app_info.xml, aistubs and other confusing things

Message boards : Number crunching : app_info.xml, aistubs and other confusing things
Message board moderation

To post messages, you must log in.

1 · 2 · 3 · Next

AuthorMessage
Stephen "Heretic" Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 3653
Credit: 80,718,960
RAC: 163,603
Australia
Message 1933883 - Posted: 5 May 2018, 23:17:46 UTC
Last modified: 5 May 2018, 23:19:59 UTC

. . An attempt to "upgrade" my Lunatics 0.45 Beta 6 setup to run SoG r3584 insteaad of r3557 has opened the proverbial can of worms and given rise to millions <slight exageration> of questions.

. . I understand that Richard had increased the size of the Lunatics aistub to mamoth proportions to accommodate every version of each app in existence, but in poring over the aistubs in my project folder I have discovered a great deal of what seems to be unneccesary redundancy.

. . How do I correct this without crippling the machine??

. . The simplest and most coherent of the aistubs file is for the CPU based MB tasks yet even this to my mind has a redundant section for the version.

<app_info> 
    <app>
        <name>setiathome_v8</name>
    </app>
    <file_info>
        <name>MB8_win_x64_AVX_VS2010_r3330.exe</name>
        <executable/>
    </file_info>
    <file_info>
        <name>libfftw3f-3-3-4_x64.dll</name>
	<executable/>
    </file_info>
    <file_info>
 	<name>mb_cmdline_win_x64_AVX_VS2010.txt</name>
    </file_info>
    <app_version>
        <app_name>setiathome_v8</app_name>
        <version_num>800</version_num>
        <platform>windows_x86_64</platform>
        <api_version>7.5.0</api_version>
        <file_ref>
            <file_name>MB8_win_x64_AVX_VS2010_r3330.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>libfftw3f-3-3-4_x64.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>mb_cmdline_win_x64_AVX_VS2010.txt</file_name>
            <open_name>mb_cmdline.txt</open_name>
        </file_ref>
    </app_version>
    <app_version>
        <app_name>setiathome_v8</app_name>
        <version_num>800</version_num>
        <platform>windows_intelx86</platform>
        <api_version>7.5.0</api_version>
        <file_ref>
            <file_name>MB8_win_x64_AVX_VS2010_r3330.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>libfftw3f-3-3-4_x64.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>mb_cmdline_win_x64_AVX_VS2010.txt</file_name>
            <open_name>mb_cmdline.txt</open_name>
        </file_ref>
    </app_version>
</app_info>


. . Isn't the second instance of the app_version section purely redundant???? Would removing it actually cause a problem? To me it seems not but I have been wrong before ...

Stephen

?
ID: 1933883 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 5331
Credit: 356,930,717
RAC: 1,089,560
United States
Message 1933904 - Posted: 6 May 2018, 1:31:52 UTC - in response to Message 1933883.  

<platform>windows_x86_64</platform>


Is for 64 bit Systems.

<platform>windows_intelx86</platform>


Is for 32 bit Systems.

So if you don't intend to run any 32 bit Operating Systems, you can safely remove all sections defining 32 bit Systems. One way way to cut in half the size of the app_info.xml file.

The other way to reduce the size of the app_info.xml is to remove all sections defining plan classes you don't intend to run their applications like:

<plan_class>cuda22</plan_class>


etc. etc. for the other plan_classes like cuda32, cuda42 and cuda50.

So chop down the .aistub files to just 64 bit MB and AP and SoG plan classes.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1933904 · Report as offensive
Stephen "Heretic" Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 3653
Credit: 80,718,960
RAC: 163,603
Australia
Message 1933909 - Posted: 6 May 2018, 2:24:46 UTC - in response to Message 1933904.  
Last modified: 6 May 2018, 2:34:41 UTC

<platform>windows_x86_64</platform>


Is for 64 bit Systems.

<platform>windows_intelx86</platform>


Is for 32 bit Systems.

. . I thought I must have been missing something .. that clears that up and now I know how to determine what is or isn't unnecessary.

So if you don't intend to run any 32 bit Operating Systems, you can safely remove all sections defining 32 bit Systems. One way way to cut in half the size of the app_info.xml file.

. . I will be doing that part for sure.

The other way to reduce the size of the app_info.xml is to remove all sections defining plan classes you don't intend to run their applications

. . And that as well :)

. . It also solves the problem of why I could not get the r3584 download to work, it is the x86 version not the x64 version .... again D'oh!

. . So many traps for players (variables).

. . Thanks Keith.

Stephen

:)
ID: 1933909 · Report as offensive
Bruce
Volunteer tester

Send message
Joined: 15 Mar 02
Posts: 106
Credit: 104,787,273
RAC: 11,229
United States
Message 1933924 - Posted: 6 May 2018, 4:58:27 UTC
Last modified: 6 May 2018, 5:02:44 UTC

Keith is right. You as a single user don't really need everything that the install puts into the app_info.xml.
Richard does a super job with the installer, but because it is a one size fits all app he throws in everything including the kitchen sink and a old pair of boots. Why, because someone might need them. This way the installer covers the widest range of equipment, this is a good thing, but it makes for a large app_info.

When I was running Win7, before I switched to Linux Mint, this was my entire app_info.xml. I posted it before, but here it is again.

<app_info>
<app>
<name>setiathome_v8</name>
</app>
<file_info>
<name>MB8_win_x64_AVX_VS2010_r3330.exe</name>
<executable/>
</file_info>
<file_info>
<name>libfftw3f-3-3-4_x64.dll</name>
<executable/>
</file_info>
<file_info>
<name>mb_cmdline_win_x64_AVX_VS2010.txt</name>
</file_info>
<app_version>
<app_name>setiathome_v8</app_name>
<version_num>800</version_num>
<platform>windows_x86_64</platform>
<file_ref>
<file_name>MB8_win_x64_AVX_VS2010_r3330.exe</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>libfftw3f-3-3-4_x64.dll</file_name>
</file_ref>
<file_ref>
<file_name>mb_cmdline_win_x64_AVX_VS2010.txt</file_name>
<open_name>mb_cmdline.txt</open_name>
</file_ref>
</app_version>
<app>
<name>astropulse_v7</name>
</app>
<file_info>
<name>AP7_win_x86_SSE2_OpenCL_NV_r2887.exe</name>
<executable/>
</file_info>
<file_info>
<name>libfftw3f-3-3-4_x86.dll</name>
<executable/>
</file_info>
<file_info>
<name>ap_cmdline_win_x86_SSE2_OpenCL_NV.txt</name>
</file_info>
<app_version>
<app_name>astropulse_v7</app_name>
<version_num>705</version_num>
<platform>windows_intelx86</platform>
<avg_ncpus>0.04</avg_ncpus>
<max_ncpus>0.2</max_ncpus>
<plan_class>cuda_opencl_100</plan_class>
<cmdline></cmdline>
<coproc>
<type>CUDA</type>
<count>0.5</count>
</coproc>
<file_ref>
<file_name>AP7_win_x86_SSE2_OpenCL_NV_r2887.exe</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>libfftw3f-3-3-4_x86.dll</file_name>
</file_ref>
<file_ref>
<file_name>ap_cmdline_win_x86_SSE2_OpenCL_NV.txt</file_name>
<open_name>ap_cmdline.txt</open_name>
</file_ref>
</app_version>
<app>
<name>setiathome_v8</name>
</app>
<file_info>
<name>MB8_win_x86_SSE3_OpenCL_NV_SoG_r3584.exe</name>
<executable/>
</file_info>
<file_info>
<name>libfftw3f-3-3-4_x86.dll</name>
<executable/>
</file_info>
<file_info>
<name>mb_cmdline_win_x86_SSE3_OpenCL_NV_SoG.txt</name>
</file_info>
<app_version>
<app_name>setiathome_v8</app_name>
<version_num>822</version_num>
<platform>windows_intelx86</platform>
<avg_ncpus>0.04</avg_ncpus>
<max_ncpus>0.2</max_ncpus>
<plan_class>opencl_nvidia_SoG</plan_class>
<cmdline></cmdline>
<coproc>
<type>CUDA</type>
<count>0.5</count>
</coproc>
<file_ref>
<file_name>MB8_win_x86_SSE3_OpenCL_NV_SoG_r3584.exe</file_name>
<main_program/>
</file_ref>
<file_ref>
<file_name>libfftw3f-3-3-4_x86.dll</file_name>
</file_ref>
<file_ref>
<file_name>mb_cmdline_win_x86_SSE3_OpenCL_NV_SoG.txt</file_name>
<open_name>mb_cmdline.txt</open_name>
</file_ref>
</app_version>
</app_info>

Hope it helps.
Bruce
ID: 1933924 · Report as offensive
Stephen "Heretic" Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 3653
Credit: 80,718,960
RAC: 163,603
Australia
Message 1933926 - Posted: 6 May 2018, 5:44:43 UTC - in response to Message 1933924.  

Keith is right. You as a single user don't really need everything that the install puts into the app_info.xml.
Richard does a super job with the installer, but because it is a one size fits all app he throws in everything including the kitchen sink and a old pair of boots. Why, because someone might need them. This way the installer covers the widest range of equipment, this is a good thing, but it makes for a large app_info.

When I was running Win7, before I switched to Linux Mint, this was my entire app_info.xml. I posted it before, but here it is again.

Hope it helps.


. . I noticed that you were running 64 bit app for AP but the x86 (32 bit) app for r3584. That wouldn't run for me and I thought it meant I needed a 64 bit version of r3584 which I cannot find. This is so confusing. And which version of AP is best, I noticed your file referred to vers 705, mine lists 700, 701, 703 and 705. Every aistub file I look at makes me more confused.

Stephen

? ?
ID: 1933926 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 5331
Credit: 356,930,717
RAC: 1,089,560
United States
Message 1933930 - Posted: 6 May 2018, 6:22:05 UTC - in response to Message 1933926.  

Stephen, I will repeat this one more time . . . . and then I will shut up.

The version number is meaningless. There is only one app for each application/plan_class type. That's it ONE. One AP, One MB, One SoG.

Raistmer found it was pointless to make the SoG app 64 bit because it was slower than 32 bit. So it is 32 bit. It runs on 32 or 64 bit just fine.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1933930 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 12097
Credit: 123,955,090
RAC: 46,893
United Kingdom
Message 1933936 - Posted: 6 May 2018, 7:06:40 UTC

Specifically, the additional aistubs are present to avoid trashing cached work that you might have downloaded earlier - whether stock, a known Beta application, because you ran the wrong bitness of the installer last time, because you haven't updated for a couple of years, or anything else I could think of.

Once your pre-existing cache has run through, and you only have tasks fetched under the current set of installer options, you can throw away the app versions that you're not using. But why bother? It's complicated to think through, prone to error, and you have to stop BOINC while you do it.
ID: 1933936 · Report as offensive
Bruce
Volunteer tester

Send message
Joined: 15 Mar 02
Posts: 106
Credit: 104,787,273
RAC: 11,229
United States
Message 1933938 - Posted: 6 May 2018, 7:19:44 UTC
Last modified: 6 May 2018, 7:38:59 UTC

Stephen.
I believe all GPU apps are 32bit, most CPU are also 32bit, but there are some that are 64bit.
If you look you will see that the MB CPU app is 64bit with Platform of windows_x86_64.
The AP GPU app is 32bit with Platform of windows_intelx86. The MB GPU app is listed the same way.
As far as version numbers, those can be almost anything. 700 or 800 usually refer to CPU apps, anything else probably means a GPU app. For AP 700 means CPU and 705 means the latest GPU app.
Version numbers are just to show which are the newest or the oldest. If you are running the r2887 AP app, then it is 705. The r3584 SoG MB app is 822 and so on.
You could probably call them version Bob and Frank if you wanted to, but I don't think I'd try that.

My OS was Win7 Pro 64bit. With a 32bit OS you are limited to 32bit apps only, but with a 64bit OS you can run either 32bit or 64bit, it does not matter. My app_info demonstrates that.
It's always best to run your cache dry before you shut down BOINC to edit your app_info.xml

I hope this helps to clarify things for you.
Bruce
ID: 1933938 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 12097
Credit: 123,955,090
RAC: 46,893
United Kingdom
Message 1933940 - Posted: 6 May 2018, 7:30:29 UTC - in response to Message 1933938.  

It makes life much easier if you stick to the stock release version numbers shown on the Applications page
ID: 1933940 · Report as offensive
Bruce
Volunteer tester

Send message
Joined: 15 Mar 02
Posts: 106
Credit: 104,787,273
RAC: 11,229
United States
Message 1933945 - Posted: 6 May 2018, 7:46:48 UTC - in response to Message 1933940.  

It makes life much easier if you stick to the stock release version numbers shown on the Applications page


That's true and is excellent advice.
Hope I didn't step on any toes when I mentioned the kitchen sink.
Bruce
ID: 1933945 · Report as offensive
Stephen "Heretic" Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 3653
Credit: 80,718,960
RAC: 163,603
Australia
Message 1933995 - Posted: 6 May 2018, 15:47:58 UTC - in response to Message 1933936.  
Last modified: 6 May 2018, 15:48:39 UTC

Specifically, the additional aistubs are present to avoid trashing cached work that you might have downloaded earlier - whether stock, a known Beta application, because you ran the wrong bitness of the installer last time, because you haven't updated for a couple of years, or anything else I could think of.

Once your pre-existing cache has run through, and you only have tasks fetched under the current set of installer options, you can throw away the app versions that you're not using. But why bother? It's complicated to think through, prone to error, and you have to stop BOINC while you do it.


. . I'm trying to sneak up on it by sorting out the aistubs then just doing the aimerge with BOINC shut down.

. . When I added the aistub for the r3584 package it would not run ... and I'm trying to work out why ... by eliminating everything that is unnecessary so I have only the relevant apps in the app_info.xml but I am finding it somewhat confusing.

. . I can see why most people avoid it.

. . Do I have to remover the older apps from the project folder so the newer apps will run? And why is the AP app an SSE one rather the older AVX??

. . Sorry but I don't don't know enough of the development path to know which way to jump. I am trying to get a clearer picture in my head.

Stephen

:(
ID: 1933995 · Report as offensive
Stephen "Heretic" Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 3653
Credit: 80,718,960
RAC: 163,603
Australia
Message 1933997 - Posted: 6 May 2018, 15:54:39 UTC - in response to Message 1933938.  
Last modified: 6 May 2018, 15:55:52 UTC

Stephen.
I believe all GPU apps are 32bit, most CPU are also 32bit, but there are some that are 64bit.
If you look you will see that the MB CPU app is 64bit with Platform of windows_x86_64.
The AP GPU app is 32bit with Platform of windows_intelx86. The MB GPU app is listed the same way.
As far as version numbers, those can be almost anything. 700 or 800 usually refer to CPU apps, anything else probably means a GPU app. For AP 700 means CPU and 705 means the latest GPU app.
Version numbers are just to show which are the newest or the oldest. If you are running the r2887 AP app, then it is 705. The r3584 SoG MB app is 822 and so on.
You could probably call them version Bob and Frank if you wanted to, but I don't think I'd try that.

My OS was Win7 Pro 64bit. With a 32bit OS you are limited to 32bit apps only, but with a 64bit OS you can run either 32bit or 64bit, it does not matter. My app_info demonstrates that.
It's always best to run your cache dry before you shut down BOINC to edit your app_info.xml

I hope this helps to clarify things for you.


. . It does until I open the files. It makes sense when you say it simply but ... well if the app attaches the ver number to the listing why does it trash everything if they disagree? I don't know enough of the process to feel confident.

. . Thanks for the input.

Stephen

?
ID: 1933997 · Report as offensive
Stephen "Heretic" Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 3653
Credit: 80,718,960
RAC: 163,603
Australia
Message 1934003 - Posted: 6 May 2018, 16:03:03 UTC - in response to Message 1933940.  
Last modified: 6 May 2018, 16:08:33 UTC

It makes life much easier if you stick to the stock release version numbers shown on the Applications page


. . Thank you Richard, that fills in a lot of the blanks.

. . Also if I follow your pattern and have two entries for r3584, one saying version 820 and one saying 822 then it should not trash existing WU is that correct?

Stephen

. .
ID: 1934003 · Report as offensive
Profile Brent Norman Special Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2325
Credit: 254,493,465
RAC: 611,339
Canada
Message 1934024 - Posted: 6 May 2018, 18:42:15 UTC - in response to Message 1934003.  
Last modified: 6 May 2018, 18:44:06 UTC

. . Also if I follow your pattern and have two entries for r3584, one saying version 820 and one saying 822 then it should not trash existing WU is that correct?
Yes, IF you are running 8.22

Just look at what you are running and delete all the other numbers - for the same device.

HINT: Add carriage returns in the app_info between the sections, and app versions, it makes it so much easier to read!!
Put 5 or 6 in between AP and MB sections and 1 for each app version, and it all becomes much cleaner to read.

PS: If you think the NVIDIA section is long, you should look at an ATI one - geesh!

EDIT: I have often wondered what would happen if # comments were placed within the file for notes ....
ID: 1934024 · Report as offensive
Stephen "Heretic" Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 3653
Credit: 80,718,960
RAC: 163,603
Australia
Message 1934069 - Posted: 6 May 2018, 23:33:55 UTC - in response to Message 1934024.  

. . Also if I follow your pattern and have two entries for r3584, one saying version 820 and one saying 822 then it should not trash existing WU is that correct?
Yes, IF you are running 8.22

Just look at what you are running and delete all the other numbers - for the same device.

HINT: Add carriage returns in the app_info between the sections, and app versions, it makes it so much easier to read!!
Put 5 or 6 in between AP and MB sections and 1 for each app version, and it all becomes much cleaner to read.

PS: If you think the NVIDIA section is long, you should look at an ATI one - geesh!

EDIT: I have often wondered what would happen if # comments were placed within the file for notes ....


. . Now someone who is on the same wavelength as me .... :)

. . I wanted to add CRs but I was afraid BOINC might not handle it and I was sure adding #'s to render sections as comments would have crashed it out horrendously.

. . This why I am working at the aistub level, more wiggle room while I am working ...

Stephen

:)
ID: 1934069 · Report as offensive
Stephen "Heretic" Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 3653
Credit: 80,718,960
RAC: 163,603
Australia
Message 1934753 - Posted: 10 May 2018, 1:38:36 UTC
Last modified: 10 May 2018, 1:58:01 UTC

. . OK, I'm a git I guess but the last attempt trashed all the GPu Tasks again.

. . Here are the aistubs I am using

MB-CPU-AVX (this is working)

<app_info> 


    <app>
        <name>setiathome_v8</name>
    </app>


    <file_info>
        <name>MB8_win_x64_AVX_VS2010_r3330.exe</name>
        <executable/>
    </file_info>
    <file_info>
        <name>libfftw3f-3-3-4_x64.dll</name>
	<executable/>
    </file_info>
    <file_info>
 	<name>mb_cmdline_win_x64_AVX_VS2010.txt</name>
    </file_info>


    <app_version>
        <app_name>setiathome_v8</app_name>
        <version_num>800</version_num>
        <platform>windows_x86_64</platform>
        <api_version>7.5.0</api_version>
        <file_ref>
            <file_name>MB8_win_x64_AVX_VS2010_r3330.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>libfftw3f-3-3-4_x64.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>mb_cmdline_win_x64_AVX_VS2010.txt</file_name>
            <open_name>mb_cmdline.txt</open_name>
        </file_ref>
    </app_version>


 </app_info>


AP-GPU-OpenCL (Don't know if working, no AP tasks ATM)

<app_info>


    <app>
        <name>astropulse_v7</name>
    </app>

    <file_info>
        <name>AP7_win_x86_SSE2_OpenCL_NV_r2887.exe</name>
        <executable/>
    </file_info>
    <file_info>
        <name>libfftw3f-3-3-4_x86.dll</name>
        <executable/>
    </file_info>
    <file_info>
        <name>AstroPulse_Kernels_r2887.cl</name>
    </file_info>
    <file_info>
        <name>ap_cmdline_win_x86_SSE2_OpenCL_NV.txt</name>
    </file_info>

    <app_version>
        <app_name>astropulse_v7</app_name>
        <version_num>710</version_num>
        <platform>windows_x86_64</platform>
        <avg_ncpus>0.04</avg_ncpus>
        <max_ncpus>0.2</max_ncpus>
        <plan_class>opencl_nvidia_100</plan_class>
       	<cmdline></cmdline>
        <coproc>
            <type>CUDA</type>
            <count>1</count>
        </coproc>
        <file_ref>
            <file_name>AP7_win_x86_SSE2_OpenCL_NV_r2887.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>libfftw3f-3-3-4_x86.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>AstroPulse_Kernels_r2887.cl</file_name>
        </file_ref>
        <file_ref>
            <file_name>ap_cmdline_win_x86_SSE2_OpenCL_NV.txt</file_name>
            <open_name>ap_cmdline.txt</open_name>
        </file_ref>
    </app_version>


</app_info>



MB-GPU-SoG r3584 (definitely not working)

<app_info> 

    <app>
        <name>setiathome_v8</name>
    </app>
 
   <file_info>
        <name>MB8_win_x86_SSE3_OpenCL_NV_SoG_r3584.exe</name>
        <executable/>
    </file_info>
    <file_info>
        <name>libfftw3f-3-3-4_x86.dll</name>
	<executable/>
    </file_info>
    <file_info>
        <name>MultiBeam_Kernels_r3584.cl</name>
    </file_info>
    <file_info>
        <name>mb_cmdline_win_x86_SSE3_OpenCL_NV.txt</name>
    </file_info>

    <app_version>
        <app_name>setiathome_v8</app_name>
        <version_num>820</version_num>
        <platform>windows_intelx86</platform>
        <avg_ncpus>0.04</avg_ncpus>
        <max_ncpus>0.2</max_ncpus>
        <plan_class>opencl_nvidia_SoG</plan_class>
        	<cmdline></cmdline>	
	<coproc>
            <type>CUDA</type>
            <count>1</count>
        </coproc>
        <file_ref>
            <file_name>MB8_win_x86_SSE3_OpenCL_NV_SoG_r3584.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>libfftw3f-3-3-4_x86.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>mb_cmdline_win_x86_SSE3_OpenCL_NV.txt</file_name>
            <open_name>mb_cmdline.txt</open_name>
        </file_ref>
    </app_version>
 
   <app_version>
        <app_name>setiathome_v8</app_name>
        <version_num>822</version_num>
        <platform>windows_intelx86</platform>
        <avg_ncpus>0.04</avg_ncpus>
        <max_ncpus>0.2</max_ncpus>
        <plan_class>opencl_nvidia_SoG</plan_class>
        	<cmdline></cmdline>	
	<coproc>
            <type>CUDA</type>
            <count>1</count>
        </coproc>
        <file_ref>
            <file_name>MB8_win_x86_SSE3_OpenCL_NV_SoG_r3584.exe</file_name>
            <main_program/>
        </file_ref>
        <file_ref>
            <file_name>libfftw3f-3-3-4_x86.dll</file_name>
        </file_ref>
        <file_ref>
            <file_name>mb_cmdline_win_x86_SSE3_OpenCL_NV.txt</file_name>
            <open_name>mb_cmdline.txt</open_name>
        </file_ref>
    </app_version>

</app_info>


. . Can anyone see where I have stuffed up? I cannot.

<edit>
. . I just reviewed the message to make sure it was formatted OK. Please let it not be this. It looks like there is a space missing at the beginning of the <file_info> section and another at the second <app_version> section. Could this be causing it to fail? I went back to the aistub to confirm this and they were missing, so I added them. But in reviewing the structure to make sure there were no other similar errors I discovered several lines were aligned using <tab> chars not spaces like 90% of the other lines. I have corrected this as well so they are now all coherent, but could these discrepancies be the cause of the problem??


Stephen

? ?
ID: 1934753 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 5331
Credit: 356,930,717
RAC: 1,089,560
United States
Message 1934759 - Posted: 10 May 2018, 2:12:29 UTC - in response to Message 1934753.  
Last modified: 10 May 2018, 2:13:30 UTC

Make your mind up.

Decide what version number you want to assign to SoG tasks and remove the entry for the other version. There needs to be only one version called out for each task type and application. The fact that you labelled SoG tasks with both 820 and 822 version numbers is why you threw your cache away.

Whatever version label for tasks you have in your cache that IS NOT the version number you have specified in the app_info is why you threw your cache away.
Seti@Home classic workunits:20,676 CPU time:74,226 hours
ID: 1934759 · Report as offensive
Stephen "Heretic" Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 3653
Credit: 80,718,960
RAC: 163,603
Australia
Message 1934770 - Posted: 10 May 2018, 4:36:58 UTC - in response to Message 1934759.  

Make your mind up.

Decide what version number you want to assign to SoG tasks and remove the entry for the other version. There needs to be only one version called out for each task type and application. The fact that you labelled SoG tasks with both 820 and 822 version numbers is why you threw your cache away.

Whatever version label for tasks you have in your cache that IS NOT the version number you have specified in the app_info is why you threw your cache away.


. . Well that is contrary to what Richard has done with the aistub for Lunatics. He has an entry for every conceivable version number all pointing to the current app selected for use. That is how he prevents the cache being ghosted (as I understand it anyway). The problem occurs when there is NO entry matching the version number of the tasks listed in BOINC manager. This is why I included entries to cover both the version number 820 to NOT ghost the cached tasks, and the version number 822 to set the new downloads.

. . I think it is a point which needs to be cleared up, as well as my question about syntax and the effects of extra or missing spaces and the use of TAB cahracters. I wonder what the definitive source would be on that. Jason isn't here anymore so I guess that leaves Richard.

Stephen

? ?
ID: 1934770 · Report as offensive
Profile Brent Norman Special Project $250 donor
Volunteer tester

Send message
Joined: 1 Dec 99
Posts: 2325
Credit: 254,493,465
RAC: 611,339
Canada
Message 1934772 - Posted: 10 May 2018, 5:02:28 UTC - in response to Message 1934770.  

.. and the use of TAB cahracters
I would say no to TABs!

I think I see the problem ... are these the sections you have (intact) in the file? I mean paste them together and it is complete?
ID: 1934772 · Report as offensive
Stephen "Heretic" Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 3653
Credit: 80,718,960
RAC: 163,603
Australia
Message 1934774 - Posted: 10 May 2018, 5:37:36 UTC - in response to Message 1934772.  

.. and the use of TAB cahracters
I would say no to TABs!

I think I see the problem ... are these the sections you have (intact) in the file? I mean paste them together and it is complete?


. . Yes, they are aistub files. I decided to work with that format to keep each section separate while working on it.

. . Then I copy them to the project folder and run aimerge.

. . I was doubtful that a space or 2 would bring it down but I wasn't sure, but TAB's seemed like a no-no to me too.

Stephen

. .
ID: 1934774 · Report as offensive
1 · 2 · 3 · Next

Message boards : Number crunching : app_info.xml, aistubs and other confusing things


 
©2018 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.