LHC@home

Hi Ben and Chrulle!

Sorry about posting OT questions here. But I
really would like to know when Malaria Control goes live, or at least account
creation is enabled.

As you seem to be involved with the project, maybe you
can give a rough timetable or point to the info
on the Malaria Control website.

Thanks in advance,

Michael

---

Team Linux Users Everywhere

Beta is probably coming soon. (weeks)
As soon as account creation is opened we will announce it here.

---

Chrulle
Research Assistant & Ex-LHC@home developer
Niels Bohr Institute

Beta is probably coming soon. (weeks)

OK, I can wait.

As soon as account creation is opened we will announce it here.

I thought so.

Thanks,

Michael

---

Team Linux Users Everywhere

Talking of Malaria control, I'm impressed that it got to be a CERN thing at all.

After all, mosquitos and parasites don't fit into CERNs traditional fields of study, the RN or CERN; and Africa is outside the E of CERN. (I'm not complaining, I think it is great).

I wonder if anyone can fill me in on the history of how it got to be a CERN project?

River~~

---

I wonder if anyone can fill me in on the history of how it got to be a CERN project?

River~~

Did you Click here for more information?

Michael

I wonder if anyone can fill me in on the history of how it got to be a CERN project?

River~~

Did you Click here for more information?

Michael

yes I did. I repeat it is a great project and it is great that it is being supported at CERN, but I could not see anywhere on that site how the wonderful idea came about that a nuclear research organisation (Recherche Nuclaire) for Europe (Europeanne) is doing a biomedical project for Africa.

I am asking, not how the project came about or how it is needed (which are answered on their site) but how it came about that CERN is supporting it.

R~~

There's another BOINC project where CERN is probably involved, to my knowledge they do it together with institutes in Portugal - anything new about HEP@Home?

Must be quite close related to LHC I guess, so could it be that we will crunch those WUs when LHC pauses?

There's another BOINC project where CERN is probably involved, to my knowledge they do it together with institutes in Portugal - anything new about HEP@Home?

Nothing more than a paper :(

Michael

---

Team Linux Users Everywhere

Heres an unofficial update on MalariaControl.

The Project Developers and us Alpha Testers and still trying to find the last few bugs (not Mosquitoes) in the application. There have been some problems with checkpointing, particuarly when your client runs the benchmarks or exits. This will cause the result to have invalid results, of course this is not known until the result is validated and the participant ends up getting ZERO credit. They have been making good progress the last week with fixes, if you follow the main page or the message board you will see that we are now getting some good results. There may still be one more bug to find. I'm an alpha tester and I'm guessing that once this problem is solved, then the project would be opened up for beta (I can't promise that, but I don't see any other problems we havn't solved yet). I can't tell you when (As I don't know) but we are a lot closer now than we have been in the last four weeks. Stay tuned for more info.

There's another BOINC project where CERN is probably involved, to my knowledge they do it together with institutes in Portugal - anything new about HEP@Home?

Nothing more than a paper :(

Michael

It has been more than paper in 2004 already. I've got a test project tgz archive here from 10/2004 with sources, scripts, Atlas-stuff and ELF binaries called kvmuonsomething, a pile of things I don't know anything about so I cannot tell in which state it has been back then. The test server used to be *.gridpt.fe.up.pt but it's gone now.

I wonder if anyone can fill me in on the history of how it got to be a CERN project?

River~~

Did you Click here for more information?

Michael

yes I did. I repeat it is a great project and it is great that it is being supported at CERN, but I could not see anywhere on that site how the wonderful idea came about that a nuclear research organisation (Recherche Nuclaire) for Europe (Europeanne) is doing a biomedical project for Africa.

I am asking, not how the project came about or how it is needed (which are answered on their site) but how it came about that CERN is supporting it.

R~~

It came about because CERN although a particle physics laboratory also research computer science and a mandate to transfer this research to the public. The http protocol for example was created at CERN. As other examples of research not directly linked to particle physics CERN is also involved in medical research.
See: here

Because we already had a lot of expertise in boinc projects here at CERN it was an ideal place to start and spin-off another project. Although CERN is not directly contributing money to the project, the management has graciously allowed us to work on the project as long as it does not interfere with our other duties.

---

Chrulle
Research Assistant & Ex-LHC@home developer
Niels Bohr Institute

... the management has graciously allowed us to work on the project as long as it does not interfere with our other duties.

and, it seems from the url, allowing cern computers time and space to host the servers.

I'm impressed at their flexibility for a good cause :)

R~~

if you have an RSS reader then add it to your reader that way at least if they announce new account creation is open you will get the message to let you know

---

Join us in Chat (see the forum) Click the Sig


Join UBT

and, it seems from the url, allowing cern computers time and space to host the servers.

I may be wrong, but I think the production servers may be hosted elsewhere, with only the alpha setup on the CERN equipment while undergoing alpha testing.

There was mention of new servers and a new url for the beta, IIRC.

But it's definitely cool of the CERN guys to help get it started. They deserve a big pat on the back. ;->

Hi Chrulle, Krunchin-Keith,

I'am reading that there are problems validating Linux against
Windows results.

Quote from here:

I've noticed the large proportion of invalid results with 519, all except one from linux hosts (the other one is Keith's win 98 box). As Chris points out, this almost looks like a linux-windows conflict, but not an easily reproducible one, as some of the linux results come back valid.

If you need another Linux host, just drop me an email.

Michael

---

Team Linux Users Everywhere

Hello!
Why only a linux client, I also have different windows clients waiting with Amd cpu'st ready to help!
See You soon!

---

Hans Sveen
Oslo, Norway

Hello!
Why only a linux client, I also have different windows clients waiting with Amd cpu'st ready to help!
See You soon!

This is only my guess, based on other projects.

Usual procedure is for the early dev work to be done on a single platform, then branch out to other platforms when the code is working on one.

The reason to confine early prototypes to a single platform are to get rid of the more obvious bugs in a single environment.

Which platform is chosen for the prototype depends partly on which platform the devs are most happy with, and my also depend on cost of licences for dev tool packages. Either way the choice could often be Linux: many programming types prefer Linux, and the software is the ultimate in cheap.

It is absolutely certain that a windows version will be along at some stage - no DC project is going to turn away the biggest cadre of computers in the world! In the meantime, the beauty of BOINC is those winboxes can be doing some other project a big favour, even if their primary mnission is for Malaria Control.

River~~

---

It is absolutely certain that a windows version will be along at some stage...

If you were referring to MC specifically, there already is one.

Your synopsis regarding the logical reason to choose one particular platform for development is apt, however not expanding the dev pool during alpha or beta phase can cause problems farther down the road. This is pretty much what we are seeing at MC at the moment.

Thankfully the devs are working on both the linux & win platforms simultaneously, so we have found the incompatibility pretty early. Had the devs chosen to go with a single platform, we wouldn't have identified the current issue of inter-platform incompatibility. If the problem was only tagged when adding a new platform into the production environment, many users would be crying foul over 0 credits...and the devs would have more headaches "fixing" credit and trying to salvage project credibility.

IMHO, it's unlikely that MC will be able to test every possible OS before going live, however catering for the majority is a reasonable 2nd place finish. Why test them all? Well, even something as simple as a single math function call into a shared library can return a slightly different result between two *nix OS's, linux & solaris for eg.

It is absolutely certain that a windows version will be along at some stage...

If you were referring to MC specifically, there already is one.

OK, noted thanks

Your synopsis regarding the logical reason to choose one particular platform for development is apt, however not expanding the dev pool during alpha or beta phase can cause problems farther down the road.

My impression is that most BOINC projects start alpha with just one platform, and go to multi platform around the end of alpha or start of beta.

I'd agree that by the time a project leaves beta it should have working apps on the popular platforms - in my opinion that is at least Win, Lin and OS X, the variation is whether a dev team chooses to put the multi platform debugging into the alpha or beta phase, and each choice has its advantages.

R~~

Beta is probably coming soon. (weeks)
As soon as account creation is opened we will announce it here.

MalariaControl.net moves out of Alpha and into Beta. The site has moved to a new web address of http://www.malariacontrol.net/ and is hosted by the Centre Universitaire d'Informatique (CUI) and the Division Informatique (DINF) at University of Geneva. Account creation will be enabled later this week or early next week. A limit of 500 user accounts will be set to start while they test the server backend under this load and test Validation of results. Hope to see you there.

---

BOINCin since 10-Apr-2004 ~~~ Alpha and Beta Tester ~~~ Join team USA