Interesting speculation, but no more than that. I think the situation is more that Einstein is well funded (largely because it's associated with the glamour of proving the correctness of the prediction of gravitational waves - though that proof came directly from LIGO, and didn't involve '@Home' volunteers).

And because it's well funded, it's well resourced, and it can afford to "pay its taxes" in the form of contributing staff development hours back to the infrastructure it relies on. The smaller 'one man' projects are just grateful to be able to help themselves to more-or-less working code (and sometimes need us volunteers to explain the finer points to them). They don't have the luxury of spare time to allow them to contribute something back.

I need to restart both programs. Is this normal?

You don't need to restart BoincTask. When you look at it, at the far left is computers.

Restart BOINC, then click on the computer tab and it should show your computers attached.

Click on your computer and it should reattach to it.

Huh, couldn't figure out what you just described. I ran it again, it's been about an hour, and moved 24 tasks both ways (with now 5 waiting to run), so for a machine with 3 high end GPU's running 4 tasks at once, with 36 CPU tasks running (reserved 1 core per GPU task for now), it seems that it might have to be ran hourly?

It again shut down BOINCtasks, and after the moving completed and I restarted BOINC, I had to manually restart that too. I don't see any other way of getting it to start other than the program icon. Am I missing something?

---

it seems that it might have to be ran hourly?

so it seems! You'll definitely want to automate it with Scheduled Tasks.

It again shut down BOINCtasks, and after the moving completed and I restarted BOINC, I had to manually restart that too. I don't see any other way of getting it to start other than the program icon. Am I missing something?

Have you played with batch files before?
If so, just right click on the .cmd file and select Edit.

If there's a REM somewhere at the beginning of a line with code after it, consider removing it. Just don't remove REM when there's documentation.

If there's a line of code you want Windows to ignore, just add REM (with a space after it) at the beginning of the line.
ie: If you don't want to see a specific: Press any key to continue...
place a REM before that PAUSE

Feel free to add empty line anywhere for legibility.
Also, on an empty line you can add: PAUSE

For your specific issues, do 1a. or 1b.
1a. add REM at the beginning of the following line:

taskkill /T /F /FI "STATUS eq RUNNING" /IM boinctasks*    >nul

so it looks like this:

REM taskkill /T /F /FI "STATUS eq RUNNING" /IM boinctasks*    >nul

1b. at the end of the file, remove the REM if there is one before:

START "" /MAX "C:\Program Files\eFMer\BoincTasks\boinctasks64.exe"

2. as for the Boinc Client not starting, it should! lol
The command line is:

"C:\Program Files\BOINC\boinc.exe" --detach

If you do 1b above, with BoincTasks, there's an option to start the Boinc Client 5 secs after it has started.

Keep me posted :-D

Editing batch files! Talk about taking me back to my DOS days... Man, lotsa cobwebs, but you explained it quite well. I'll putz with it this evening. Thanks!

---

Editing batch files! Talk about taking me back to my DOS days... Man, lotsa cobwebs, but you explained it quite well. I'll putz with it this evening. Thanks!

My experience is mostly from the Win3.11 days!
I tried to keep the .cmd file as simple as possible in order for others to improve/modify it however they want.

If you're stuck in the cobwebs, this is the site I was using:
http://ss64.com/nt/

Does DA have much of an interest in SETI any longer, or are we now sort of the ugly redheaded stepchild of BOINC?

. . As a surrogate dad to a gorgeous Ranga I take exception to that comment :)

Oh, and just ran it again, this time it was 19 tasks moved each way, and 9 tasks are waiting to run. One question, before running it, I manually shut down BOINC, but it seems to shut down BOINCtasks as well? The last line in the window says something about restarting BOINC, but it doesn't restart, at least it's not in the taskbar at the bottom. I need to restart both programs. Is this normal?

. . As long as it "knows" your BOINC directory location it will restart the BOINC client. But it doesn't shut down BOINC manager (but I do manually) so it does not restart it. BOINC tray (the icon on the right side of task bar) restarts on one of my rigs but not the other two, but kicks off when manager is restarted so I have no problem with that.

. . As long as it "knows" your BOINC directory location it will restart the BOINC client. But it doesn't shut down BOINC manager (but I do manually) so it does not restart it. BOINC tray (the icon on the right side of task bar) restarts on one of my rigs but not the other two, but kicks off when manager is restarted so I have no problem with that.

Stephen,
Are you still running my semi-automated Alpha from pre-Mr Kevvy?
...or did you get my latest front-end to work with Mr Kevvy's .exe

. . As long as it "knows" your BOINC directory location it will restart the BOINC client. But it doesn't shut down BOINC manager (but I do manually) so it does not restart it. BOINC tray (the icon on the right side of task bar) restarts on one of my rigs but not the other two, but kicks off when manager is restarted so I have no problem with that.

Stephen,
Are you still running my semi-automated Alpha from pre-Mr Kevvy?
...or did you get my latest front-end to work with Mr Kevvy's .exe

I am running the pre-Mr Kevvy. So I guess this doesn't apply to the combined app?

Does DA have much of an interest in SETI any longer, or are we now sort of the ugly redheaded stepchild of BOINC?

. . As a surrogate dad to a gorgeous Ranga I take exception to that comment :)

Lol, Sorry, Stephen, I actually had to Goodle it to translate it into American. You Aussies, love your slang. ;-) And wait a minute, if you are the surrogate, does that also make you a... :-O

I use Ginger for the most part, or in select company another slightly more colorful, un-PC name. ;-)

---

Stephen,
Are you still running my semi-automated Alpha from pre-Mr Kevvy?
...or did you get my latest front-end to work with Mr Kevvy's .exe

I am running the pre-Mr Kevvy. So I guess this doesn't apply to the combined app?

It might...and you pointed out a variable I assumed Al already knew about...if he read the README! ;-)
Al, as Stephen pointed out:

As long as it "knows" your BOINC directory location it will restart the BOINC client.

please check if Boinc is installed in:

C:\Program Files\BOINC\

If not, you'll need to make changes to this line:

"C:\Program Files\BOINC\boinc.exe" --detach

...but then again, my script would not be able to start Mr Kevvy's .exe if it wasn't in that directory....grrrrrr

Hopefully Al can give me ideas as to why it doesn't restart the Boinc Client.
Valiant effort Stephen. It was a very good lead.

Cheers,
Rob :-)

PS: Don't know if anyone noticed on SG WoW; I was in the Top10 for the 1st 4 hrs.
I say that's mission accomplished for my 1st WoW with only 2 rigs and 3 good GPUs.
And to banter Petri, he ***still*** hasn't passed me after 9hrs. I'm now in 15th and he's only in 30th.
At least I have handed over the Leo torch to Mr Kevvy, who's now in 10th with no sign of slowing down.
Woop-woop to DQOs!!!
Can't wait to see if there an advantage of running some form of DQO with Petri's app.
STOP THE PRESS: Jason mentionned a possible "...extended wider alpha..." soonish.

Can't wait to see if there an advantage of running some form of DQO with Petri's app.
STOP THE PRESS: Jason mentionned a possible "...extended wider alpha..." soonish.

I've been sounding like a broken record when I ask him when his "special application" would make it off Beta and into the general public. His output on CUDA tasks is phenomenal with his special code. With his app I would go back and start running CUDA again. At this point, it's not worth it since the OpenCL app cleans the floors on CUDA50 with OpenCL optimized settings.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

At this point, it's not worth it since the OpenCL app cleans the floors on CUDA50 with OpenCL optimized settings.

In the 3.75 months I've been active in the forums, I still haven't seen conclusive data.
I tried but I didn't know back then that NV_SoG can run 2tasks on GTX 750 Ti.
My picture though shows Cuda50 is in the lead when doing 3onCuda50 vs 1onNV_SoG...with Mr Kevvy's .exe
Keep in mind that I was processing 6 guppi on CPUcores on both rigs
...and with the Cuda50 rig, I could easily have done a 7th, plus uped to 4nonVLAR/GPU.

Jimbolous is interested after WoW to help out in a better test format:
His HP Z400 with 2x GTX 750 Ti (using Cuda50)
vs
my HP Z400 with 2x GTX 750 Ti (using NV_SoG)

But if Petri & Jason have an alpha GPU Cuda## app available by then, that test might not happen.

Does DA have much of an interest in SETI any longer, or are we now sort of the ugly redheaded stepchild of BOINC?

. . As a surrogate dad to a gorgeous Ranga I take exception to that comment :)

Lol, Sorry, Stephen, I actually had to Goodle it to translate it into American. You Aussies, love your slang. ;-) And wait a minute, if you are the surrogate, does that also make you a... :-O

I use Ginger for the most part, or in select company another slightly more colorful, un-PC name. ;-)

. . Yes I am ... a childless bachelor ! :)

. . I started calling her "bluey" at one point and was nearly divorced on the spot, yet her then boyfriend calling her ginger was just fine with her :). Though ginger is fine too, as in Ginger in Gilligan's Island. BTW "bluey" is more Aussie slang for a redhead. But as for the surrogacy, it is a role I am privileged to have, since her RL father is a waste of space. That is unkind to say but as fathers go, perfectly justified.

. . But this is now wildly off topic, my apologies one and all.

Can't wait to see if there an advantage of running some form of DQO with Petri's app.
STOP THE PRESS: Jason mentionned a possible "...extended wider alpha..." soonish.

. . Picture me in the meerkat lookout mode ....

PS: Don't know if anyone noticed on SG WoW; I was in the Top10 for the 1st 4 hrs.
I say that's mission accomplished for my 1st WoW with only 2 rigs and 3 good GPUs.
And to banter Petri, he ***still*** hasn't passed me after 9hrs. I'm now in 15th and he's only in 30th.
At least I have handed over the Leo torch to Mr Kevvy, who's now in 10th with no sign of slowing down.
Woop-woop to DQOs!!!

. . Well done but it is only seeing one of my rigs so I am sightly handicapped :(

. . Mr Kevvy can lead the way

Jimbolous is interested after WoW to help out in a better test format:
His HP Z400 with 2x GTX 750 Ti (using Cuda50)
vs
my HP Z400 with 2x GTX 750 Ti (using NV_SoG)

. . That would make for an interesting comparison. Using SoG I am running 4 simultaneous SoG WUs on my GTX 970 from one CPU core (on a pentium D 3GHz) and the runtimes are approx 28 mins, which is not too bad. My problem is the upgrade to Win10 on that machine was a mistake. It does not support the Nvidia NForce4 SLI chipset on this ASUS motherboard, and neither ASUS nor Nvidia offer win10 drivers for it also. This lack of proper drivers is causing me endless headaches with lockups and other probs, all related to machine functions not so much BOINC itself. For that matter I am running 3 concurrent SoG WUs on my GTX950 as well and they are running in approx 36 to 38 mins. Also using only one CPU core. So SoG is not the problem some ppl think.

@ Stephen is that time total for 3 work units or what the time is after you divide the total by 3?

I would feel a lot better about tools like these if the server was properly informed about the switch (ie: the recorded stats are wrong and record that the GUPPI ran on your GPU). If the accounting was accurate I'd be all for these tools (and even help work on them).

Actually, I'm fairly certain that the app version that each task actually ran with is reported back to the server as part of the scheduler request. For instance, if you take a look at the task detail for one that one of my hosts just reported, Task 5095021623, you'll see that it was originally assigned to "Anonymous platform (NVIDIA GPU)". However, it actually ran as a CPU task, as shown in my Event Log:

8/15/2016 2:01:48 PM | SETI@home | [cpu_sched] Starting task blc5_2bit_guppi_57451_66326_HIP117463_OFF_0014.17591.416.18.27.126.vlar_1 using setiathome_v8 version 800 in slot 1

The sched_request_setiathome.berkeley.edu.xml file in which the task result was reported confirms this:

<result>
<name>blc5_2bit_guppi_57451_66326_HIP117463_OFF_0014.17591.416.18.27.126.vlar_1</name>
<final_cpu_time>17550.240000</final_cpu_time>
<final_elapsed_time>20616.165501</final_elapsed_time>
<exit_status>0</exit_status>
<state>5</state>
<platform>windows_intelx86</platform>
<version_num>800</version_num>
<final_peak_working_set_size>38137856</final_peak_working_set_size>
<final_peak_swap_size>47706112</final_peak_swap_size>
<final_peak_disk_usage>51128</final_peak_disk_usage>
<app_version_num>800</app_version_num>
<stderr_out>
...

For a task that ran on the GPU as originally scheduled, such as Task 5097687040, the task result is reported as:

<result>
<name>31jl10aa.16178.25757.7.34.128_0</name>
<final_cpu_time>504.491600</final_cpu_time>
<final_elapsed_time>5278.716400</final_elapsed_time>
<exit_status>0</exit_status>
<state>5</state>
<platform>windows_intelx86</platform>
<version_num>800</version_num>
<plan_class>cuda50</plan_class>
<final_peak_working_set_size>138244096</final_peak_working_set_size>
<final_peak_swap_size>165892096</final_peak_swap_size>
<final_peak_disk_usage>28370</final_peak_disk_usage>
<app_version_num>800</app_version_num>
<stderr_out>
...

So, the info is there, and could be extracted just as the CPU Time and Elapsed Time is currently being extracted from those <result> blocks to populate the task result pages that we see online. In fact, the Peak Working Set Size, Peak Swap Size, and Peak Disk Usage have only been added to the task pages in the last year or so, I think.

The question, however, is whether there might be unintended consequences to do so. Certainly, the current server code wasn't designed to make changes to the Application version based on the reported result, even though the info is there. Perhaps that's for a very good reason tied to the integrity of the database. Then again, maybe it's just because the original designers didn't anticipate anybody fiddling with the task assignments on the client side to the extent that's now being discussed.

In any event, I would think the programming changes to parse that info would be fairly trivial. Convincing someone to make such a policy change, however, would likely not be. Good luck with that one!