IRC logs of #boinc for Tuesday, 2011-09-27

00:51 *** Haldrik has joined #boinc

01:08 *** SimoneCnt has quit IRC

02:38 *** desti has quit IRC

02:46 *** Caterpillar has joined #boinc

02:52 *** desti has joined #boinc

03:13 *** Tank_Master has quit IRC

03:28 *** up365 has quit IRC

03:41 *** Marcofe has joined #boinc

03:41 *** Marcofe has quit IRC

03:41 *** Marcofe has joined #boinc

03:55 *** _Danilo_ has quit IRC

04:14 *** up365 has joined #boinc

04:43 *** Edgeman has quit IRC

04:46 *** Caterpillar has quit IRC

05:28 *** Caterpillar has joined #boinc

05:48 *** |Caterpillar| has joined #boinc

07:38 *** Aeternus has joined #boinc

07:38 *** Aeternus has quit IRC

07:38 *** Aeternus has joined #boinc

07:39 *** Edgeman has joined #boinc

09:03 *** up365 has quit IRC

09:05 *** Marcofe has quit IRC

09:14 *** efc has quit IRC

09:20 *** Edgeman has quit IRC

09:22 *** yomshleeshee has joined #boinc

09:47 *** yomshleeshee_ has joined #boinc

09:50 *** yomshleeshee has quit IRC

10:09 *** Caterpillar has quit IRC

10:33 *** SimoneCnt has joined #boinc

10:43 *** yomshleeshee_ is now known as yomshleeshee

11:21 *** up365 has joined #boinc

12:13 *** Tank_Master has joined #boinc

12:30 *** Simone1 has joined #boinc

12:31 *** SimoneCnt has quit IRC

13:24 *** |Caterpillar| is now known as Caterpillar

13:48 *** denom_ has joined #boinc

13:54 *** denom has joined #boinc

14:06 *** denom_ has quit IRC

14:20 *** denom_ has joined #boinc

14:36 *** denom_ has quit IRC

14:51 *** denom_ has joined #boinc

15:06 *** denom_ has quit IRC

15:10 *** Simone1 is now known as Simone

15:19 <yomshleeshee> Anyone here know how a program's "Estimated app speed" in the client manager is calculated?

15:21 *** denom_ has joined #boinc

15:33 *** whynot has joined #boinc

15:37 *** denom_ has quit IRC

15:38 *** DerMeister has joined #boinc

15:51 *** Marcofe has joined #boinc

15:51 *** Marcofe has quit IRC

15:51 *** Marcofe has joined #boinc

15:53 *** denom_ has joined #boinc

16:08 *** denom_ has quit IRC

16:23 *** denom_ has joined #boinc

16:37 *** denom_ has quit IRC

16:40 *** up365 has quit IRC

16:42 *** up365 has joined #boinc

16:47 *** Marcofe has quit IRC

16:53 *** denom_ has joined #boinc

17:01 *** whynot has quit IRC

17:05 *** yomshleeshee has quit IRC

17:09 *** denom_ has quit IRC

17:13 *** DerMeister has quit IRC

17:20 <wdsmia> &wx 50319

17:20 <Romulus> wdsmia: Temperature: 68.7F / 20.4C | Humidity: 65% | Pressure: 29.86in / 1011.1hPa (Steady) | Conditions: Light Rain | Wind Direction: NE | Wind Speed: 4.0mph / 6.4km/h ; Tonight - Mostly cloudy. A 20 percent chance of showers through midnight. Low in the mid 50s. North wind 5 to 10 mph.; Wednesday - Mostly sunny. High in the mid 70s. Northwest wind around 5 mph.; Wednesday Night - Partly cloudy. (1 more message)

17:21 <wdsmia> &more

17:21 <Romulus> wdsmia: Low in the mid 50s. West wind near 10 mph.;

17:36 *** denom_ has joined #boinc

17:41 *** Caterpillar has quit IRC

18:02 *** mweltin has quit IRC

18:04 *** z0lty has joined #boinc

18:04 <z0lty> lo

18:05 <z0lty> i'm trying to set up boinc on freebsd...i've sucessfully attach to projects

18:05 <z0lty> but nothing happend

18:05 <z0lty> boinc_cmd --get_cc_status show me nothing

18:06 <z0lty> in get_state i have nrpc_failures: 3

18:06 <z0lty> what shoul i do ?

18:08 <z0lty> i tried to restart boinc deamon

18:12 *** mweltin has joined #boinc

18:16 *** Haldrik has quit IRC

18:20 *** Tank_Master has quit IRC

18:21 *** Tank_Master has joined #boinc

18:54 *** Edgeman has joined #boinc

18:57 *** licensed has joined #boinc

18:57 *** licensed has quit IRC

18:57 *** licensed has joined #boinc

19:22 *** zombie67 has joined #boinc

19:35 *** zombie67 has quit IRC

19:41 *** Aeternus has quit IRC

19:46 *** zombie67 has joined #boinc

19:46 <CoderForLife> is there any freebsd work available for your projects, z0lty?

19:47 <CoderForLife> &wx 45140

19:47 <Romulus> CoderForLife: Temperature: 59.7F / 15.4C | Humidity: 85% | Pressure: 28.80in / 975.2hPa (Steady) | Conditions: Mostly Cloudy | Wind Direction: West | Wind Speed: 0.0mph / 0.0km/h ; Overnight - Mostly cloudy this evening...then becoming partly cloudy. Scattered showers. Lows in the lower 50s. Southwest winds around 5 mph. Chance of rain 40 percent.; Wednesday - Mostly cloudy with a 50 percent chance (1 more message)

19:47 <CoderForLife> &more

19:47 <Romulus> CoderForLife: of showers. Highs in the upper 60s. Southwest winds 5 to 10 mph.; Wednesday Night - Mostly cloudy. A chance of showers...mainly in the evening. Lows in the lower 50s. West winds around 5 mph. Chance of rain 50 percent.;

19:47 <CoderForLife> &social

19:47 <Romulus> http://lledgerock.com/boinc_logs/boinc-current.png

19:47 *** yomshleeshee has joined #boinc

19:47 *** yomshleeshee has left #boinc

19:48 *** RomW has joined #boinc

19:48 <CoderForLife> hey RomW

19:48 <RomW> Howdy

19:48 <CoderForLife> how're things?

19:48 <RomW> Pretty good...

19:49 <RomW> T4T is moving BOINC in an interesting direction

19:49 <CoderForLife> what brings you back to this little quiet corner of the 'net?

19:49 <RomW> Hadn't stopped by in awhile

19:49 <CoderForLife> of that, I am sure =)

19:50 <CoderForLife> my job here is to stand guard, and chase away old friends by chatting them up with nonsense =)

19:50 <RomW> Well some of the guys at Oracle are pretty excited...  So we are getting things sorted out

19:50 <CoderForLife> ok, so what's T4T?

19:50 <xcamel> 'elo Rom & Don...

19:50 <CoderForLife> hey Jay

19:51 <RomW> Howdy Jay

19:51 <Tank_Master> RomW! howdy

19:51 <xcamel> Last load of hay in for the season tonight

19:51 <RomW> T4T is a BOINC based gateway for real CERN grid jobs

19:51 * CoderForLife sees the room spring spontaneously to life

19:51 * Tank_Master likes T4T

19:51 <CoderForLife> ah  cool  got a link?

19:51 <xcamel> interesting..

19:52 <RomW> Basically CERN put together a VM for downloading and running CERN Grid jobs using scientific linux

19:52 <CoderForLife> Google was my friend

19:53 <RomW> http://boinc01.cern.ch/test4theory/

19:53 <Romulus> Title: Test4Theory (at boinc01.cern.ch)

19:53 <CoderForLife> yup - found that

19:54 <RomW> Needless to say, getting BOINC and Virtualbox to play well with each other has been interesting

19:54 <CoderForLife> heh

19:54 <CoderForLife> did you run out of four letter words?

19:54 <RomW> lol

19:54 <RomW> nope

19:55 <CoderForLife> heh

19:55 <RomW> It is actually going quite well

19:55 <CoderForLife> "well" - that's a little-used four letter word =)

19:56 <RomW> There were a few bugs that needed to be fixed in VirtualBox which Oracle took care of as soon as we said it would keep us from deploying

19:56 <RomW> lol

19:57 <RomW> All I can really say is programming COM-based apps on Mac and Linux are a pain in the butt

19:58 <RomW> Had to switch over to using the CLI version of the API set

19:58 <RomW> since I only figured out how to easily do things on Windows

19:59 <RomW> anyway, things are moving right along now...

19:59 <CoderForLife> I love a good report

19:59 <Tank_Master> :)

20:00 <wdsmia> Lo Rom, Don, Jay oh and the other one TM :)

20:00 <CoderForLife> *lo Doug

20:00 <RomW> I suspect things are about to get real interesting in the BOINC universe

20:00 <Tank_Master> heya wdsmia

20:00 <RomW> Howdy wdsmia

20:00 <CoderForLife> now that would be a change >yawn<

20:01 <RomW> Well, as near as I can tell there are a huge number of grid applications that run on Linux, but the devs are not really interested in porting to Windows or Mac...

20:02 <RomW> If this works well, then the development effort in running there code is reduced.

20:02 <RomW> their code

20:02 <CoderForLife> sounds good

20:03 <Tank_Master> which is a plus

20:03 <RomW> yup

20:04 <RomW> Last number thrown out in one of the grid publications is something like 2000 apps on university grids

20:04 <wdsmia> didnt I hear this song before oh in 2004 lol

20:04 <RomW> maybe

20:04 <RomW> probably

20:04 <RomW> :)

20:04 <CoderForLife> heh

20:05 <wdsmia> then its about time :)

20:05 <RomW> Well the VM technology is far enough along it is much easier now

20:06 <RomW> debian automatically installed vbox's guest extensions during setup

20:06 <RomW> and the guest extensions supported shared folders

20:08 <RomW> I think the first time Ben, project manager at CERN, tried this he tried using a web server in the guest VM and python scripts on both side to handle file copy operations and moving data inside and outside the VM

20:08 <RomW> needless to say, there were many points of failure

20:14 <RomW> So what has been happening in this part of the net?

20:16 <CoderForLife> um..

20:17 <wdsmia> lots of weather

20:17 * CoderForLife looks at the ceiling

20:17 <CoderForLife> tomorrow is my birthday

20:17 <RomW> congats

20:17 <RomW> Happy Birthday

20:17 <CoderForLife> it's nice to be here

20:19 <CoderForLife> I officially weigh 128 pounds less this morning than I did in May 2009

20:19 <CoderForLife> so - nice to be here

20:20 <RomW> Wow, that takes dedication...

20:20 <CoderForLife> well, it beats the alternative I was facing

20:21 <RomW> Good deal all the way around then...

20:21 <CoderForLife> yup - thanks

20:23 <PovAddict> seen on the internets: "So, I think I'm going to humiliate myself into losing weight. I'm getting a Wifi scale and some other health hardware, and I'm going to have it automatically post everything to the interwebs. If I don't stick to my plan, it'll mock me on Facebook and all of you will get to laugh at me, thus giving me an incentive to not be a lazy ass. This sounds like a plan."

20:24 <RomW> CFL, Have you done much work around validating API contract consistancy?

20:24 <RomW> Howdy Pov

20:25 <CoderForLife> I was more afraid of the Grim Reaper than any human pressure

20:25 <CoderForLife> nothing beyond web services/XML schema type stuff, and some automated test tools for regression testing

20:26 * PovAddict is trying TDD

20:26 <CoderForLife> bravo Pov

20:26 <RomW> TDD?

20:26 <CoderForLife> go ahead Pov\

20:26 <PovAddict> test-driven development

20:27 <CoderForLife> bingo

20:27 <Tank_Master> im just wrighting a bit of code for one of the second life alternitive viewers

20:27 <PovAddict> you first write a test for the behavior / new feature / bug fix you want

20:27 <RomW> I haven't heard of that before

20:27 <PovAddict> which will fail

20:27 <RomW> ah

20:27 <PovAddict> then you write code to make the test pass

20:27 <CoderForLife> one of the agile principles - if testing is good, then test from the start

20:28 <PovAddict> and check if that test passes (code change worked) and all the other tests still pass (no regressions introduced)

20:28 <RomW> hmmmm, i had read about a variation on that, i suspect it might have been touched in some agile stuff I read

20:28 * CoderForLife notes that all intellectual conversations eventually become Coding discussions

20:29 <RomW> lol

20:29 <PovAddict> in theory, if you stick to that, eg. if always write a (failing) test before adding any new feature, you'll end up with all such new features having good test coverage

20:29 <CoderForLife> and rightly so, I might add

20:29 <PovAddict> which is much easier than trying to add tests to existing code

20:30 <RomW> How does TDD deal with random error states such as network transfers and the like?

20:30 <PovAddict> well, TDD also "forces" you to architect the code so that it's testable, for example making it easy to mock actual network access

20:30 <RomW> ah

20:32 <CoderForLife> brb  - Mother Nature demands my attention

20:32 <Tank_Master> dont forget to send your donations

20:32 <PovAddict> eg. you don't test if all workunit input files are downloaded (by doing real http downloads), you test if the relevant code does all the "download file" calls correctly, probably by making it use a different httpdownloader object that records all such calls and returns dummy data instead of doing any real network access

20:34 <RomW> yeah, I started (meaning I put in a skeleton library) creating some code to simulate a core client to test GUI RPCs for something else I am working on...

20:34 <RomW> I haven't gotten very far into it though

20:34 <PovAddict> another interesting technique I have seen is mutation testing (though I didn't see any good implementation for C++...)

20:35 <RomW> mutation testing?

20:35 <PovAddict> basically, a tool goes through your code and introduces bugs into it, by deleting random lines, inverting conditionals, etc

20:35 <PovAddict> then you run your unit tests and watch in horror as they *still pass*, highlighting that your tests need more work :)

20:36 <RomW> lol

20:37 <CoderForLife> she told me she wants to see you soon, Tank_Master

20:42 <CoderForLife> well - about time for me to do some day-ending chores

20:42 <CoderForLife> gn all

20:42 <RomW> gn

20:42 <CoderForLife> thanks for the T4T link Rom

20:42 <RomW> yw

20:42 <RomW> That project is going to keep us on our toes for a few months

20:46 <PovAddict> do apps use a complex setup of libraries, filesystem hierarchy for data, multiple processes, etc?

20:47 <RomW> which apps?

20:47 <PovAddict> in the CERN grid

20:47 <RomW> Honestly I don't know

20:48 <RomW> I know that a good chunk of their libraries have to be validated before general use

20:48 <PovAddict> I think that's the main huge advantage of using a VM

20:48 <RomW> Well their VM has something of a virtual filesystem

20:49 <PovAddict> if it was a self-contained binary with no multiprocessing or anything weird, it wouldn't be too hard to port IMHO

20:50 <RomW> they have some Linux kernel extension that translates placeholder links into http requests to a public facing webserver to actually download the file on first use

20:50 <PovAddict> oh god, kernel extension

20:50 <RomW> so far the experience hasn't been bad

20:51 <RomW> but that might be related to the fact that the Linux VM is locked down and so you only ever see the agent's status messages

20:51 <PovAddict> yeah, if you use a VM there's no problem

20:52 <PovAddict> the app could be using an extremely complex system with multiple processes and users and kernel extensions and tricky IPC

20:52 <PovAddict> and from BOINC you still see it as a single process (VirtualBox.exe)

20:52 <RomW> yup

20:52 <PovAddict> my question is if the cern apps really are that insane :)

20:53 <RomW> And VirtualBox does a better job at CPU/Network throttling...

20:56 <RomW> I don't know if it is a matter of them being insane, as everybody using the same tools so everybody will get consistant results

20:56 <RomW> imagine if everyone used a different FFT library

20:57 <RomW> if one of them had a bug that made one group believe they cracked something important when all they did was stumble across a bug...

20:57 <PovAddict> or worse, viceversa...

20:57 <RomW> true

20:58 <RomW> this way, the domain experts validate the libraries everybody uses

20:59 <RomW> and they don't run into executables not being able to be shared across the whole org

21:01 <Tank_Master> i would like to see the ability to run more than one T4T app at once, or for it to be multi threaded

21:01 <Tank_Master> otherwise, i havent have any issues with running T4T

21:02 <Tank_Master> i do think it might be useful if a project itself could tell the client to detach and reattach to the project, either via the same URL or a differnt one

21:03 <RomW> We might have some of those issues resolved when Daniel gets back.

21:05 <RomW> Wouldn't a reset be better?

21:20 <RomW> Well, I'll catch you all around... gn

21:20 *** RomW has quit IRC

21:32 <desti> http://www.youtube.com/watch?v=Phr7Za10Gpc

21:32 <Romulus> Title: 津波で青森のおじさんが・・・ 【tsunami hit Oirase City in Japan】 - YouTube (at www.youtube.com)

21:36 *** efc has joined #boinc

21:38 *** licensed has quit IRC

22:16 *** zombie67 has quit IRC

23:08 *** efc has quit IRC

23:08 *** efc has joined #boinc

23:37 *** zombie67 has joined #boinc

23:48 *** zombie67 has quit IRC

Generated by irclog2html.py 2.4 by Marius Gedminas - find it at mg.pov.lt!