weeks later.   Makes me wonder if they have a policy that makes it okay if no one is at the wheel sometimes

00:04 <jasong> errrr, makes them think it's okay

00:10 *** infinisoft has quit IRC

00:11 <Soul_keeper>   ION struggles to play DVD's apparently ...

00:11 <Romulus> <> (at

00:32 *** mchelen has joined #boinc

00:37 <Soul_keeper> so how are you people's carbon footprint doing today ?

00:38 *** mchelen has quit IRC

00:38 *** mchelen has joined #boinc

00:54 * efc stomps the world with his carbon footprint

00:55 <Romulus> New news from boinc: BOINC news March 9, 2009

01:04 <Soul_keeper>

01:04 <Romulus> Title: YouTube - Car Skydiving (at

01:12 <Soul_keeper> gts 250's just hit newegg, nice

01:13 <MTughan> 1GB or 512MB?

01:14 <Soul_keeper> both

01:14 <Soul_keeper> like 8 models

01:14 <Soul_keeper> most don't have pictures yet, i checked last night and they weren't there

01:14 <MTughan> Because IIRC, the 512MB doesn't necessarily have to be anything different from the 9800GTX+, but the 1GB version does bring a few changes.

01:15 <Soul_keeper> yeah some could just be rebranded, but the board size is supposed to be 1.5" shorter if they follow the new design

01:15 <MTughan> It's supposed to be, but only the 1GB version is mandated to be. The 512MB version isn't required to be different.

01:16 <MTughan> Also, only uses one PCIe power connector.

01:16 <Soul_keeper> we won't know till we get some reviews and what not tho

01:17 <MTughan>

01:17 <Romulus> <> (at

01:17 <Soul_keeper>

01:17 <Romulus> <> (at

01:17 <Soul_keeper>

01:17 <Romulus> <> (at

01:17 <Soul_keeper> the XFX 512 one is clearly shorter

01:18 <Soul_keeper> but you're right i read that review too

01:18 <MTughan> I can't really see it... Look at the length in relation to the PCIe connector on both.

01:19 <MTughan> Nope, you're right. I can tell from the head-on view.

01:19 <MTughan> Interesting cooler design on the new one.

01:20 <Soul_keeper> these will use the same amount of power as my ati 3870

01:21 <MTughan> Still not as powerful as a 4870, but it will use less power than that if it uses the same power as a 3870.

01:21 <MTughan> I see 8 GTS cards listed on Newegg. Presumably, they're all GTS 250's right now.

01:21 <MTughan> (I keep typing GTX 250... :P)

01:22 <MTughan>

01:22 <Soul_keeper> yeah i might be tempted to pick one up

01:22 <Romulus> <> (at

01:23 <Soul_keeper> i'll wait till they sale out the first wave tho, to see how that goes

01:23 <MTughan> For quality control, essentially?

01:23 <Soul_keeper> yeah i don't wanna deal with a bios bug, or something strange, also the gts 250  isn't officially listed in the nvidia linux support

01:24 <MTughan> Gigabyte sure didn't waste any time getting a different cooler on it...

01:24 <Romulus> <> (at

01:24 <Soul_keeper> although i don't think it should matter since it's just a 9800

01:25 <Soul_keeper> yeah that's a nice card, i prefer the dust buster tho :)

01:25 <MTughan> The only thing that could cause problems would either be the PCB shrink or the loss of one power connector.

01:25 <MTughan> :P Just how loud is the default cooler on a given Nvidia card?

01:25 <Soul_keeper> that gigabyte card is almost as small as the old  geforce 2 gts

01:26 <Soul_keeper> i can't heard video card fans, never one one i could

01:26 <Soul_keeper> hear*

01:26 <MTughan> You know that's a bad sign. :P

01:26 <Soul_keeper> heh yeah

01:30 <Soul_keeper> i'd bet tempted to get a 4870 if i could boinc on it ...

01:30 <MTughan> I'll keep an eye on that license discussion on MW, let you know what comes of it.

01:31 <MTughan> BTW, did you see Cluster Physik's latest post?

01:31 <Soul_keeper> yeah i'm not worried, it would likely take me forever to do anything anyways

01:31 <Soul_keeper> they can do it faster than me i bet

01:31 <Soul_keeper> if it's working in windows, then making it work in linux isn't much of a step

01:33 <Soul_keeper> understanding the whole concept/math behind their project would take most the time.

01:33 <MTughan> Yeah.

01:33 <Soul_keeper>   lol that card was fast when i had one

01:34 <Romulus> <> (at

01:34 <MTughan> Heh. How long has it been since we saw fans that small?

01:34 <MTughan> And what is that, an AGP connector?

01:34 <Soul_keeper> yeah

01:34 <Soul_keeper> think it was agp 2x

01:34 <Soul_keeper> can't remember now

01:35 <MTughan> Nice.

01:35 <Soul_keeper> that card traded blows with the voodoo5

01:35 <MTughan> I think my Radeon 9600 was AGP 8x.

01:35 <Soul_keeper> i owned both

01:35 <MTughan> I've never really seen a Voodoo card.

01:36 <Soul_keeper> they were pretty good for games like UT that supported glide

01:36 <Soul_keeper> the one i had was made in mexico  heh

01:37 <MTughan> Wow, that's old now...

01:39 <Soul_keeper>

01:39 <Romulus> <> (at

01:40 <Soul_keeper> only one GPU was supported in linux, that's why i ended up getting the gf2 at the time

01:40 <MTughan> That's a dual 7600GT board?

01:40 <Soul_keeper> that's the voodoo5

01:41 <MTughan> What two boards are there? A 7600GT and a Voodoo5?

01:41 <MTughan> 7800GT*

01:41 <Soul_keeper> yeah looks like it

01:41 <MTughan>

01:41 <Romulus> <> (at

01:42 <Soul_keeper> lol

01:42 <Soul_keeper> yeah people wanted another die shrink i guess

01:43 <MTughan> Quite cynical, but it seems to be all true.

01:43 <MTughan> Nvidia leads in all games against an HD4850 and even a 4870 in one game, Mirror's Edge.

01:43 <MTughan> He suspects PhysX was used, so tries to find the configuration, for which he's pointed to the appendix.

01:43 <MTughan> No appendix.

01:44 <Soul_keeper> it takes the lead when you turn on AA and AF

01:44 <Soul_keeper> against the 4850 atleast

01:44 <MTughan> That may have been true with the 38xx series, but the 48xx cards take a lot less of a hit with AA and AF.

01:44 <MTughan> Yeah, it's possibly it competed with the 4850. But not the 4870.

01:44 <Soul_keeper> just what i noticed on the tomshardware review

01:46 <MTughan> rofl... "Nvidia has a miniscule share of [the integrated graphics notebook] market, and you can usually tell those notebooks by the vertical lines on the screen shortly before they turn black."

01:47 <Soul_keeper> heh

01:47 <MTughan> Presumably referring to the problems they've had with solder material in recent months.

01:48 <Soul_keeper> ohhh

01:50 <Soul_keeper> think i'll go to sleep now

02:12 <MTughan> I'm going to do the same. Past 2AM here now.

03:21 *** FreeLarry58 has joined #Boinc

03:37 *** FreeLarry has quit IRC

03:40 *** Eythan has joined #boinc

03:51 *** desti_T2 is now known as desti

04:08 *** efc has quit IRC

04:16 <FreeLarry58> evening all

04:19 <CoderForLife> morning

04:20 <CoderForLife> &wx 45140

04:20 <Romulus> CoderForLife: Temperature: 48.5F / 9.2C | Humidity: 66% | Pressure: 29.97in / 1014.8hPa (Falling) | Conditions: Clear | Wind Direction: ESE | Wind Speed: 0.0mph / 0.0km/h ; Today - Mostly cloudy with a chance of showers and thunderstorms. Highs in the lower 70s. Southeast winds 10 to 15 mph...becoming southwest 15 to 20 mph this afternoon. Chance of rain 50 percent.; Tonight - Showers and (1 more message)

04:20 <CoderForLife> &more

04:20 <Romulus> CoderForLife: thunderstorms. Breezy with lows in the lower 40s. Southwest winds 15 to 25 mph with gusts up to 35 mph. Chance of rain 100 percent.; Wednesday - Partly sunny. A chance of showers and thunderstorms in the morning. Highs in the mid 40s. West winds 15 to 20 mph with gusts up to 30 mph. Chance of rain 40 percent.;

04:33 <CoderForLife> brb  breakfast

04:40 <CoderForLife> b

04:58 *** mydog2 has quit IRC

05:27 <CoderForLife> time to get ready for w@#% - laterz

06:18 <XioNYC> &wx LGA

06:18 <Romulus> XioNYC: Temperature: 37F / 3C | Humidity: 76% | Pressure: 30.33in / 1027hPa (Rising) | Conditions: Partly Cloudy | Wind Direction: NE | Wind Speed: 4mph / 6km/h ; Today - Mostly cloudy. A chance of rain this afternoon. Highs in the mid 40s. East winds 5 to 10 mph...becoming southeast this afternoon. Chance of rain 30 percent.; Tonight - Rain likely. Lows in the upper 30s. Southeast winds 10 to 15 (1 more message)

06:18 <XioNYC> &more

06:18 <Romulus> XioNYC: mph. Chance of rain 60 percent.; Wednesday - Rain likely. Highs in the mid 50s. Southwest winds 10 to 15 mph. Chance of rain 70 percent.;

06:19 <XioNYC> Blecch.

06:24 <CoderForLife> LHC WUs seem to be available

06:26 <XioNYC> No PPC app, though... >_<

06:30 <hawmps> 'mornin

06:36 <XioNYC> In the words of a great BBC series, "Goot moaning".

06:37 <hawmps> goat moanin

07:54 <wdsmia> &wx 50319

07:54 <Romulus> wdsmia: Temperature: 37.1F / 2.8C | Humidity: 97% | Pressure: 29.70in / 1005.6hPa (Falling) | Conditions: Rain | Wind Direction: NNE | Wind Speed: 0.0mph / 0.0km/h ; Today - Mostly cloudy. A chance of light rain in the morning...then a slight chance of light rain in the afternoon. A chance of light snow and light sleet late in the afternoon. High in the upper 30s. North wind 5 to 10 mph shifting (1 more message)

07:54 <wdsmia> &more

07:54 <Romulus> wdsmia: to the northwest 10 to 20 mph with gusts to around 30 mph in the afternoon. Chance of precipitation 30 percent.; Tonight - Partly cloudy. A 30 percent chance of light snow and light sleet early in the evening. Windy. Much colder. Low around 10. Northwest wind 20 to 30 mph with gusts to around 40 mph.; Wednesday - Sunny...breezy...colder. High in the lower 20s. Northwest wind 15 to 25 mph.;

08:27 *** wdsmia_w has joined #Boinc

08:32 *** mchelen has quit IRC

08:49 *** wdsmia_w has quit IRC

09:21 *** wdsmia_w has joined #Boinc

09:29 *** factor has quit IRC

09:35 *** factor has joined #boinc

09:48 *** PovAdct_w has joined #boinc

09:50 *** fwerp2 has joined #boinc

09:50 *** Fwerp has quit IRC

10:12 <zombie67> Maybe I need to remove .Net 3.5....

10:12 <Romulus> <> (at

10:14 <zombie67> I am going to remove all .NET, and then install only 2.0 SP2

10:14 *** fwerp2 has quit IRC

10:15 <zombie67> whoops.  Here is the correct URL:

10:15 <Romulus> <> (at

10:20 *** Fwerp has joined #boinc

10:26 *** Aeternus has joined #boinc

10:31 <PovAdct_w> MW uses .NET?!

10:36 <zombie67> CCC does, I think

11:11 *** ELGono has quit IRC

11:12 *** Fwerp has quit IRC

11:19 *** zombie_67 has joined #boinc

11:28 *** zombie67 has quit IRC

11:29 *** zombie__67 has joined #boinc

11:30 *** zombie_67 has quit IRC

11:37 *** zombie__67 has quit IRC

11:37 *** zombie67 has joined #boinc

11:49 *** ELGono has joined #boinc

11:54 <MTughan> PovAdct_w: Yes, CCC and the ATI drivers need .NET if I understand it correctly.

11:55 *** PovAdct_w has quit IRC

12:33 *** DerMeister has joined #boinc

12:37 <MTughan> Hynix has made a deal with Rambus over SDRAM royalties.

12:37 <MTughan>

12:37 <Romulus> Title: Hynix agrees to pay Rambus royalties on RAM sales - The Tech Report (at

12:40 <Soul_keeper> lol MTughan that milkyway dev has some fanboys

12:41 <MTughan> Yeah, well look how much faster it crunches than the CPU app.

12:41 <MTughan> You don't think some people are going to love it?

12:41 <Soul_keeper> yeah give the masses points and they love you

12:42 <MTughan> :P Look at my thread about OS X apps. Even though I'm only taking about 10-15 minutes off, plenty of people wanting them.

12:43 <MTughan> BTW, I do actually have recompiled apps. Just haven't gotten around to getting them all and testing them, because I've been doing yoyo's applications.

12:43 <CoderForLife> &seen AFLAC

12:43 <Romulus> CoderForLife: AFLAC was last seen in #boinc 30 weeks, 5 days, 14 hours, 37 minutes, and 48 seconds ago: * AFLAC ZZZZZZZZZZZzzzzzzzzzzzzz

12:44 <Soul_keeper> can't find it

12:48 <MTughan>

12:48 <Romulus> <> (at

12:51 <Soul_keeper> crunch3r has the ppc/mac compiles on his server btw

12:52 <MTughan> For MW?

12:52 <Soul_keeper> yes

12:52 <Soul_keeper> hold on

12:52 <MTughan> I saw just TSP and SETI a couple days ago.

12:52 <Soul_keeper>

12:52 <Romulus> Title: Index of /~deselt (at

12:53 <Soul_keeper> i think that's crunch3r's server

12:53 <MTughan> No, that's actually the server of the University where MW is...

12:54 <MTughan> deselt sounds like Travis Desell's username.

12:54 <Soul_keeper> then that's probably who it is

12:54 <MTughan> If you see Travis on the forums, that's who it is. He and Dave <something> are the admins.

12:55 <Soul_keeper> k

12:55 <MTughan> Dave Przybylo

12:55 <Soul_keeper> i don't post there, only post i made was that GPL one

12:55 <MTughan> I can never remember his last name.

12:55 <MTughan> Yep, but if you're ever looking for info and come across Travis, that's who it is.

12:55 <MTughan> Anyway, checking out the apps now...

12:56 <Soul_keeper> download the xcode zip

12:56 <Soul_keeper> it has the powerpc and osx binary in it

12:56 <MTughan> Oh, there... K.

12:58 <MTughan> lol... The PPC applications in there is a combination PPC and i386.

12:58 <MTughan> No wonder it's so much bigger, 272KB compared to 952KB.

13:02 <Soul_keeper> what did you want ?  just ppc ?

13:02 <MTughan> Yes.

13:03 <MTughan> If it's labelled PPC, just do PPC. I've corrected the Xcode file locally anyway.

13:03 <Soul_keeper> i'd grab the latest GPL code and compile one myself

13:04 <MTughan> GPU or CPU code?

13:04 <MTughan> And you or me?

13:05 <Soul_keeper> for your ppc

13:05 <Soul_keeper> he didn't do it with that Makefile so you'll have to figure it out on your own tho

13:05 <MTughan> I can strip the i386 code out if need be.

13:06 <MTughan> And like I said, I've already corrected the Xcode project locally, just trying to get it to compile now.

13:07 <Soul_keeper> that's the "no license one"

13:08 <Soul_keeper> i compiled the latest GPL version yesterday

13:08 <Soul_keeper> it has an entry for PPC in the makefile

13:08 <MTughan> Yes it does.

13:10 <MTughan> If the Xcode project one doesn't have a license, why is it still being used?

13:15 <zombie67> uninstalling .NET 3.0 and 3.5 solved nothing.

13:15 <Soul_keeper> i dunno

13:16 <MTughan> zombie67: Try installing 3.5 and 2.0 separately, even if 3.5 shows 2.0 being installed.

13:16 <zombie67> huh?

13:16 <MTughan> Install .NET 3.5 and 3.0 again, and then download and install .NET 2.0.

13:17 <zombie67> I tried to look for the 2.0 d/l, but couldn't tell which of the MANY was the correct one.  I use XP64

13:18 <Soul_keeper>  MTughan here is your buddy  heh

13:18 <MTughan> That's not me.

13:18 <Soul_keeper> nah that's deslter or whatever

13:18 <MTughan> zombie67:

13:18 <Romulus> <> (at

13:19 <zombie67> Thanks!

13:28 *** siofwolves has joined #boinc

13:37 <MTughan> WTF?

13:37 <Romulus> <> (at

13:45 <FreeLarry58> ultra thin parking garage MTughan

13:46 <FreeLarry58> or for the very lazy to ride ferris wheel

13:46 <MTughan> "must be an American thing ... who else wouldn't want to get out of their car when they go places to sight-see?"

13:47 <FreeLarry58> lol wouldn't know - last time i was on one didn't have a seat

13:48 * FreeLarry58 wonders if the person at top of wheel forgets and tries to drive off to favorite fast food joing

13:49 * FreeLarry58 watch out for that first speed bump - its a dozzie

13:51 *** PovAdct_w has joined #boinc

14:01 <hawmps> 'lo Larry

14:04 <FreeLarry58> lo jay

14:04 <MTughan> Long-time Windows user takes Linux for a spin.

14:04 <Romulus> <> (at

14:04 <MTughan> And it seems I typed the URL wrong...

14:05 <MTughan>

14:05 <Romulus> <> (at

14:08 *** yoyo[RKN] has joined #boinc

14:14 *** Eythan has quit IRC

14:17 <FreeLarry58> howdy yoyo[RKN] - i see those new ogr wu's are long ons -- aprox 40 hrs indicated on the new machine

14:17 <MTughan> 'lo yoyo

14:18 <MTughan> Got around to ecm again, didn't realize there was a make check target in there. :P

14:18 <yoyo[RKN]> hi

14:18 <MTughan> I ran my build by it, it passed.

14:18 <MTughan> So I have an i386 build and a PowerPC build at this point.

14:18 <PovAdct_w> home ->

14:18 <yoyo[RKN]> FreeLarry58 : dnet server seems to send big ogr wus, which I put into boinc wus

14:19 *** PovAdct_w has quit IRC

14:19 <yoyo[RKN]> I saw 80h ogr wus

14:19 <FreeLarry58> lol its ok that machine has lots of memories - it's actually predicting 43.5 hrs to run

14:20 <yoyo[RKN]> just wait for the first 14% step

14:20 <FreeLarry58> and 80 hrs about what seti ap takes on that machine to (120+ hrs for same unit on other machines)

14:21 <FreeLarry58> currently its at 42.8%

14:21 <yoyo[RKN]> &wx sxf

14:21 <Romulus> yoyo[RKN]: Temperature: 39°F / 4°C | Humidity: 93% | Pressure: 29.68in / 1005hPa | Conditions: Light Rain | Wind Direction: SSE | Wind Speed: 5mph / 7km/h | Updated: 6:50 PM CET; Chance of Rain. High:46 F.; Chance of Rain. Low:33 F.; Chance of Rain. High:42 F.; Chance of Snow. Low:33 F.; Rain. High:42 F.; Chance of Rain. Low:39 F.;

14:22 <MTughan> yoyo[RKN]: Are you involved any with the development of ecm?

14:22 <yoyo[RKN]> yes, there is a thread in mersenneforum

14:23 <MTughan> Because I get a compiler error with some assembler when trying x86_64.

14:23 <MTughan> "32-bit absolute addressing is not supported for x86-64"

14:23 <yoyo[RKN]>

14:23 <Romulus> <> (at

14:24 <MTughan> This is ecm, not gmp.

14:24 <MTughan> I know gmp doesn't have any specific assembler.

14:25 <yoyo[RKN]> for gmp I do not have any contact

14:25 <MTughan> gmp for x86_64 works fine, compiles and passes make check.

14:28 <MTughan> I've got to head off to class... I'll look through it later, although it says that gmp 4.3 brings x86_64 asm and some speedups.

14:28 <yoyo[RKN]> ok

14:51 <hawmps> spork >>

15:55 <xcamel> re-hi

15:59 <PovAddict> gmp has lots of specific assembly code :)

16:00 <PovAddict> but has many versions, selects appropriate one, and has C code as fallback

16:04 *** zombie67 has quit IRC

16:05 <MTughan> PovAddict: Yeah, but using the C code doesn't help speed. ;)

16:05 <PovAddict> but I mean you're unlikely to get asm compiler errors from gmp

16:05 <MTughan> I don't get compiler errors from gmp.

16:07 *** up365 has quit IRC

16:20 <xcamel> spork >> (web hack)

16:47 *** wdsmia_w has quit IRC

16:50 *** Eythan has joined #boinc

17:06 <CoderForLife> re-hi

17:06 <CoderForLife> thx for the sporkzies xcamel

17:09 <xcamel> 'elo

17:09 <FreeLarry58> you being hacked at again xcamel?

17:10 <xcamel> just the usual suspects..

17:10 <xcamel> haxbotz

17:10 <FreeLarry58> ugg waste of internet bw

17:10 <xcamel> or obama voters

17:12 <FreeLarry58> wellll not sure about him yet - i gave bush his chance now i have to wait and see if o-bAm-a can do anything worthwhile

17:13 <CoderForLife> he'll have dug a big hole by then

17:13 <FreeLarry58> you mean a deeper one than we already in?

17:13 <xcamel> much deeper.

17:14 * FreeLarry58 lost the sight of light at end a long time ago

17:14 <CoderForLife> mass extinction impactor

17:14 <FreeLarry58> ohhh the next oil field in making

17:15 <xcamel> except all that sludge will be whatever wealth you ever thought you had...

17:16 <CoderForLife> I was thinking much deeper, and with all the associated collateral damage

17:16 <FreeLarry58> naww it will be us as wealth has no oil in the paper and metal its made from

17:17 <FreeLarry58> assuming we don't go up in greatbig flash that is

17:18 * FreeLarry58 thinks time to go back to light reading again -- tca

17:19 <xcamel> Ayn Rand - Atlas Shrugged, or at least "The Fountainhead"

17:21 <CoderForLife> required reading these days

17:24 * CoderForLife waves his Cliff's Notes version of Atlas Shrugged

17:26 <CoderForLife> Who is John Galt?

17:29 <CoderForLife> down to the Coding Couch

17:33 <xcamel> I have seen a few black ball caps with "I Am John Galt" on them

17:34 <xcamel> but the wearers don't seem to get the irony in that...

17:39 <wdsmia> re-hi

17:41 <xcamel> 'lo Doug

17:42 <MTughan> Soul_keeper: Well, you just "got" a rebuke on that GPL thread, although not from any of the devs.

17:53 <Soul_keeper> they shoulda never released the binary then if what he's saying has any reason

17:54 <PovAddict> legally, he took code with no license

17:54 <Soul_keeper> as far as his comment about being an insult to ask for the source. Well if helping their project is an insult then i dunno what to tell them.

17:54 <PovAddict> so it's not even about releasing: he wasn't even allowed to modify it for personal use

17:55 <xcamel> cuda again?

17:55 <PovAddict> there's nothing in the code saying what he's allowed to do

17:55 <Soul_keeper> doesn't hurt to ask to help imo

17:55 <PovAddict> xcamel: MW ATI app

17:55 <xcamel> ah..

17:55 <PovAddict> Soul_keeper: he could have done two things... modified the newer code, licensed as GPL and clearly stating his rights; or asked for permission to the copyright holder (Travis) before touching the old license-less code

17:55 <xcamel> slipstream or something..

17:55 <Soul_keeper> notice i'm not speaking there.

17:57 <MTughan> Yeah, the MW app was a license mess before... At least it's GPL'd now, but the licenselessness from before is coming back to haunt us.

17:57 <Soul_keeper> he's probably using the project as some kinda school project, so he can put it on his resume

17:57 <Soul_keeper> i'm not trying to hiijack anything

17:59 *** yoyo[RKN] has quit IRC

18:02 <Soul_keeper>

18:02 <Romulus> Title: SDSS Data Release 7 (at

18:03 <Soul_keeper> their whole project is based completely on freely available data anyways

18:03 <MTughan> Got a warning from rr_sim... "format %d expects type int, but argument 6 has type size_t"

18:03 <PovAddict> wrong win

18:03 <PovAddict> ^ MTughan

18:04 <MTughan> Oops... Thanks. :)

18:12 *** DerMeister has quit IRC

18:22 <xcamel> the sloan scope is an amazing piece of hardware for it's size..

18:30 *** zombie67 has joined #boinc

18:30 <zombie67> MT?

18:30 <Romulus> Somebody said MT is right, zombie67

18:30 <zombie67>

18:31 <Romulus> <> (at

18:32 <MTughan> I don't have an nQueens account.

18:51 <xcamel>

18:51 <Romulus> <> (at

18:51 <xcamel> and..

18:51 <xcamel>

18:51 <Romulus> <> (at

18:54 *** Aeternus has quit IRC

18:55 <zombie67> MT, can I PM him with a way to contact you?

18:55 <MTughan> Sure.

18:55 <zombie67> what method do you prefer?

18:55 <MTughan> Tell him to email me at

18:55 <zombie67> ah

18:56 <MTughan> If he wants to cross check that, it's the email address I use on the BOINC mailing lists.

18:57 <MTughan> I'm going to dinner. bbl

18:57 <PovAddict>

18:57 <Romulus> <> (at

18:57 <MTughan> HTTP 503.

18:57 <PovAddict> refresh till it works

18:57 <PovAddict> dunno what's up with that server

18:57 <MTughan> Yep... What is it?

18:57 <PovAddict> zoomed in

18:57 <Romulus> <> (at

18:58 <MTughan> That's a lot of bottles...

18:58 <MTughan> I'm gone.

19:03 <zombie67> cool.  looks like they are in a pile waiting to be recycled.  it's all goodness.  The problem are all the bottles that DON'T make it into that pile

19:04 <PovAddict> it's not a real picture

19:04 <PovAddict> it's to give you an idea of numbers

19:04 <PovAddict> "Depicts two million plastic beverage bottles, the number used in the US every five minutes."

19:05 <zombie67> again, I don't care how big the number, so long as they get recycled.

19:05 <PovAddict>

19:05 *** infinisoft has joined #boinc

19:05 <Romulus> <> (at

19:05 <PovAddict>

19:05 <Romulus> <> (at

19:05 <PovAddict> "Depicts 170,000 disposable Energizer batteries, equal to fifteen minutes of Energizer battery production."

19:06 <zombie67>

19:06 <Romulus> Title: What does one TRILLION dollars look like? (at

19:06 <PovAddict> I saw that one (in your facebook)

19:06 <zombie67> now THAT is amazing

19:07 <PovAddict> I remember someone was sued for a few trillion dollars for music piracy

19:07 <PovAddict> RIAA calculated like a hundred thousand per song

19:07 <PovAddict> absolutely ridiculous

19:08 <zombie67> RIAA can lick my left nut

19:10 <zombie67> every time one machine (XP64) machine boots, I get the following error message pop-up:

19:11 <zombie67> The procedure entry point LocalizeString could not be located in the dynamic link library STLang.dll

19:12 <PovAddict> open task manager, see if the error popup appears in the list of Applications

19:12 <PovAddict> right click, Go to process... should tell you what program caused the popup

19:14 <zombie67> sstray.exe is listed.  but right click has no "go to process"

19:14 <zombie67> oh, apps, not processes

19:14 <PovAddict> I mean in applications, not in process list

19:15 <PovAddict> while the error is still shown

19:15 <zombie67> that took me to csrss.exe

19:15 <PovAddict> argh

19:15 <PovAddict> would be worse if it was svchost.exe; then it could be almost any service in your system :P

19:16 <xcamel> 'Released Gitmo detainee becomes So. Afghan terrorist leader"

19:16 <xcamel> friggin brilliant.

19:16 <infinisoft> Actually you can check the command line for each svchost and at least see which service group it was

19:17 <PovAddict> xcamel: hah, that's what they get for releasing... idiots

19:17 <PovAddict> I heard cases here where a Plain Old Robber gets caught stealing on the same day he got released

19:18 <zombie67> should I trust this?

19:18 <Romulus> <> (at

19:18 <PovAddict> "stray files, orphaned startup tasks, corrupted drivers" <- if you want to fix those, get CCleaner

19:19 <Soul_keeper> i don't trust any of those things ... glad i use linux :)

19:19 <zombie67> many pages seem to point to the audio driver for that error message

19:21 <xcamel>

19:21 <Romulus> <> (at

19:21 <xcamel> prisoner #0008

19:22 <PovAddict> zombie67: it seems that program isn't free, I see mentions of serials on the web

19:22 <zombie67> yeah, i am going to avoid it

19:22 <PovAddict> very possibly the program you download tells you "we found all these problems in your computer, pay us $$ for a version that actually fixes them"

19:22 <PovAddict> it says "free scan", not "free fix" :)

19:23 <zombie67> right.  like AV/spyware programs.  It makes me very wary, when they have a vested interest in perpetuating virus/worms

19:24 <zombie67> okay, I finally removed ALL .NET versions

19:24 <PovAddict> there are also a LOT of antispyware programs that are really fake

19:24 <zombie67> rebooted

19:24 <zombie67> and am now doing an install

19:24 <zombie67> .NET takes FOREVER to install

19:24 <zombie67> this is my last attempt before I swap the card into my CUDA box.

19:29 <PovAddict>

19:29 <Romulus> <> (at

19:29 <PovAddict> ^ see, just like that :)

19:29 * KathrynM groan

19:29 <KathrynM> groans too

19:33 <MTughan> xcamel: Guantanamo Bay isn't all good. They've detained a Canadian there who's been proven to be innocent.

19:33 <MTughan> Omar Khadr.

19:33 <Romulus> New news from boinc: BOINC news March 9, 2009

19:33 <xcamel> they "proved" this guy innocent too..

19:36 <Soul_keeper> i thought that was a german guy

19:37 <MTughan> Khadr?

19:38 <xcamel> 88 released, 43 known to have gone back to terrorist activities, 12 have become terrorist leaders, 9 have already committed terrorist acts.

19:39 <MTughan>

19:39 <Romulus> Title: Omar Khadr - Wikipedia, the free encyclopedia (at

19:40 <Soul_keeper>   awesome, nvidia is throwing around millions to get more cuda devs

19:40 <Romulus> <> (at

19:41 <MTughan> "Interestingly, NVIDIA says that it will still invest in companies who are working on GPU technology even if the firm is working with ATI GPUs."

19:41 *** infinisoft has quit IRC

19:41 <MTughan> Ha, just means they'll be looked at, and passed by.

19:42 <PovAddict> it means they'll invest a few cents

19:42 <PovAddict> (metaphorically speaking)

19:42 <xcamel> nvidia needs to take that code open source

19:42 <MTughan> Is OpenCL open source?

19:42 <PovAddict> (just like someone who said $40 million of revenue (MySQL AB.) was pocket change)

19:42 <Soul_keeper> MTughan, yes

19:43 <Soul_keeper> it'll be like opengl, available to everyone

19:43 <MTughan> I was just wondering if it wasn't despite the name.

19:43 <MTughan> Like Apple's original Apple Public Source License.

19:43 <PovAddict> "like opengl" would mean an open specification and possibly-closed reference implementation

19:44 <Soul_keeper> pretty much the same members on the opencl board as the opengl board

19:45 <Soul_keeper>

19:45 <Romulus> <> (at

19:48 <xcamel> rumor has it that the M$ GPU code is or will be open source

19:48 <MTughan> Emphasis on rumor.

19:48 <xcamel> otherwise they face the same fate as 3DFx and voodoo

19:49 <xcamel> I was there when that all went down in flames

19:50 <xcamel> and beta testing riva 128 chips at the same time

19:50 <zombie67>

19:50 <Romulus> <> (at

19:51 <xcamel> matrox was also in it up to their necks, and blew it in chunks.

20:01 <xcamel> I was a bit sharper tool in the shed back then, programming for 68030, 68040, and Targa gfx boards

20:04 <Soul_keeper> :)

20:04 *** siofwolves has quit IRC

20:21 <MTughan>

20:21 <Romulus> <> (at

20:22 *** efc has joined #boinc

20:24 <Soul_keeper>     where would you sit in a feild like this ?

20:24 <Romulus> <> (at

20:29 <MTughan> Okay, the latest response from Cluster Physik was just smug...

20:31 <xcamel> sk - for what event?

20:31 <Soul_keeper> baseball

20:31 <Soul_keeper> season starts right around my bday

20:32 <zombie67> DING DING DING!  After another 4 hours today, I finally got the 2nd machine working!!

20:32 <Soul_keeper> opening season game is the day after my bday

20:32 <zombie67> uninstalled all versions of .NET, each one taking 10-15 minutes.

20:33 <PovAddict> BOINC apps really shouldn't depend on .NET...

20:33 <zombie67> reinstalled .NET 3.5, takes about an HOUR to d/l and install (what's up with that?!) and that also installed all the previous versions

20:33 <zombie67> Reinstalled 8.12 driver and CCC

20:33 <zombie67> So no idea why it worked this time. Just random luck I guess.

20:33 <Soul_keeper> what is the problem ?

20:33 <Romulus> I guess the problem is that "WorldCommunityGrid" is viewed by gridrepublic as a single project in the resource partition process, Soul_keeper

20:34 <zombie67> I don't think the app requires .NET, ATI driver and CCC require .NET

20:34 <Soul_keeper> forget problem

20:34 <Romulus> Gotcha.

20:34 <MTughan> He couldn't get his second ATI card working.

20:34 <Soul_keeper> working with boinc, or just working in general ?

20:34 <PovAddict> forget the problem

20:34 <Romulus> 10-4, PovAddict.

20:34 <zombie67> altogether, about 9 hours down the drain

20:35 <zombie67> just passed 750k in CPDN

20:35 <zombie67> oo!  a hydrogen task!

20:37 <zombie67> just passed 2M in PS3GRID

20:37 <PovAddict> took you a while

20:38 <PovAddict> j/k

20:46 <zombie67> POV:  see my last post in this short thread.  Am I confused, or did I find a problem with the credit awarding logic?

20:46 <zombie67>

20:46 <Romulus> <> (at

20:48 <PovAddict> zombie67: interesting, I'll send it to boinc_projects; or directly to the one who developed that feature (I think it was Kevin Reed)

20:50 <zombie67> Thanks!

20:50 <zombie67> so I'm not crazy?

20:51 <PovAddict> at least it made some sense

20:52 <KathrynM> omfg.  my boss is actually letting me take a sick day

20:52 <PovAddict> :O

20:52 <MTughan> Doesn't usually?

20:52 <KathrynM> nope

20:52 <KathrynM> you basically have to be in the hospital to miss work

20:53 <KathrynM> no substitute teachers here

20:53 <KathrynM> I have no voice and I can barely hear because my ears are so clogged up

20:54 <PovAddict> then what's the point of making you go to work

20:54 <PovAddict> not like you can teach

20:54 <KathrynM> warm body in the classroom

20:55 <zombie67> baby sitter

20:55 <KathrynM> pretty much

20:55 <zombie67> if there is no teacher, then all the parents may have to take off work too

20:55 <zombie67> huge domino

20:55 *** Fwerp has joined #boinc

20:55 <Fwerp> Recent average credit 749.58

20:55 <KathrynM> well, I'm in different classes all day.  Our periods are only 25 minutes


20:55 <zombie67> when the teacher isn't there, the sub's primary role is to take care of the kids, not teach

20:56 <KathrynM> zactly

20:56 <KathrynM> I used to be a sub in the states

20:57 <efc> my mom too

20:58 *** kmeyer has joined #boinc

20:58 <kmeyer> boinc is giving me zero tasks

20:58 <kmeyer> any ideas why/

20:59 <zombie67> a sick teacher w/o a sub can impact the economy

21:01 <zombie67>

21:01 <Romulus> <> (at

21:01 <PovAddict> kmeyer: what project?

21:01 <kmeyer> rosetta

21:01 <kmeyer> among others

21:01 <KathrynM> this sick teacher is going to bed

21:01 <KathrynM> laters

21:02 <PovAddict> good luck getting that sickness away!

21:02 <xcamel> hope you feel better soon!

21:05 <zombie67> what is SLI called for ATI cards?

21:05 <zombie67> crossfire?

21:06 <MTughan> CrossFireX.

21:06 <zombie67> thanks

21:07 <zombie67> .19d knows multiple GPUs, right?  does that include crossfire?

21:07 <MTughan> "- multi GPU support (still experimental, can be turned off by exclusion of all but one GPUs in

21:07 <MTughan>    app_info.xml)"

21:08 <MTughan> Presumably, that means not Crossfire, in the same way that two Nvidia GPUs can't be in SLI to work in BOINC.

21:08 <zombie67> dang

21:08 <MTughan> Wanted to put the two cards together?

21:08 *** mchelen has joined #boinc

21:09 <zombie67> well, you can put the 2nd card in a plain PCI slott, right?

21:09 <zombie67> slot

21:09 <MTughan> No, it has to be PCIe.

21:09 <zombie67> hrm

21:09 <zombie67> really?

21:10 <MTughan> Yep, PCI and PCIe have different physical slots to prevent you from doing that.

21:10 * xcamel still lusts after Tesla

21:10 <MTughan> PCI is a parallel bus, PCIe is a serial bus AFAIK.

21:11 <zombie67> okay, let me re-ask

21:11 <zombie67>

21:11 <Romulus> <> (at

21:11 <zombie67> 2 PCI-e x8 slot wired as x4

21:11 <zombie67> 1 PCI-e x16 graphics slot

21:12 <zombie67> you can use one of those non-graphics slots for the 2nd card?

21:12 <MTughan> Yes, but there's only a possibility it'll fit.

21:13 <zombie67> okay.  So then THAT is why I want do be able to use crossfire/SLI

21:13 <MTughan> Because a PCIe x8 slot is shorter than an x16 slot, you'll have to see if the back end of the slot doesn't have any plastic over it. If it doesn't, that'll allow the rest of the back of the card to go out the rear of the slot.

21:14 <zombie67> that is a way to increase crunching capacity w/o replacing a mobo or whole machine

21:15 *** infinisoft has joined #boinc

21:16 <MTughan> Aren't all three PCIe slots right next to each other? And aren't the two cards you got both dual-slot cards?

21:18 <zombie67> yeah, it wouldn't work for my two boxes.

21:18 <zombie67>

21:18 <Romulus> <> (at

21:19 <zombie67> but in general, it would be a good thing to support.

21:19 <MTughan> What would be? Crossfire?

21:20 <zombie67> yes, and SLI

21:20 <zombie67>

21:20 <Romulus> <> (at

21:20 <zombie67> push push!

21:20 <MTughan> The problem with SLI and Crossfire isn't in the application.

21:21 <zombie67> oh?

21:21 <Romulus> Rumor has it oh is he on, zombie67

21:21 <MTughan> IIRC, both of them take two GPUs and make them appear as one to the host OS and its programs.

21:23 <zombie67> okay, so let's take MW as an example

21:23 <zombie67> so then couldn't I just up the setting to 4 tasks at once?

21:23 <MTughan> But what GPU does it run on? It only sees one, so it sends all the jobs to that one.

21:24 <MTughan> And I don't think Crossfire is smart enough to distribute that load across both GPUs.

21:25 <MTughan> Because I think it's just designed to spread the load of pixel crunching, not GPU specific GPGPU.

21:26 <zombie67> hm

21:26 <zombie67> 6.6.15 is out

21:26 <zombie67> QMC will be wrapping up the current project soon

21:27 <zombie67>

21:27 <Romulus> <> (at

21:36 <zombie67> I now have 4 machines running QCN.  1 a mac laptop, 3 XP machines with external usb sensors

21:45 *** BadBarbarian has joined #boinc

21:48 <zombie67> BB!

21:50 *** mchelen has quit IRC

21:52 *** mchelen has joined #boinc

22:06 <zombie67> MT:  quick Q, I think

22:06 <MTughan> Shoot.

22:06 <zombie67> my 8-core machine, with default 2 in the app info

22:06 * PovAddict bets a dollar it will be as quick as last night's

22:07 * MTughan is listening to "Don't Download This Song" by Weird Al Yankovic

22:07 <zombie67> a resource share (all things being equal) of 20% MW will make MW run 2 tasks at all times, right?

22:07 <MTughan> The irony is I downloaded it. xD

22:07 <zombie67> this is with ncpus of 9

22:07 <PovAddict> [01:04:14] <zombie67>  sorry, I thought this would be quick and simple

22:07 <MTughan> It should, provided the debts are equal.

22:08 <zombie67> POV, I know, I have a bad habit of asking Qs I think will be quick

22:08 <PovAddict> not necessarily... it may very well run 9 tasks at once (and then compensate with other projects to meet the share)

22:08 <zombie67> okay, so if I crank up the resource share so that it is 40%, what happens?

22:09 <zombie67> POV, yes, but I am assuming for this equal debt

22:09 <MTughan> PovAddict: MW's GPU app_info tells BOINC is uses 0.5 cores.

22:09 <MTughan> That's what he meant by "with the default 2", because it'll run 2 WUs at once.

22:10 <PovAddict> it depends, for example, on work it gets from other projects

22:10 <zombie67> (all things being equal)

22:10 <PovAddict> for example, suppose you download 10 tasks... it will start *those* if it didn't get work from other projects yet :)

22:11 <zombie67> [19:08]  <zombie67> okay, so if I crank up the resource share so that it is 40%, what happens?

22:11 <zombie67> will it run 4 tasks, even though the app_info says 2?

22:11 <PovAddict> the app_info says two tasks per core

22:11 <PovAddict> two cores free, four tasks

22:13 <zombie67> MT?

22:13 <Romulus> I guess MT is right, zombie67

22:13 <zombie67> that goes without saying

22:13 <PovAddict> lol

22:13 <PovAddict> PovAddict?

22:13 <Romulus> hmm... PovAddict is the only one who could really help you right now, PovAddict

22:13 <zombie67> TRVTH

22:14 <MTughan> rofl

22:14 <PovAddict> zombie67?

22:14 <Romulus> Somebody said zombie67 is considering upgrading to leopard, PovAddict

22:14 <PovAddict> MTughan?

22:14 <Romulus> MTughan is in need of a pony for xmas, PovAddict

22:14 <PovAddict> o_O

22:14 <zombie67> wow, how old is THAT?

22:14 <MTughan> WTF?!

22:14 <PovAddict> roflcopter

22:14 <MTughan> MTughan is compiling apps for Milkyway.

22:14 <MTughan> How does that thing learn anyway?

22:15 <zombie67> '/

22:15 <PovAddict> MTughan?

22:15 <Romulus> I guess MTughan is in need of a pony for xmas, PovAddict

22:15 <PovAddict> forget MTughan

22:15 <Romulus> Okay!

22:15 <PovAddict> MTughan is compiling apps for Milkyway.

22:15 <PovAddict> MTughan?

22:15 <Romulus> Somebody said MTughan is compiling apps for Milkyway, PovAddict

22:15 <MTughan> Thanks... :P

22:15 <zombie67> so MT, what about my question?

22:15 <PovAddict> [mar dic 23 2008] [02:53:44] <Soul_keeper> MTughan is in need of a pony for xmas

22:15 <PovAddict> (= tuesday december 23th)

22:15 <MTughan> zombie67: What question?

22:15 <PovAddict> [00:11] <zombie67> will it run 4 tasks, even though the app_info says 2?

22:15 <PovAddict> hmm

22:16 <PovAddict> DA?

22:16 <Romulus> DA is a pain, PovAddict

22:16 <PovAddict> !

22:16 <MTughan> zombie67: The app_info doesn't specify two directly, does it?

22:16 <zombie67> okay, so if I crank up the resource share so that it is 40%, what happens?

22:16 <zombie67> will it run 4 tasks, even though the app_info says 2?

22:16 *** mchelen1 has joined #boinc

22:16 <MTughan> It just specifies the avg_ncpus to be 0.5, right?

22:16 <zombie67> yes

22:16 <zombie67> and that means....?

22:17 <MTughan> So it'll run as many tasks as it think it needs to fill up the CPUs.

22:17 <MTughan> But on average, I think that means it'll use another core, which it doesn't need to.

22:17 <PovAddict> I don't think it's possible to answer your question zombie67, it depends on many other factors, and I don't understand the scheduler enough

22:17 <zombie67> okay, so "yes"?

22:17 <zombie67> let me rephrase

22:17 <PovAddict> we need a scheduler simulator (with a better user interface than what it exists), to do "what if" tests

22:18 <Soul_keeper> PovAddict, did you ever get that pony ? :)

22:18 <PovAddict> ...

22:18 <zombie67> It can/may run more than 2 tasks in that scenario?

22:18 <PovAddict> possibly

22:18 <zombie67> assuming "yes", and assuming 4 tasks at once, are they running at half speed on the GPU?

22:19 <PovAddict> possibly

22:19 <MTughan> Yes.

22:19 <PovAddict> maybe even crash if they fill VRAM

22:20 <PovAddict> even if BOINC supported ATI GPUs directly, it's still lacking features with CUDA (like VRAM checking)

22:20 <MTughan> PovAddict: These won't crash with VRAM. I've run 30 tasks at once without filling VRAM.

22:20 <Soul_keeper> geeze

22:20 <zombie67> wdsmia said yesterday around this time, that you could get more credits/hr by running more than two tasks at once

22:20 <Soul_keeper> so much for jokes

22:20 <zombie67> so I am trying to understand if that is correct or not

22:21 <PovAddict> depends if the MW app uses all the GPU cores fully or not

22:21 <PovAddict> (remember GPUs have hundreds of "cores")

22:21 <zombie67> I thought 4 tasks, two at a time, were equal to 4 tasks at once

22:21 <MTughan> I've seen average 96%.

22:22 <zombie67> so I could squeze out a bit more by doing (say) 3 at a time?

22:22 <PovAddict> if you run four CPU tasks on a dual core, it's on average the same (might be slightly faster for I/O, slightly slower for CPU cache issues)

22:22 <Soul_keeper> you could write a "kernel" to run however many tasks you want at a time essentially

22:23 <PovAddict> same if you run two 2-thread tasks on a dual core

22:23 <PovAddict> so the thing here is... how many cores, of the 200 the GPU has, can the MW app take advantage of?

22:23 * Soul_keeper scurries back into the shadows

22:23 <zombie67> like HT can give a bit more than without HT

22:23 <PovAddict> if it uses 50 of 200, then running 4 tasks instead of 2 gives you 2x speedup

22:23 <MTughan> &math 2 / 0.04019

22:23 <Romulus> MTughan: 49.7636227917

22:24 <PovAddict> since running 2 tasks you have half the GPU *idle*

22:24 <MTughan> PovAddict: I've seen on average 96% load with one task running.

22:24 <PovAddict> but I don't know how many sharers it uses

22:24 <PovAddict> MTughan: load measured where?

22:24 <MTughan> GPU-Z.

22:24 <PovAddict> hmm interesting tool

22:24 <zombie67> wait...96% running one or running two tasks at once?

22:24 <MTughan> One.

22:24 <PovAddict> you'd have to test it

22:25 <PovAddict> run a second one and see if the first one drops 50% performance

22:25 <zombie67> so if that is the case, then running the default two is already over the 100% and using as much capacity as possible

22:25 <zombie67> then there should be NO benefit in running more than 2

22:25 <MTughan> &math 4 / 0.08031

22:25 <Romulus> MTughan: 49.8069978832

22:26 <MTughan> At least it's linear...

22:26 <zombie67> the problem is that the tasks run so FAST, you can never really be sure how many are running at any one time

22:27 <zombie67> ding ding ding and ove

22:27 <zombie67> r

22:27 <zombie67> POV:  you were right.

22:27 <zombie67> it wasn't a quick Q

22:28 <PovAddict>

22:28 <Romulus> <> (at

22:28 <PovAddict> 256 cores

22:28 <PovAddict> "Note that the current utilization on this system is 91%!! Somebody’s running something big — maybe Folding(at)Home, development for the next Pixar flick, or maybe Vista without SP1."

22:28 <PovAddict> lol @ latter

22:29 <zombie67> brb

22:29 *** zombie67 has left #boinc

22:34 <MTughan> PovAddict: Sure you have the right image link? I get redirected to the homepage.

22:35 *** mchelen has quit IRC

22:35 <PovAddict>

22:35 <Romulus> <> (at

22:35 <MTughan> Wow, that's a lot of cores...

22:35 *** zombie67 has joined #boinc

22:35 <PovAddict> zombie67:

22:35 <Romulus> <> (at

22:36 <zombie67> awesome!  what is it?

22:36 <zombie67> what cpu, i mean

22:37 <PovAddict> probably just virtualized CPUs...

22:37 <zombie67> i am running gpu-z

22:37 <PovAddict> probably a giant cluster with separate *physical* CPUs

22:37 <zombie67> says gpu load 96%

22:37 <PovAddict> or one of those CPUs with different "logical units" (whatever that means) on each core

22:38 <zombie67> wow!  running 6 MW tasks!

22:38 <zombie67> load is 97%

22:38 <PovAddict> but do they seem to run at 1/6th speed?

22:38 <zombie67> answered my own Q about more than 2

22:38 <zombie67> how can i tell?

22:39 <PovAddict> look at how fast the progress bar moves! :)

22:39 <zombie67> progress bar is odd with these tasks, and the clock is only CPU time

22:40 <PovAddict> then run one at a time, wait for it to complete

22:40 <PovAddict> run two, see if it completes in a similar amount of time

22:40 <zombie67> They do seem to be completing more slowly

22:40 <PovAddict> same with six

22:40 * FreeLarry58 still not happy with gpugrid on the one machine that can run it - indicats a little over 5 minutes cpu time but has been running for close to 48hrs and still <50%

22:41 <zombie67> I wish there were two columns, once for CPU time and one for GPU time

22:42 <FreeLarry58> now the machine is stuck just running gpu wu as in high priority mode and rest of projects just sit there with cpu running way <100%

22:42 <zombie67> wow, I have ncpus set to 9, and 14 tasks are running

22:42 <zombie67> one is non-cpui

22:42 <zombie67> non-cpu

22:43 <PovAddict> zombie67: I think there is no easy way to get "GPU time"

22:44 <MTughan> zombie67: The application records wall time in the output, which can be viewed online.

22:44 <Soul_keeper> i still have 180 seti cuda WU's to finish here ...

22:45 <zombie67> ah, let me check the web page

22:46 <zombie67> they all seem to be about 1.7 to 2

22:46 <zombie67>

22:46 <Romulus> <> (at

22:46 <MTughan> CPU seconds? No, look in the output.

22:46 <PovAddict> if anyone finds anything in cuda api to get GPU time, let me know...

22:46 <MTughan>

22:46 <Romulus> <> (at

22:46 *** up365 has joined #boinc

22:46 <zombie67> oh, crap, you have to look at each one

22:46 <MTughan> WU completed. It took 1.64063 seconds CPU time and 162.846 seconds wall clock time @ 2.66 GHz.

22:47 <Soul_keeper> i figure it'll take me just under 12 days to complete them, that's far more than the 10 days of "work buffer" considering it's been at them for over 5 days and sent those in

22:47 <PovAddict> zombie67: I can write you a scraper tomorrow to get all of them

22:47 <FreeLarry58> MW and seti gpu wu's wont run on the machine that can do gpugrid ;( don't know why

22:48 <zombie67> the problem:  There are so many results, and who can tell how many were running at the time that particular task was crunched?

22:48 <MTughan> Yeah, that is a problem.

22:48 <PovAddict> doesn't log start and end time, does it?

22:48 <zombie67> the only real way to tell, I think, is to play with the settings, and then look at your daily stats

22:48 <zombie67>

22:48 <Romulus> <> (at

22:49 <zombie67> that is a result

22:49 <MTughan> PovAddict: Actually, the BOINC logs would, right?

22:49 <PovAddict> yes that may work

22:49 <MTughan> I've got to close some tabs in FF here... 99 open.

22:50 <PovAddict> Konqueror using 184MB with a single tab open

22:50 <PovAddict> because I had 20 tabs open a while ago

22:50 <MTughan> Firefox taking 733MB.

22:50 <PovAddict> heh

22:50 <MTughan> I'm not running ChatZilla.

22:50 <PovAddict> Macs for some reason seem to use too much RAM

22:50 <PovAddict> almost seem to leak

22:51 <zombie67> with which app?

22:51 <MTughan> Think again. This is a Win7 machine.

22:51 <MTughan> zombie67: No, overall. OS X applications are very heavy on RAM.

22:51 <PovAddict> I did a CTCP VERSION and it said you were on Vista, heh

22:51 <MTughan> Really?

22:51 <PovAddict> xchat 2.8.6-1 Windows Vista [Intel /3.27GHz]

22:51 <zombie67> I don't see that

22:51 <PovAddict> report the bug :D

22:52 <PovAddict> sleep ->

22:52 <zombie67> now I am down to 9 CPU tasks running an 0 MW tasks

22:52 <zombie67> ugh

22:52 <MTughan> Night.

22:52 <zombie67> GN MT!

22:52 <PovAddict> me, not him

22:52 <MTughan> No, night to Pov. :P

22:52 <zombie67> ah

22:52 <PovAddict> bai

22:53 <zombie67> BOINC really needs ATI support...this is messy and confusing

22:53 * PovAddict still welcoming suggestions on a name for the async boinc gui rpc library

22:53 <PovAddict> zombie67: it has yet to get fully working cuda support ;)

22:53 * PovAddict really goes

22:54 <Soul_keeper> by the time they get complete  cuda and stream support opencl will start being used ...

22:55 <zombie67> doing 9 CPU tasks on an 8 cpu machine is bad, and the idle GPU is horrible

22:56 *** Soul_keeper has left #boinc

22:59 <MTughan> bbiab, gonna try updating Xchat first...

22:59 *** MTughan has quit IRC

23:00 *** MTughan has joined #boinc

23:00 <zombie67> there is an update to x-chat aqua?

23:01 <MTughan> Dunno.

23:01 *** up365 has quit IRC

23:01 <zombie67> answer = nope

23:01 <zombie67> you use that too?

23:02 <MTughan> On OS X?

23:03 <zombie67> yes

23:03 <MTughan> Sorry, just trying to concentrate on other stuff too...

23:03 <MTughan> No, I use Colloquy on OS X.

23:03 <zombie67> ah

23:04 <zombie67> I find x-chat to be more OSX'y, but with less features

23:04 <MTughan> I find it the other way around.

23:04 <zombie67> really?

23:04 <MTughan> Could be that XChat Aqua doesn't use Cocoa, but Carbon, so it stands out when it isn't themed like everything else.

23:04 <zombie67> what features do you miss with colloquy?

23:05 <MTughan> Not much. I find it easier to program for Colloquy.

23:05 <zombie67> ah

23:05 <zombie67> I don't do that

23:05 *** desti_T2 has joined #boinc

23:05 <zombie67> fyi, I sent the NQ admin your address

23:06 *** MTughan has quit IRC

23:07 *** MTughan has joined #boinc

23:07 <MTughan> No idea what happened there...

23:22 *** desti has quit IRC

23:22 *** ball has joined #boinc

23:22 <ball> How do I retire a system from boinc gracefully?

23:23 <ball> This machine will just catch fire if I try that much more.

23:23 <ball> I would like it to finish its current work unit and then not fetch a new one.

23:23 <efc> no new work, wait till it finishes work, disconnect all projects

23:25 <ball> Thanks, I think that'll do it.

23:26 <ball> We'll find out in a week or so.

23:27 <efc> I don't think it hurts anybody's feelings if you just pull the plug either. just delays somebody's credits.

23:28 * ball nods

23:28 <ball> I wanted to do it the right way thoughh.

23:34 <zombie67> ah!  now I know what you meant when you said you were working on MW apps

23:36 <MTughan> :)

23:38 <zombie67> Mr. Expert on ME boards!

23:38 <zombie67> MW

23:55 *** up365 has joined #boinc

Generated by 2.4 by Marius Gedminas - find it at!