[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: hdparsim




Hi Mark,

    That sounds like a good solution. I'll go ahead and remove group 
write from the 2009-01-06 release on the CUE for now. If I recall, you 
had already set up a group account at some point that we could use. Can 
you remind me again how to log into it?

Regards,
-David

Mark M. Ito wrote:
> David,
>
> I think the ultimate solution would be to have a librarian account and 
> make the files writable only by the owner, i. e., the librarian. Then 
> more than one person could be "librarian" at any one time. Then when 
> you log in as librarian, you know you are possibly modifying public 
> stuff.
>
> In this particular case for this particular set of directories, for 
> now, I would go ahead and exclude group write. I don't think that that 
> will cut off too many efforts at present. Whereas as lot of us may 
> start dumping stuff in there if this is not done.
>
>  -- Mark
>
> P. S. Can I forward this conversation to the group?
>
> David Lawrence wrote:
>>
>> Yes, they should be, but I'm frankly not 100% sure how. What I'd like 
>> is to make them read-only, but still allow halld group members to 
>> change the permissions if needed to allow maintenance (e.g. compile 
>> binaries for another OS). Do we have to set the SUID bit to get this 
>> behavior? Any other ideas?
>>
>> I have gone ahead and copied the snapshot of the whole 
>> release-2009-01-06 directory from 2 nights ago back into the builds 
>> area and renamed the one you installed to 
>> "release-2009-01-06.corrupted" for now. We can discuss the long-term 
>> preventative solution tomorrow.
>>
>> Regards,
>> -David
>>
>> Mark M. Ito wrote:
>>> David,
>>>
>>> Whoops. Looks like I am writing my plugin to the public build:
>>>
>>> install -p .lib/Linux_Fedora8-i686-gcc4.1.2/hdparsim.so 
>>> /group/halld/Software/builds/release-2009-01-06/lib/Linux_Fedora8-i686-gcc4.1.2 
>>>
>>>
>>> I'll delete that.
>>>
>>> Shouldn't those directories be protected against this?
>>>
>>>  -- Mark
>>>
>>>
>>> Mark M. Ito wrote:
>>>> Next error:
>>>>
>>>> marki@ifarml5> hd_root --plugin=hdparsim -PDEFTAG:DPhoton=HDParSim 
>>>> --plugin=photoneff_hists output.hddm
>>>> Created JCalibration object of type: JCalibrationFile
>>>> Generated via: fallback creation of JCalibrationFile
>>>> Runs: requested=1 found=1 Validity range=1-10000
>>>> URL: file:///group/halld/calib
>>>> context: default
>>>> Reading Magnetic field map from Magnets/Solenoid/solenoid_1500 ...
>>>> 32481 entries found ( Nx=81 Ny=1 Nz=401 )  at 0xa0c27a8
>>>> Read 840 values from FDC/lorentz_deflections in calibDB
>>>>   lorentz_deflections columns (alphabetical): bx bz nx nz x z
>>>> Initializing plugin 
>>>> "/scratch/marki/hdp/lib/Linux_Fedora8-i686-gcc4.1.2/hdparsim.so" ...
>>>>
>>>> ***ERROR : Couldn't find plugin "photoneff_hists"!***
>>>> ***        To see paths checked, set JANA_PRINT_PLUGIN_PATHS env. 
>>>> var. and re-run
>>>>
>>>> I don't have JANA_PRINT_PLUGIN_PATHS defined. Should I?
>>>>
>>>>
>>>> David Lawrence wrote:
>>>>>
>>>>> Hi Mark,
>>>>>
>>>>>    You'll want to check out the code from:
>>>>>
>>>>> https://halldsvn.jlab.org/repos/trunk/src/programs/Simulation/plugins/hdparsim 
>>>>>
>>>>>
>>>>> Let me know if you have trouble compiling it.
>>>>>
>>>>> You can test the photon part first with the following:
>>>>>
>>>>> genphoton -M 10000 -Pmin 0.05 -Pmax 8 -Thetamin 0.5 -Thetamax 125 
>>>>> -o genphoton.ascii
>>>>> genr8_2_hddm genphoton.ascii
>>>>> hd_root --plugin=hdparsim -PDEFTAG:DPhoton=HDParSim 
>>>>> --plugin=photoneff_hists output.hddm
>>>>>
>>>>> Of course, you'll need to have the /genphoton/, /genr8/_/2_hddm/ 
>>>>> and /hd/_/root/ programs compiled. I think only /genphoton/ is not 
>>>>> compiled by default. It lives in the src/programs/Simulation 
>>>>> directory. (If it complains about "uint" when compiling, you'll 
>>>>> need to do a global search and replace of "unsigned int" for 
>>>>> "uint". That fix was only committed this weekend to the trunk.)
>>>>>
>>>>> Regards,
>>>>> -David
>>>>>
>>>>> Mark M. Ito wrote:
>>>>>> What should I check out first, if anything?
>>>>>>
>>>>>> So far I've only done:
>>>>>>
>>>>>>  source /group/halld/Software/builds/release-2009-01-06/setenv.csh
>>>>>>
>>>>>> by following the getting started on JLab wiki page.
>>>>>>
>>>>>
>>>>
>>>>
>>>
>>
>

-- 

------------------------------------------------------------------------
 David Lawrence Ph.D.
 Staff Scientist                 Office: (757)269-5567   [[[  [   [ [       
 Jefferson Lab                   Pager:  (757)584-5567   [  [ [ [ [ [   
 http://www.jlab.org/~davidl     davidl@jlab.org         [[[  [[ [[ [[[
------------------------------------------------------------------------