Strange...
You have deleted out the folders and subfolders after uninstalls in windows ???
Sorry if my english is a bit bad .. :)

Strange...
You have deleted out the folders and subfolders after uninstalls in windows ???
Sorry if my english is a bit bad .. :)

---

Strange...
You have deleted out the folders and subfolders after uninstalls in windows ???
Sorry if my english is a bit bad .. :)

Sorry,
when validation of the message, I thought about a validation error and validated too many times ...

---

Ongoing same problem regarding sending in completed results -- now at 9 hours.

Ongoing same problem regarding sending in completed results -- now at 9 hours.

Not validating wu's. Have 5 wu's that are ready to report, but are just sitting there. Everything seemed to work fantastic up until yesterday. Great work on behalf of the Collatz team!

Have a great day!!!

I am wondering if Slicker is not available -- from yesterday onward for some reason as there has been no updates regarding the sudden problems from 16 hours ago.

Hi Slicker,

I'm using Linux primarily. So if you make me have access to the source code, I will compile Linux binaries for nvidia and cpu (sorry, no I don't have amd toolkit) and can try cross-platform test if you give Windows workunits/results pairs.

I had been programming for seti@home beta project in C++ with Eric Korpela in 2005.

---

Luckiest man in the world

Strange...
You have deleted out the folders and subfolders after uninstalls in windows ???
Sorry if my english is a bit bad .. :)

Yes it did it for me too, you also have to reattach using http:// NOT https://

Hi Slicker,

I will help you compile and test Linux applications (cpu and nvidia gpu). If you need, I can also build Windows application only for cpu (I don't have gpu environment for Windows, because my Windows' run on old machines and vmware.) To tell the truth, I have Intel C++/C compilers which optimize applications run according to its capability (SSE4.2 or AVX, AVX512 etc.) both for Linux and Windows. So I can help you in that direction if I have access to the source code.

Sincerely,

Hi Slicker,

I will help you compile and test Linux applications (cpu and nvidia gpu). If you need, I can also build Windows application only for cpu (I don't have gpu environment for Windows, because my Windows' run on old machines and vmware.) To tell the truth, I have Intel C++/C compilers which optimize applications run according to its capability (SSE4.2 or AVX, AVX512 etc.) both for Linux and Windows. So I can help you in that direction if I have access to the source code.

Sincerely,

The issue isn't with being able to compile the applications. The delay is due to the same code returning invalid results when compiled with GCC on Linux. That exact same code works correctly using VS2017 on Windows. I removed the new Boost code changing out the 128 bit integers with GMP and/or custom bit shift functions but the results are the same. That leads me to believe it has to do with the sieve generation and/or the 64 or 128 bit emulation/optimization. I've now switched gears and am working on a universal version that works with GCC/MinGW/MinGW-w64 so except for a few platform specific definitions, the code will be exactly the same across all platforms. It will then require testing those apps against the existing and older non-sieve apps to make sure the results all still match.

Good luck Slicker and thank you very much for your work! My GTX 1080 under Ubuntu is hungry for work! :)

---

Hello,

after detach/attach, the collatz work dir:

projects\boinc.thesonntags.com_collatz

stay empty, and my host (windows + ATI, ID:139469) can't download work:

30/04/2018 23:04:41 | collatz | Detaching from project
30/04/2018 23:04:56 | | Fetching configuration file from https://boinc.thesonntags.com/collatz/get_project_config.php
30/04/2018 23:05:13 | collatz | Master file download succeeded
30/04/2018 23:05:18 | collatz | Sending scheduler request: Project initialization.
30/04/2018 23:05:18 | collatz | Requesting new tasks for CPU and AMD/ATI GPU
30/04/2018 23:05:19 | collatz | Scheduler request completed: got 0 new tasks
30/04/2018 23:05:19 | collatz | Not sending work - last request too recent: 45 sec
30/04/2018 23:07:21 | collatz | Sending scheduler request: To fetch work.
30/04/2018 23:07:21 | collatz | Requesting new tasks for AMD/ATI GPU
30/04/2018 23:07:23 | collatz | Scheduler request completed: got 0 new tasks

Is there an ETA on the *nix apps yet?

I've got a new "universal" GCC based app that compiles with MinGW for windows and GCC on Linux and BSD Unix (a.k.a. OS X).
GMP = gone. Boost = Gone. I'm using a new UInt256 but custom library instead. The bugs were fixed today and I'm testing now.... So far so good. It takes about 2 hours for each test but if all goes well, I'm hoping to have CPU apps in a few days. (Yeah, I know people want the GPU apps and those will follow shortly. It's just easier to debug the CPU apps to get the kinks out and then hope the GPU apps work equally as well since they are tougher to debug.

@Alessandro : There is no work available for all platforms / OS / GPUs at the moment, Slicker is working on priorities right now, be patient !

Thanks again for all the hard work Slicker, et ne baisses pas les bras ;)


Edit : I hadn't seen Slicker latest message just above, so mine is even more correct :)

@Alessandro : There is no work available for all platforms / OS / GPUs at the moment, Slicker is working on priorities right now, be patient !

Thanks again for all the hard work Slicker, et ne baisses pas les bras ;)


Edit : I hadn't seen Slicker latest message just above, so mine is even more correct :)

Also be sure to use use "http://" and NOT "https://" when you sign on again.

Just wondering who the user is that was cheating. These work units take some time to process. I'm using stock apps and each work unit takes approximately 13 minutes to process.

Just wondering who the user is that was cheating. These work units take some time to process. I'm using stock apps and each work unit takes approximately 13 minutes to process.

For a 1080 use these codes:
verbose=1
kernels_per_reduction=48
sleep=1
threads=8
lut_size=17
reduce_CPU=0
sieve_size=25
cache_sieve=1

Put them in the config file in c:\program data\boinc\projects\collatz

This is what I am seeing when I use the codes on one of my own 1080's:
264.94 1.31 28,082.07

For comparison your numbers are:
756.76 0.45 37,056.03