They should block all the computers having BOINC older than 6.something. Enough said. I know it won't fix this problem but some other problems it will.

There is no reason to use BOINC 4 or BOINC 5 any longer.

HTH.

---

Manned mission to Mars in 2019 Petition <-- Sign this, please.

They should block all the computers having BOINC older than 6.something. Enough said. I know it won't fix this problem but some other problems it will.

There is no reason to use BOINC 4 or BOINC 5 any longer.

HTH.

This remind me of the joke about a drunkard who was looking for his lost keys not where he had lost them but under a lamppost because "there is more light".
Tullio

---

They should block all the computers having BOINC older than 6.something. Enough said. I know it won't fix this problem but some other problems it will.

There is no reason to use BOINC 4 or BOINC 5 any longer.

HTH.

This remind me of the joke about a drunkard who was looking for his lost keys not where he had lost them but under a lamppost because "there is more light".
Tullio

Hi Tullio and Henri, yeah I wander what problems Henri is referring to.
I (still) use 5.10.45 and 6.1.0.(Crunch3r) and don't 'see', whats wrong with them?
Anyway, the problems with optimized AP app. has nothing to do with the BOINC version used.
And I certainly don't think it's a good idea of blocking about 30% off the hosts, using an older version of BOINC.
This is something on which the project has to decide, which version is really outdated, IMHO. :)

---

They should block all the computers having BOINC older than 6.something. Enough said. I know it won't fix this problem but some other problems it will.

There is no reason to use BOINC 4 or BOINC 5 any longer.

HTH.

I don't know why this comment was made when the issues in this thread have to do with different AP versions and not different BOINC versions, but I still disagree with the suggestion. Just because some people cannot see any reason to use the older versions, doesn't in fact mean that there isn't a reason to use the older versions.

On top of that, how would this affect the majority of BOINC users who are set-it-and-forget-it types who don't even check the boards but are still producing valid science for the project?

I think this is just one of those necessarily "evils" that the minority who post here should accept and get used to.

On top of that, how would this affect the majority of BOINC users who are set-it-and-forget-it types who don't even check the boards but are still producing valid science for the project?

Some old BOINC 5 clients are NOT producing valid credits.

People using that old clients probably don't even know they are crunching. They are wasting their electricity, memory and HDD. They are wasting the credits of wingmen actually liking this project.

HTH.

---

Manned mission to Mars in 2019 Petition <-- Sign this, please.

Hi, Fred J. Verster!

Hi Tullio and Henri, yeah I wander what problems Henri is referring to.
I (still) use 5.10.45 and 6.1.0.(Crunch3r) and don't 'see', whats wrong with them?

5.10.45 is new enough, I think.

What is wrong with too old clients? They are wasting the credits of wingmen.

Why are people still using BOINC 4 or BOINC 5? What is wrong with BOINC 6? Nothing, I guess.

HTH.

---

Manned mission to Mars in 2019 Petition <-- Sign this, please.

Some old BOINC 5 clients are NOT producing valid credits.

People using that old clients probably don't even know they are crunching. They are wasting their electricity, memory and HDD. They are wasting the credits of wingmen actually liking this project.

HTH.

I am using BOINC 5.10.45 on SETI, Einstein, QMC, CPDN, CPDN Beta and LHC always obtaining valid results except in Astropulse. The problem does not reside in the BOINC client. In Astropulse I have been using an untested optimized Linux app based on 4.36 because the stock app was too slow. I have stopped all Astropulse work until I can get an optimized app based on 5.0.
Tullio

---

People, if you encounter problems with BOINC 6, report them here or somewhere, so David and Rom can fix the problems.

BOINC 6 system requirements for Windows machines:
Operating system: Windows 98 or later, Windows 2000 with Service Pack 4
Hardware: Pentium 233 MHz (Recommended: Pentium 500 MHz or greater)

I guess, no one of you are crunching with a computer with Windows 95 or CPUs like 66 MHz Pentium...

HTH.

---

Manned mission to Mars in 2019 Petition <-- Sign this, please.

People, if you encounter problems with BOINC 6, report them here or somewhere, so David and Rom can fix the problems.

BOINC 6 system requirements for Windows machines:
Operating system: Windows 98 or later, Windows 2000 with Service Pack 4
Hardware: Pentium 233 MHz (Recommended: Pentium 500 MHz or greater)

I guess, no one of you are crunching with a computer with Windows 95 or CPUs like 66 MHz Pentium...

HTH.

What about those of us who just don't like the way Boinc 6 sets up on the system and prefer the older structure? Honestly, there's nothing wrong with 5.10.45, 5.10.28, or 5.10.13 (all of which I use), so what's your problem with these versions that makes you want to just cut them off?

-Dave

---

On top of that, how would this affect the majority of BOINC users who are set-it-and-forget-it types who don't even check the boards but are still producing valid science for the project?

Some old BOINC 5 clients are NOT producing valid credits.

People using that old clients probably don't even know they are crunching. They are wasting their electricity, memory and HDD. They are wasting the credits of wingmen actually liking this project.

HTH.

Well that's very interesting considering BOINC doesn't produce results. BOINC simply launches the science app and manage CPU time, transfers work back to the project servers and all sorts of other administrative tasks, but it does not produce results.

As other people have stated to you already, there is nothing wrong with v5 clients. FYI - I'm using BOINC 6 on all my hosts.

As other people have stated to you already, there is nothing wrong with v5 clients. FYI - I'm using BOINC 6 on all my hosts.

http://setiathome.ssl.berkeley.edu/sah/forum_thread.php?id=46265&nowrap=true

HTH.

---

Manned mission to Mars in 2019 Petition <-- Sign this, please.

As other people have stated to you already, there is nothing wrong with v5 clients. FYI - I'm using BOINC 6 on all my hosts.

http://setiathome.ssl.berkeley.edu/sah/forum_thread.php?id=46265&nowrap=true

HTH.

Yeah, I was a part of that thread. I think you have misunderstood the part about claiming incorrect credit due to the change in the way credits are granted and the actual science that is returned.

Even if these old clients claim 0 credit, they are still returning valid work. This doesn't mean there's something "wrong" with them, only that every once in a while people miss a few credits. This isn't sufficient reason to cut them off, especially since many are the set-it-and-forget-it types, the project would be cutting off quite a few clients just to appease people who spend way too much time looking at their credits.

...and as far as I'm concerned, this is a non-discussion anyway. I personally asked Dr. Korpela myself back when that discussion was hot and he told me that the project is not willing to institute a minimum BOINC version at this time or the foreseeable future. Its just not going to happen.

People, if you encounter problems with BOINC 6, report them here or somewhere, so David and Rom can fix the problems.

BOINC 6 system requirements for Windows machines:
Operating system: Windows 98 or later, Windows 2000 with Service Pack 4
Hardware: Pentium 233 MHz (Recommended: Pentium 500 MHz or greater)

I guess, no one of you are crunching with a computer with Windows 95 or CPUs like 66 MHz Pentium...

HTH.

I have a Win95 system with 200MHz Pentium MMX, turning in valid results and correct <fpops_cumulative> based credit claims using BOINC 5.2.12. For compatibility I run the same version of BOINC on my Win98 and Win2k systems also. There are more active hosts running Windows NT 4.0 than Win95, they also cannot run BOINC 6.

                                                                Joe

What about those of us who just don't like the way Boinc 6 sets up on the system and prefer the older structure? Honestly, there's nothing wrong with 5.10.45, 5.10.28, or 5.10.13 (all of which I use), so what's your problem with these versions that makes you want to just cut them off?

-Dave

I agree. I only install 5.10.45 on PC's that I bring on. I like the service install where I can set to my network ID and do the work when the PC is not in use. I have tried a few PC's with 6.2.19, and I do not like the set up.

---

What about those of us who just don't like the way Boinc 6 sets up on the system and prefer the older structure? Honestly, there's nothing wrong with 5.10.45, 5.10.28, or 5.10.13 (all of which I use), so what's your problem with these versions that makes you want to just cut them off?

Using those versions is ok at the moment, I think, but some older BOINC 5 clients and all the BOINC 4 clients are underclaiming, I think.

"Setting-and-forgetting" is not the correct way to do science. Everyone should have some responsibility keeping their programs up to date. Just about five minutes once a year to check the BOINC site, downloading and installing.

If BOINC is so difficult to update, how did they manage to install the BOINC 4/5 in the first place??

HTH.

---

Manned mission to Mars in 2019 Petition <-- Sign this, please.

What about those of us who just don't like the way Boinc 6 sets up on the system and prefer the older structure? Honestly, there's nothing wrong with 5.10.45, 5.10.28, or 5.10.13 (all of which I use), so what's your problem with these versions that makes you want to just cut them off?

-Dave

I agree. I only install 5.10.45 on PC's that I bring on. I like the service install where I can set to my network ID and do the work when the PC is not in use. I have tried a few PC's with 6.2.19, and I do not like the set up.

The "service" install is available in BOINC v6, only under a new name known as the Protected Mode install. Just in case you didn't know.

"Setting-and-forgetting" is not the correct way to do science. Everyone should have some responsibility keeping their programs up to date. Just about five minutes once a year to check the BOINC site, downloading and installing.

Nowhere on the site does it say any participant has to keep up to date, nor does it imply any personal time out of the participant to know what's going on. It is not your place or mine to tell other participants how they should "do science".

Further, the entire point of BOINC is to make it as easy as possible to contribute your spare cycles to science without having to involve yourself once you've installed the app. Any further participation is on the user's volition and is not mandatory.

If BOINC is so difficult to update, how did they manage to install the BOINC 4/5 in the first place??

Nobody said BOINC was difficult to install or update. In fact, the difficulty isn't the source of the issue here. The source of the issue is that BOINC works extremely well especially when left alone, and that's exactly what many people have done.

lets not get to hasty. 5.10.45 is a fine piece of work all other previous versions are caca

---

In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope

lets not get to hasty. 5.10.45 is a fine piece of work all other previous versions are caca

Agreed--I use 5.10.45 and 5.10.41 with absolutely no difficulties

---

I see no reason to reject anything newer than that. Once they get a stable build of 6.4 released, then perhaps we could revisit the arguments for restricting BOINC versions.

There used to be talk about really old clients not calculating credit correctly (version 4.8, I believe?). However, I do not know of any credit-calculation algorithm changes since BOINC 5.8.

Unless Henri has valid argument(s) for his point, there's not much support here for the suggestion to block below version 6.