[Clas_offline] New test build

[Nikolai Pivnyuk]

fsgen generator - in my my personal directory. 10,000 events were produced.
gsim and gpp - from "old" /group/clas/builds/bin/ directory.
gpp.bos was processed with "old" and "new" user_ana versions.
-------------------------------------------------------------------
user_ana interactively - from "old" /group/clas/builds/bin/ directory.
user_ana on batch farm - from "new"
/group/clas/builds/volatile_test_03_27_14/bin/ directory.
--------------------------------------------------------------------
On level of comparison of histograms (from "old" and "new" recsis ntuples)
- no differences are observed.
--------------------------------------------------------------------
The results of direct comparison of "old" and "new" recsis .bos files
are presented in the directory
/volatile/clas/clasg10/pivnyuk/g10/mc/builds_New_vs_Old

Please read Readme file and have a look through diff_new_old file .
I have no enough qualification to comment on diff_new_old file.

Regards,
Nikolai

> Thanks! I am asking Mac about what his criteria for testing should be for
> the new build to be successful.
> -Will
>
> On Mar 28, 2014, at 2:02 PM, Nikolai Pivnyuk <pivnyuk at jlab.org> wrote:
>
>> Will,
>> I have processed gpp.bos with "old" build (interactively) and with
>> "new" (test_03_27_14) one (on batch farm). The input gpp.bos file was
>> small, just 1000 events only.
>>
>> I have plot and compared a few histograms from from both recsis.nt "old"
>> and recsis.nt "new". No difference is observed.
>>
>> The closer look - next week.
>>
>> Regards,
>> Nikolai
>>
>>> Hello Everyone,
>>> At the CLAS offline/CLAS12 software meeting yesterday we discussed what
>>> to
>>> do with the CLAS6 software production build. We now have a test build
>>> in
>>> /group/clas/builds/volatile_test_03_27_14/ where only this build will
>>> exist for testing purposes temporarily. Those who were having trouble
>>> with
>>> the current production build will be responsible for testing the new
>>> build. This new build will be swapped to the current production
>>> location
>>> (/group/clas/builds/) after users have tested it. This volatile build
>>> should not be used for production purposes, it will be removed after
>>> the
>>> testing is finished or if no testing is done. If you have any
>>> questions,
>>> just let me know!
>>> -Will
>>>
>>> To use these test builds:
>>> source /group/clas/builds/volatile_test_03_27_14/environment.csh
>>> or
>>> source /group/clas/builds/volatile_test_03_27_14/environment.sh
>>>
>>>
>>> Note: Relevant minutes from the software meeting (Mac's minutes)
>>>
>>> NEW VERSION OF USER_ANA:
>>> - Nikolai Pivnyuk has been having problems running batch
>>> jobs with user_ana.  Will Phelps (the CLAS librarian) will
>>> release a new version in a test area.  A few people will
>>> be designated as 'testers' of the code.  If they report
>>> that all is ok within a couple of weeks, the new code will
>>> be official.  If it's not ok, back to the drawing boards.
>>> If the testers don't do anything, we don't do anything.
>>
>
>