Okay & Thanks for the Info, I'll wait then ...

Regarding Linux the 'NCI' Monkeys apps, here's a how to create the app_config.xml file on Ubuntu16.04 (and similar)

How To create an app_config.xml on Ubuntu 16.04 to instruct Boinc Manager that the Monkeys apps only use 0.0001 CPU’s.

Note this is for a repository installation of Boinc+Bonic Manager. It should work for other similar versions of Linux too.

The app_config.xml file needs to be created in the following folder:
var/lib/boinc-client/projects/boincgridcomputingcenter.org/
You won’t normally have access to that folder but you should be able to open it using the following:

Open a Terminal (top left, search for Ter and click Terminal),
type,

gksu nautilus /var/lib/boinc-client/projects/

One way of creating a new .xml file is to copy and paste an existing .xml file (one of the job_monkeys_v1.xml file for example) then rename the copy to app_config.xml. Then open it – Right click, Open With > gedit.

Delete the contents and paste in the following:

<app_config>
<app_version>
<app_name>monkeys_v1</app_name>
<avg_ncpus>0.000001</avg_ncpus>
</app_version>
<app_version>
<app_name>monkeys_v2</app_name>
<avg_ncpus>0.000001</avg_ncpus>
</app_version>
<app_version>
<app_name>monkeys_v3</app_name>
<avg_ncpus>0.000001</avg_ncpus>
</app_version>
<app_version>
<app_name>monkeys_v4</app_name>
<avg_ncpus>0.000001</avg_ncpus>
</app_version>
</app_config>

Save the file (top right).

From Boinc Manager, re-read the config files:
Options, Read config files.

The Event log should confirm the file has been read with a messages saying that it has ‘Found app config file’. It might give you some error message if there is a ‘typo’ mistake in the file.

Note this seems to apply to new tasks only; existing tasks have already been allocated CPU usage and slot. IIRC it was immediate on Windows. As the tasks only last 1h it’s not a big issue.
Also, if you restart the computer tasks will restart from 0 (as they don’t checkpoint) and then only use 1e-06 CPUs.

Monkeys v1 (NCI) - Generator/1 seed
Monkeys v2 (NCI) - Seeker/1 seed
Monkeys v3 (NCI) - Generator/Word seed
Monkeys v4 (NCI) - Seeker/Word seed

v2 and v4 are also online with WUs now for a while (about a day?) but still not counted...!
Besides, I've running v1 as NCI and the app is called like above, but in my account it shows up without NCI, so it needs a rename:
Grid computing center Monkeys v1 - Generator/1 seed 15.48

Part of the tasks v1 are counted as "Monkeys v1 (NCI) - Generator/1 seed"
Probably an application for android.
According to Bok's information - 800 hours in the last week

Monkeys v1 (NCI) - Generator/1 seed
Monkeys v2 (NCI) - Seeker/1 seed
Monkeys v3 (NCI) - Generator/Word seed
Monkeys v4 (NCI) - Seeker/Word seed

v2 and v4 are also online with WUs now for a while (about a day?) but still not counted...!
Besides, I've running v1 as NCI and the app is called like above, but in my account it shows up without NCI, so it needs a rename:
Grid computing center Monkeys v1 - Generator/1 seed 15.48

Part of the tasks v1 are counted as "Monkeys v1 (NCI) - Generator/1 seed"
Probably an application for android.
According to Bok's information - 800 hours in the last week

I am only getting "Monkeys v1 (NCI) - Generator/1 seed" on my old Galaxy S4 when it is sent the android_arm64_pie apps. Which it can't run.
On my PCs looking over the client_state.xml and data files for wuprop I see "Monkeys v1 (NCI) - Generator/1 seed" instead of "Monkeys v1 - Generator/1 seed".


So it is not clear to me why most of the Monkeys v1 apps are not displaying correctly on the wuprop site.

New russian project - odlk
Active new application - odlk1@home (Windows & Linux, x86_64)
40Mb download, low memory, ~16 min/WU, now a high level of errors

I managed to get a few tasks on odlk and they validated ok on windows but the linux ones failed, looks like it's not statically linked and fails on a GLIBC dependency

[root@dbase4 boinc.progger.info_odlk]# ldd odlk1_1.1_x86_64-pc-linux-gnu
./odlk1_1.1_x86_64-pc-linux-gnu: /lib64/libstdc++.so.6: version `GLIBCXX_3.4.21' not found (required by ./odlk1_1.1_x86_64-pc-linux-gnu)
linux-vdso.so.1 => (0x00007ffefadf7000)
libpthread.so.0 => /lib64/libpthread.so.0 (0x00007f5d36909000)
libstdc++.so.6 => /lib64/libstdc++.so.6 (0x00007f5d36601000)
libm.so.6 => /lib64/libm.so.6 (0x00007f5d362fe000)
libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00007f5d360e8000)
libc.so.6 => /lib64/libc.so.6 (0x00007f5d35d26000)
/lib64/ld-linux-x86-64.so.2 (0x00007f5d36f27000)

---

Denis@Home announced new research and new app "Carro-Rodriguez-Laguna-Pueyo Epicardial Model (Carro et al. 2011) for human ventricular cells", however there are no WUs at this moment. They are going to start simulation today, so we can expect that they will appear soon.

---

I don't think that's a new app. Given I have 5000+ hrs in it...

---

I don't think that's a new app. Given I have 5000+ hrs in it...

Well, for me it is a new one :)

---

Yup ... Been there, Done that ...

DENIS@home Carro-Rodriguez-Laguna-Pueyo Epicardial Model (Carro et al. 2011) for human ventricular cells 100,473.67

cpdnboinc Weather At Home 2 (wah2) 10,270.13 10,270.13
cpdnboinc UK Met Office HadAM3P-HadRM3P Pacific North West 4,089.85 4,089.85
cpdnboinc UK Met Office HadCM3 short 3,190.80 3,190.80
cpdnboinc Weather At Home 2 (wah2) (region independent) 226.45 226.45

Why?

https://wuprop.boinc-af.org/forum_thread.php?id=191&postid=4976

http://climateapps2.oerc.ox.ac.uk/cpdnboinc/server_status.php

cpdnboinc Weather At Home 2 (wah2) 10,270.13 10,270.13
cpdnboinc UK Met Office HadAM3P-HadRM3P Pacific North West 4,089.85 4,089.85
cpdnboinc UK Met Office HadCM3 short 3,190.80 3,190.80
cpdnboinc Weather At Home 2 (wah2) (region independent) 226.45 226.45

Why?

And now I have new hours on climateprediction (the correct one) plus all these old hours under the incorrect one. I note that ALL the cpdnboinc hours are 'red', indicating over-30-days, even hours earned yesterday.

Hope this isn't like vlhcathome-dev where a month later I was still dragging around all the obsolete named apps... although this seems to have been fixed just as the cpdnboinc thing showed up!

The monthly new app from Citizen Science Grid is out;

EXACT MNIST Batch Norm CNN Trainer

---

The monthly new app from Citizen Science Grid is out;

EXACT MNIST Batch Norm CNN Trainer

They are much quicker. 1st one completed in 44min on a 2760v1. Just 3 validated so far but about 33% higher PPD.

All the hours went under the 'EXACT MNIST Batch CNN Trainer 2.0' app instead of the new name. Average of 0.128 point/CPU second over 84 tasks vs 0.096 with last weeks app.

EXACT MNIST Batch Norm CNN Trainer isn't showing here yet.

Noticed today that I was getting credit against EXACT MNIST Batch CNN Trainer 2.0 and when I went to the project I found that my selection for the new project only had been changed to all apps! There is no 'accept new apps' option.

However 335h of CPU time now appears for the old app and not for the new app. Only 2 old tasks were picked up and they haven't completed. So I should have ~330h for the new app that was added against the old app. This is likely to get worse when I return the next few WUProp tasks, so selecting NNW for CSG until it's sorted.

Active new application - odlk2@home

EXACT MNIST Batch Norm CNN Trainer isn't showing here yet.

+1

Active new application - odlk2@home

It's been Posted, I ran about 1400 of the oblk2's before it was posted, all Hr's were lost for Wu's ran Pre Post, I've got 16 Hr's showing ... Good thing is I managed to just squeeze 1000 Hr's out of the odlk1's ... :)

...and there aren't any.

Didn't get so far:

odlk odlk1@home 21.55 0.48
odlk odlk2@home 1.23 1.23

One day [what's the point?] wonder apps!