The FreeHAL stats for the countries are obviously wrong, as I am in Canada and have some credits. The stats show that everyone is in the "other" country. http://boincstats.com/stats/project_graph.php?pr=freehal&view=countries

I have been crunching a little FreeHal on my boat anchor [barely worth using for BOINC] and FreeHal is a project with a multitude of problems.

. . . the WU's start out at 100%, they don't use clock time and then they finish no telling when.

Incorrect countries is the least of their problems.

Glad I didn't dial in a better computer.

I also take it their credit system is a little on the messed up side of things. I've started crunching for FreeHAL and have now racked up over 50,000 credits in the space of five days!?!??! Alot of WU's that take about 15 minutes are awarded 250 credits!

It's took me over a year to crunch only half of this (25k) for QMC (Along with other projects running)! Nice reward but definately making a general mess of things.

Edit: Just checking and it seems like it's only Cred hording on one of my hosts?!?!?!?! Mmmm..?

I also noticed a weird spike (6,500 more than usual) in my credits and have attributed it to FreeHAL on one host also. Very strange : http://boincstats.com/stats/user_graph.php?pr=freehal&id=704

That's because Freehal sometimes grants way too much credit.

Hello,

In my opinion, there is not only their credit system which is messed up... I thought it could be an original project, because I also tested their IA application for human dialog.

I'm often volunteer for new project which seem promising, but... I also don't want that alpha projects come into our computers without being fully tested. Alpha status doesn't mean: unleash stupid coding into the BOINC system. Coding errors could always be acceptable if the main part of the code respect the Boinc standards.

Now that I've been credited with 150000 credits for Freehal in one day;
Now that their Linux apps have almost destroy the filesystem of my very stable linux box (by creating 200 slots of working WU where there was only 2 CPU cores);
Now that it appears (on my different home computers) that their app is able to crunch in a strange mode (CPU non intensive mode???), even when the project has been told to not download work!!!

I'm willing to think that this project is a bad joke, or a kind of virus, in its way to propagate and using ressources without permission.

I'm a kind of "new" cruncher to BOINC, but... These days, I got the sad feeling that more and more unprepared projects are coming out to "infect" computers and stats. I understand that alpha and beta projects should be joined carefully, but except in these dedicated forums, nothing seems to be made to warn newbies or new volunteers.

I'll be glad that the Boinc's masters could do something to detect and automaticaly bannish stupid projects that overrule the policies of the scheduler or of the user's preferences. But it seems that there are still some glitches when "playing" with multi-cores or multi-cpu systems.

I'll be glad too that an harmonized credit system could take place, which really measure CPU time, in direct correlation with the electrical power comsumption. Some projects are giving credits only when a WU is correct at their side!!! When the CPU has run for "a few days" and that unfortunately lead to a computation error, it still has an electrical cost, so that should lead to a credit reward. At the opposite, some stupid coders give more than 500 credits for a strange WU, which doesn't seem to compute (no percentage done) and finish in a minute (on a Phenom).

If others users have more information about the reality behind or inside Freehal (reverse engineering? code exploration?), it's welcome.

I can see the logic behind not granting credits for results that error out. No useful contribution to the science -> no credit. Pretty logical in my opinion.

But that Freehal is screwed is a fact.

I see what you mean

Hello,

Mmh... I understand your point of view, but I can not join in the particular context of distributed computing.

What is the purpose of distributed computing? That many many many generous computer owners lent their CPU to some weirdo science projects, for which they have been told that it's good for the futur of mankind, etc. Without any offense or irony, if I were one of those newbie, who doesn't always have scientific background, it would be equivalent to donate for lost pets! What would be rewarded then? My noble and generous gesture? Or my comprehension of why people abandon pets? Definitively my noble gesture, I guess.

All these generous volunteers do not necessarily understand what their computer calculates! They therefore do not necessarily respond in a "scientific" manner to the fact that a poorly designed project will spoil the CPU time they give, and above all, their electricity.

Even you, us, everybody, we don't know for sure what happend in our CPU once we attached to a project. You can't even determine if your computer really computes what is presented on the project's site! But you gave your CPU time, so be rewarded for this!!!

Now, let's be more pragmatic, let's say that I agree with you, that we are fully and only focused here on one purpose: the science result. Being careful to not slip into scientistic...

How do you will react when your CPU has spent one week 24/24 to compute a great WU, full of promising science, and because of a bad coding, not because the science experiment is in a dead end, it finishes into:

• Error: Maximum CPU time exceeded
  
• Error: too many normally harmless exit(s)
  
• Error: Incorrect function. (0x1) - exit code 1 (0x1)
  

Now, let's go back to the most probable reality of distributed computing: I'm an innocent newbie that just want to participate for the... fun! So, me and my CPU and my computer are not responsible of these errors, and I have not the skill to understand what is behind. I just share my CPU time and my electrical power for (a kind of) science!!!

Could you explain where is the "science result" in these errors? And fortunately, these errors were harmless to the computer, it's not always the case (see the beginning of this post!).

If at least I'm not getting credited for my CPU time and my electrical comsumption, I should be rewarded for my patience!!!

Regards, and sorry for my english, not my native language, but I'm trying, I'm trying!!!

To be serious, I'm thinking about removing Freehal from the combined statistics. The granted credit stands in no relation ship with the delivered effort. It makes the combined statistics unrealistic, unfair and meaningless. I compare this to (not) using drugs/hormones in sports which I think nobody thinks is fair.

I tried to talk about this with Tobias, but different from when he wanted BOINCstats to show Freehal credit, he doesn't respond now.

I have no objection, at least until FreeHAL gets its act together. I don't appreciate it "hijacking" my computers, nor do I appreciate it inflating people's stats. Right now, FreeHAL is my second most important project in terms of credit, and I have only been running it for 2 months, compared to the other projects which I have been running for at least 6 times longer.

http://web.freehal.org/topic/credits

Hello Willy, BOINCstats users,

You do not need to remove FreeHAL from (combined) statictics. It will not happen again that someone gets 15 thousend of credits during two days.

Sorry Willy, I have not read your PMs because I thought I would get a mail when a private message recieves.

Tobias Schulz

you should have added some ZEROs to that number - afaik there were some users which ranted over a million a day..