It's unbelievable that we are being forced to switch over to Boinc before it works properly.

I thought the cache of work units was a nice feature.
Guess what it don't work right. It does not estimate
the number of needed workunits for 2 of the 3 machines I put boinc on properly... requiring special hand tweeking

The windows graphics are useless ...
LINUX graphics don't exist.
The connect to seti button in Linux doesn't work.
The view pending credits on the web interface page doesn't work

My Pentium III running windows 95 is still trying to upload all completed work.

The list goes on & on

How about fixing this "CRAP" be for scraping classic?

---

My Pentium III running windows 95 is still trying to upload all completed work.

Everyone is having this problem. The servers were down for an extended time, it takes time to recoop from that. Everyone is pushing, pushing, pushing, and it can only allow so many connections simutaneously before it rejects some.

I have no issues with the program. It works fine. I have a nice cache, I never liked having the graphics on in either, because all it does is slow down processing time.

The view pending credits on the web interface page doesn't work

This has been answered time and time again. It probably will NEVER work. It is a resource hog on the database. You think there is problems now, turn that back on and crash, crash, crash. I think the link should be removed, but I think they are trying to program a work around, and it may show a snapshot of pending credits, but that snapshot will never show the current state, so everyone will complain about that, saying it is useless. So just think of it as not there.

---

My movie https://vimeo.com/manage/videos/502242

Looks like you have a problem


Name 01no03aa.12221.26130.236092.68_1
Workunit 20290250
Created 10 Jul 2005 19:58:05 UTC
Sent 11 Jul 2005 19:27:31 UTC
Received 12 Jul 2005 21:08:00 UTC
Server state Over
Outcome Client error
Client state Computing
Exit status -1073741819 (0xc0000005)
Computer ID 1166227
Report deadline 25 Jul 2005 19:27:31 UTC
CPU time 77727.787
stderr out 4.45
- exit code -1073741819 (0xc0000005)



***UNHANDLED EXCEPTION****
Reason: Access Violation (0xc0000005) at address 0x7E957B7E read attempt to address 0xF6C8F59C

0: Stackwalker not initialized (or was not able to initialize)!




Validate state Invalid
Claimed credit 16.0548751845016



From the WIKI

***UNHANDLED EXCEPTION**** Reason: '(reason)'
From BOINCWiki
Table of contents [showhide]
1 General

2 Version Information

3 Example Log(s)

3.1 UNHANDLED EXCEPTION - Access Violation - Write Attempt
3.2 UNHANDLED EXCEPTION - Access Violation - Read Attempt


4 Other Related Messages

[edit]General
Message Type: Error Message

Actually, we geeks call this a "Fatal Error", this means that the whole world comes to an end …

This error can be the result of a programming error in the code for:

The BOINC Client Software
The Science Applications
Any other program running on the computer
The Operating System
On the other hand, it could be because it is Tuesday.

Seriously, this is a serious error within the running application program and it is one that computer hardware and Operating System manage when the program attempts to do something that is a no-no

---

Semper Eadem
So long Paul, it has been a hell of a ride.

Park your ego's, fire up the computers, Science YES, Credits No.

Looks like you have a problem

I HAVE A PROBLEM ??? No like I said this boinc crap is not ready for prime time.
Boinc has several issues and this is just one of them.
This same computer is running clasic at the same time with NO PROBLEM.
Don't even try to say that's causing the problem. Two other machines are
also running boinc & clasic at the same time with NO PROBLEMS.
If the old stuff works and the new stuff don't Why do I have a problem ??
boinc is at best beta and should not be forced on the world yet.

Name 01no03aa.12221.26130.236092.68_1
Workunit 20290250
Created 10 Jul 2005 19:58:05 UTC
Sent 11 Jul 2005 19:27:31 UTC
Received 12 Jul 2005 21:08:00 UTC
Server state Over
Outcome Client error
Client state Computing
Exit status -1073741819 (0xc0000005)
Computer ID 1166227
Report deadline 25 Jul 2005 19:27:31 UTC
CPU time 77727.787
stderr out 4.45
- exit code -1073741819 (0xc0000005)



***UNHANDLED EXCEPTION****
Reason: Access Violation (0xc0000005) at address 0x7E957B7E read attempt to address 0xF6C8F59C

0: Stackwalker not initialized (or was not able to initialize)!




Validate state Invalid
Claimed credit 16.0548751845016



From the WIKI

***UNHANDLED EXCEPTION**** Reason: '(reason)'
From BOINCWiki
Table of contents [showhide]
1 General

2 Version Information

3 Example Log(s)

3.1 UNHANDLED EXCEPTION - Access Violation - Write Attempt
3.2 UNHANDLED EXCEPTION - Access Violation - Read Attempt


4 Other Related Messages

[edit]General
Message Type: Error Message

Actually, we geeks call this a "Fatal Error", this means that the whole world comes to an end …

This error can be the result of a programming error in the code for:

The BOINC Client Software
The Science Applications
Any other program running on the computer
The Operating System
On the other hand, it could be because it is Tuesday.

Seriously, this is a serious error within the running application program and it is one that computer hardware and Operating System manage when the program attempts to do something that is a no-no

Gee The only NEW application running is boinc ... DUA I wonder if that might be the problem

---

Seeing that you are so open minded, all I can say at this point is, my 4 boxes run find with Boinc, so you make the judgment, have a nice day, whatever.....

---

Semper Eadem
So long Paul, it has been a hell of a ride.

Park your ego's, fire up the computers, Science YES, Credits No.

Two other machines are also running boinc & clasic at the same time with NO PROBLEMS.

That would indicate a hardware issue with this particular machine then... Why are you blaming BOINC when you say the other two machine have "NO PROBLEMS"?

>I HAVE A PROBLEM ??? No like I said this boinc crap is not ready for prime >time.
>Boinc has several issues and this is just one of them.
>This same computer is running clasic at the same time with NO PROBLEM.
>Don't even try to say that's causing the problem. Two other machines are
>also running boinc & clasic at the same time with NO PROBLEMS.
>If the old stuff works and the new stuff don't Why do I have a problem ??
>boinc is at best beta and should not be forced on the world yet.

No, you still have the problem. The results your turning in do not match the quorum of results turned in. Something on your machine is causing it to return imperfect results.

The difference is now with BOINC's validating system, you only get credit if your results are actually correct.

---

BOINC WIKI
Overclockers.com's Forum

Two other machines are also running boinc & clasic at the same time with NO PROBLEMS.

That would indicate a hardware issue with this particular machine then... Why are you blaming BOINC when you say the other two machine have "NO PROBLEMS"?

Why is it a hardware issue with my machine when it's boinc that will not run???
Clasic had no problem with this hardware.
The other machines are different processors & different operating systems.

The boinc software will apparently not run on all windows systems.
Seems to me the software has not been completely tested as was my original statement.

---

>I HAVE A PROBLEM ??? No like I said this boinc crap is not ready for prime >time.
>Boinc has several issues and this is just one of them.
>This same computer is running clasic at the same time with NO PROBLEM.
>Don't even try to say that's causing the problem. Two other machines are
>also running boinc & clasic at the same time with NO PROBLEMS.
>If the old stuff works and the new stuff don't Why do I have a problem ??
>boinc is at best beta and should not be forced on the world yet.

No, you still have the problem.

Not really I just won't contribute any more & cut down my power bill

The results your turning in do not match the quorum of results turned in. Something on your machine is causing it to return imperfect results.

Well it worked fine with classic... It's the boinc software that won't run
on this pentium III with windows 95. The boinc software is trying to do
something win95 apparently isn't able to do.

The difference is now with BOINC's validating system, you only get credit if your results are actually correct.

Validating the results has nothing to do with the software not producing proper results.

---

Validating the results has nothing to do with the software not producing proper results.

As Classic didn't validate, nobody can say for sure, whether proper results were sent or just crap. Classic excepted everything, that was one of it's serious flaws.
And one of Boincs advantages is, that such things as hardware failures will show immediately (OK, not at instant, but a reasonably short time after crunching).

---

Gruesse vom Saenger

For questions about Boinc look in the BOINC-Wiki

Hear hear!

BOINC is not yet fit for purpose, and the project team that just ignore all the valid and pertinent comments (which lead to criticism if only because of pretending that all is wonderful) do not help.

There is only one good reason for BOINC and that is to be able to run other projects with the same core management function. Unfortunately that management system has not been developed from what has been achieved (over time yes) with Classic and therefore in many respects it is a massive backward step.

Things that are supposed to be an improved version should be better than the forerunner.

The extent of manual intervention necessary to replace the automatic classic functionality is awful. Look at how many downloaded units that continue to process after the deadline (and often don't start until afterwards). The design process took too many short cuts and ignored the wider issues. The excuse that classic didn't produce valid results, or processed the wrong things is somehow blamed on us as users! We can only process what we are given with the tools we are given. Most of the problems stem from the architecture in the centre and not the client end.

The distributed processing concept is fine, it just hasn't been developed (even after all this time) to be a robust and useful application in this instance. Why provide a second rate replacement for something that worked and extremely well at that?

---

The extent of manual intervention necessary to replace the automatic classic functionality is awful.

So in Classic everything was automatic? No need for any add-ons, as caching and a GUI was build in it?

Look at how many downloaded units that continue to process after the deadline (and often don't start until afterwards). The design process took too many short cuts and ignored the wider issues. The excuse that classic didn't produce valid results, or processed the wrong things is somehow blamed on us as users! We can only process what we are given with the tools we are given. Most of the problems stem from the architecture in the centre and not the client end.

If I remember one of your former rantings correct, you were running Classic and Boinc on the same machine amd were complaining, that Boinc doesn't keep it's deadlines. You were told often enough, that that's impossible, because they cannot run at the same time smooth., but as usual, you won't listen, just rant.

The distributed processing concept is fine, it just hasn't been developed (even after all this time) to be a robust and useful application in this instance. Why provide a second rate replacement for something that worked and extremely well at that?

It's a first rate replacement for a former first rate, but now outdated system.

---

Gruesse vom Saenger

For questions about Boinc look in the BOINC-Wiki

Validating the results has nothing to do with the software not producing proper results.

As Classic didn't validate, nobody can say for sure, whether proper results were sent or just crap. Classic excepted everything, that was one of it's serious flaws.
And one of Boincs advantages is, that such things as hardware failures will show immediately (OK, not at instant, but a reasonably short time after crunching).

Where did you get this cock-a-mamey idea?? If they could varify that the results
were correct Why would people need to run the calculations in the firstplace.
If boinc knew the answers why are we doing this?

And what does hardware have to to do with the cpu running mathmatical calculations?

It's clear you have no grasp of what is going on.

---

The extent of manual intervention necessary to replace the automatic classic functionality is awful.

So in Classic everything was automatic? No need for any add-ons, as caching and a GUI was build in it?

Look at how many downloaded units that continue to process after the deadline (and often don't start until afterwards). The design process took too many short cuts and ignored the wider issues. The excuse that classic didn't produce valid results, or processed the wrong things is somehow blamed on us as users! We can only process what we are given with the tools we are given. Most of the problems stem from the architecture in the centre and not the client end.

If I remember one of your former rantings correct, you were running Classic and Boinc on the same machine amd were complaining, that Boinc doesn't keep it's deadlines. You were told often enough, that that's impossible, because they cannot run at the same time smooth., but as usual, you won't listen, just rant.

You don't rememmber right. I am the one running both not the person you are quoting. Wher was anyone told at any time this was impossible . It clearly is possible. If boing has a problem working properly with other concurrent processes the it has really major bug.

Must be the german ??

Why not trool someware else

The distributed processing concept is fine, it just hasn't been developed (even after all this time) to be a robust and useful application in this instance. Why provide a second rate replacement for something that worked and extremely well at that?

It's a first rate replacement for a former first rate, but now outdated system. [/quote]

---

Where did you get this cock-a-mamey idea?? If they could varify that the results
were correct Why would people need to run the calculations in the firstplace.
If boinc knew the answers why are we doing this?

Validation can be done in several ways.
1. Compare with truth/reality
2. Compare with theory
3. Compare with other results, done in a same (but not the same) manner

#3 is what applies to Seti (and most other Boinc projects).

And what does hardware have to to do with the cpu running mathmatical calculations?

Have you ever heard of rounding errors?
have you ever heard, that different manufacturers handle this issue different in their hardware?
If not, take a first look here in the Wiki about validation and here about floating point errors.

BTW: What's the value of Pi, that has to be used for correct results?

It's clear you have no grasp of what is going on.

Those, who sit in glasshouses...

---

Gruesse vom Saenger

For questions about Boinc look in the BOINC-Wiki

You don't rememmber right. I am the one running both not the person you are quoting. Wher was anyone told at any time this was impossible . It clearly is possible. If boing has a problem working properly with other concurrent processes the it has really major bug.

One of the most important features (not in a Bill Gates sense of this word) is the use of formerly unused CPU-cycles. So, if proper designed, anything else on a puter, that's running Boinc, is of a higher priority than Boinc and gets served first.

And on my puter it works just this way. It keeps the CPU usage on a constant 100% level, but doesn't interfere with the real "work" I'm doing. And if this other stuff is heavy on the CPU as well, Boinc will stop crunching at all. No manual fiddeling (although I sometimes do it for fun and testing purposes), no disturbtion, just fine.

---

Gruesse vom Saenger

For questions about Boinc look in the BOINC-Wiki

Where did you get this cock-a-mamey idea?? If they could varify that the results
were correct Why would people need to run the calculations in the firstplace.
If boinc knew the answers why are we doing this?

Validation can be done in several ways.
1. Compare with truth/reality
2. Compare with theory
3. Compare with other results, done in a same (but not the same) manner

#3 is what applies to Seti (and most other Boinc projects).

And just how would they know which result was the right one?

And what does hardware have to to do with the cpu running mathmatical calculations?

Have you ever heard of rounding errors?
have you ever heard, that different manufacturers handle this issue different in their hardware?
If not, take a first look here in the Wiki about validation and here about floating point errors.

So are you trying to say boinc can't do the math properly with different cpu's??
Sure sounds like a bionc client problem to me

Go Troll somewhere else

---

So are you trying to say boinc can't do the math properly with different cpu's??
Sure sounds like a bionc client problem to me

Nobody can do proper math with a computer, it's all just iterations, rounding errors...

Go Troll somewhere else

Ditto

---

Gruesse vom Saenger

For questions about Boinc look in the BOINC-Wiki

You don't rememmber right. I am the one running both not the person you are quoting. Wher was anyone told at any time this was impossible . It clearly is possible. If boing has a problem working properly with other concurrent processes the it has really major bug.

One of the most important features (not in a Bill Gates sense of this word) is the use of formerly unused CPU-cycles. So, if proper designed, anything else on a puter, that's running Boinc, is of a higher priority than Boinc and gets served first.

And on my puter it works just this way. It keeps the CPU usage on a constant 100% level, but doesn't interfere with the real "work" I'm doing. And if this other stuff is heavy on the CPU as well, Boinc will stop crunching at all. No manual fiddeling (although I sometimes do it for fun and testing purposes), no disturbtion, just fine.

Gee nice dodge, Some good theory.
The more you say the more you make my point
BOINC HAS Problems

Have a nice day TROLL

---

And just how would they know which result was the right one?

They don't. All they are looking at is whether several machines turn in similar results. Thus considering the science 'valid'.