Dear participants!

I'm trying to run autonomous subproject.
The subproject uses my algorithm “pseudo-associative DLSs”.
Jobs are generated by my program. Jobs check is performed by Belyshev colleague program (family_mar.exe).

I ask everyone to support the subproject.
I really need your help!
I have thousands of tasks, several algorithms, I perform dozens of experiments.
I can't handle all the work!

Perhaps this attempt will help to further translate the stand-alone subproject onto the BOINC platform.
I am ready to do it right now. But I do not know how to run subprojects in the BOINC platform.

Here is the first block of tasks
https://cloud.mail.ru/public/NB13/7aGzktuGe

Try everyone! It works!
Read on for details.

The archive contains 10 WU.
After you unpack the archive, run the script check.bat

The program works only in the Windows system.
After executing the script, you will get the results in the results_block1.txt file.

Please test it and tell how it works.

Thanks!

PS. Please ask your questions.

Hello, ice00!

I ask you to help me in this autonomous subproject.
I ask you to do the following, if possible

1. Give me the opportunity to upload tasks to the site.
2. Allow users to download tasks.
3. Allow users to upload results to the site.
4. Provide the opportunity for me to download the results.
5. Post here the executable family_mar.exe for the Linux system.

About item 5.
You know the family_mar.exe checker.
You compiled this program for the Linux system when we wanted to prepare an experiment of Belyshev.

Next, I ask you to test my first task (link above).
I generate WU myself. You just check.
Can you do it?
Please tell us how it works in Linux.

Thanks!

https://ibb.co/MG6hhSq
Hi, cpu use about 2%.. i dont know if is it ok..??
win cmd dont show azbuka fonts in default , thats why we see only "???" will better create program on windows where is region or L. set to english.. but im not sure about this..
Really good, about new subproject.

Thanks for testing!

Did you to run the script to the end?
10 WU should boot and check.

I do not know how to make the English text appear in the window :(

still running, now run : "34432 "
i will write another message when it end..)

Thanks!

When loading each WU, a new line of ??? will appear.
And so 10 times.

??? probably appears as source code maybe are not with UTF8 for Russian text, but instead force to use a LOCARE_RU with a text ASCII coded that is for native Russian program (at least this is what I see in some sources related to the same author). I try even to use google translate onto that text into the source but it fails too.

Probably it shows up correctly only for Windows installation that have native Russian language installed.

You should see that it works when you will find some solution TXT file produced by the program (the average rate should be something like 1:3 as in 15.967 families tested so far in server, 5.108 produces solutions).

i see , these computers are on Russian colleague soo so the question goes to them. becouse it show correctly only on native russian windows/lin instalation. for me is it no problem
but if program run automatically and dont need to write some command , no problem) for later and for more people to get into run program/project they will must to do some with english or other l. ..
still running ,stable.

It is 2018, please use Unicode/utf-8 and avoid all text encoding problems this way.

You should see that it works when you will find some solution TXT file produced by the program (the average rate should be something like 1:3 as in 15.967 families tested so far in server, 5.108 produces solutions).

No!
Another experiment is now running, not the one you tested on the server.
In my experiment almost every WU produces solutions!
That be seen in the file results_block1.txt.

But the family_mar.exe check program in both experiments is the same.

You should see that it works when you will find some solution TXT file produced by the program (the average rate should be something like 1:3 as in 15.967 families tested so far in server, 5.108 produces solutions).

Have you tested 15.967 families in an experiment of Belyshev?
Can you send me the results?

It is 2018, please use Unicode/utf-8 and avoid all text encoding problems this way.

Thanks.

I think that the problem with the encoding can be solved.
I can post the source code for the family_mar program from the author.

ice00
as I understand it, you have already solved this problem for this program.
Demis helped you with this.
You also compiled this program for a Linux system.

still running ,stable.

Soon be completed.
On my computer, one WU runs for 40-60 minutes.

Thanks.

Here is the second block of tasks
https://cloud.mail.ru/public/6kPm/J61HehrNK

The archive contains WU 11-20.

See instructions here
https://boinc.multi-pool.info/latinsquares/forum_thread.php?id=57&postid=367#367

i have finnished records from first block. i have upload or just post.?
they are in : results_block1.txt

start second. )

i have finnished records from first block. i have upload or just post.?
they are in : results_block1.txt

start second. )

Oh! How good!

ice00 has not yet answered my question about uploading results to the site.

Now you can send me the results_block1.txt file to the home box, e-mail
natalimak1@yandex.ru

Thank you very much!

Here is the block #3 of tasks
https://cloud.mail.ru/public/4D7n/sZA51hsuo

The archive contains WU 21-30.

See instructions here
https://boinc.multi-pool.info/latinsquares/forum_thread.php?id=57&postid=367#367

ice00
I found a letter by Demis in which he sent you the sources for family_mar.
Demis wrote in a letter

Sources translated to English.

See the archive in this letter family_mar.tgz.gz

So the problem with the translation of the source code of the program into English has been solved.
Is not it?

jozef j
I got the results_block1.txt.
Thanks!

Found 169 CF ODLS in results_block1.txt.

This is the processing of results

Найдено марьяжных КФ:
count[1] = 347
count[2] = 24
count[4] = 2
Всего: 373
Найдено соквадратов: 403
КФ соквадратов: 373

After processing it turned out 373 CF ODLS.

Until I checked the solutions for uniqueness.
We already have a fairly large database and some solutions are either repeated or isomorphic.
Therefore, among the solutions found, not all are unique.