Software meeting: usually every Friday, 10:30 o'clock
Minutes of Calice Software Meetings
- 15 May, 2009
- 8 May, 2009
- 16 May, 2008
- 4 April, 2008
- Minutes of Calice Software Meeting 15st February 2008
- Minutes of Calice Software Meeting 8st February 2008
- Minutes of Calice Software Meeting 1st February 2008
- Minutes of Calice Software Meeting 25th January 2008
- Minutes of Calice Software Meeting 18th January 2008
- Minutes of Calice Software Meeting 21th December 2007
- Minutes of Calice Software Meeting 17. Dezember 2007
- Minutes of Calice Software Meeting 7. Dezember 2007
1. 15 May, 2009
many changes done by the data quality group in the HitWriteEngine, which should be propagated to the afs software, since also other people want to use it
to be discussed with Joergen how this can be done (codes may not be compatible with the actual RootTreeWriter version)
- additional class to analyze the data quality root files - to be put in a separate git repository
- all the useful information gained with the data quality studies should be made public: Andrea will set up a flcwiki page to start with
- list of noisy cells needed (Serguei can provide it)
- next step: simple processor to be called during the reconstruction, which removes these cells (this might slow down the reconstruction additionally, but this is the fastest solution for the moment)
- I'll go for shifts at Fermilab, so no meetings in the next 2 weeks
2. 8 May, 2009
- Angela, Niels, Beni: Software changes related to applying of temperature corrections will be propagated at the beginning of next week to cvs, such that other groups (e.g. Muenchen) can use this
- Beni: check of data base folders containing gain values
3. 16 May, 2008
There is a new track finder written by Nicola and Sebastian (it is the old processor of Nicola, used in his analysis, but rewritten in a modular approach). It will go, once we have a new release, in our package with Hamburg programs, calice_hh_proc.
- Still some open issues before new software release:
- in the HCAL reconstruction code - Niels is trying to solve them
- in the Tail Catcher reconstruction (Niels will check if this has to do with our recent HCAL changes, or if it was wrong from the beginning).
4. 4 April, 2008
We've decided to try to get rid of magic numbers in our software packages. The idea is to put all the constants in a header file, and all the processor should take the constants from there. This will be done step by step, so if you stumble upon such a number, please send an email to firstname.lastname@example.org with the number and the location.
- From the list of things we miss: a working event disply...
5. Minutes of Calice Software Meeting 15st February 2008
- Agreement on using doxygen for documention of processors
6. Minutes of Calice Software Meeting 8st February 2008
Announcement of conddbextractor: simple tool to extract conditions data from the database
conddbextractor and mcfileimitator can be found at: flc45:/data/richters/calice-devel/(conddbextractor|mcfileimitator) (git repositories), if you need any assistance, please ask via e-mail (email@example.com)
7. Minutes of Calice Software Meeting 1st February 2008
7.1. Status of new tag/release
- Roman's deadline prolonged with a week (15.02.)
8. Minutes of Calice Software Meeting 25th January 2008
8.1. Status of new tag/release
- intend to release a new pro_test version of our software
comment: the release is now on the afs, just do flcini caliceSoftpro_test . Then you can use the executable caliceMarlin which knows about the newly introduced processors. Feel free to test it, and report any errors, suggestions, comments etc.
what is new in this release: calice_sim package adapted by Sebastian; contains ganging and digitisation processors (digitisation of the fine modules is ok, problems with the coarse modules). The package contains also some old code for the ECAL digitisation (perhaps Roman's task to get the latest software for ECAL and Tail Catcher digitisation?).
we have a calic_hh_proc package, which will contain not only Marlin processors, but any piece of code which is useful to us (feel free to contribute). Sebastian is going to put there a Monte Carlo file imitator, which is a C++ program able of reading a configuration file and of writing as output an lcio file (useful for debugging purposes)
- the official deadline for a new release is 8th February, so we set our deadline to 6th of February, so that people have time to come with comments, and Roman has time to integrate our changes
the RootTreeWriter package will contain new engines from Erika and Nanda
8.2. MC production status (Oli)
- October 2006 data, pi minus, energies of 6, 8, 10, 12, 15, 18, 20, for the full detector chain (ECAL + HCAL + Tail Catcher), for LHEP and LHEP_BERT physics lists are simulated
- for more details about these files, have a look at
http://www-flc.desy.de/hcal/tilehcal/simulationDB/index.html (Date of production: 2008)
- plans to simulate also electrons and positrons for the same run period, and later for 2007
9. Minutes of Calice Software Meeting 18th January 2008
9.1. Calice event display
- we are currently missing a working Calice event display.
- In the past, there was an old event display in calice_gui, written by Goetz, which used the cell index and mapping classes.
- Another possibility would be CED (C Event Display) - old C implementation of an event display.
In MarlinUtil, there is MarlinCED, but which does not draw the wire frame.
- One idea would be to use this MarlinCED event display, and create a stand alone processor for drawing the wire frame. Niels will talk to Frank about this.
9.2. HCAL specific package
- HCAL specific software (e.g. as the one used by Beni for the gain analysis, and by Alex and Nils for the temperature corrections, etc.) should go to the already created package calice_hh_proc in the afs git repository
- Note that this package may contain any useful piece of code, not only Marlin processors
9.3. Bug report
- Sebastian: in calice_torso, 'libname' should be changed to 'lib1name', or something else, since the 'libname' gets overwritten by the FindROOT.cmake module.
9.4. Grid scripts for Monte Carlo production
- Oli started Monte Carlo production on the grid
- At some point, the resulting scripts should be stored on an official place, such that people can use them (there is a CVS repository of Ivan, who's working on the MC production for the whole detector, but Oli's scripts diverged from the Ivan's ones)
- Probably we will need in the future somebody to take care of this.
9.5. Status of the next PRO release
- There is still work on the digitization chain (mapping problem)
- Niels and Sebastian will commit their changes by Tuesday afternoon (even if there is still a bug with the coarse modules)
10. Minutes of Calice Software Meeting 21th December 2007
Enjoy the holidays, dream about snow and leave the software for the next year
11. Minutes of Calice Software Meeting 17. Dezember 2007
Erika volunteered to think about a list of useful tree/n-tuple variables and will have a look in writing engines for the RootTreeWriter
comment: The Doxygen documentation concerning the installation is currently broken, if you don't use the latest doxygen version (SL3/SL4 does not), I will try to fix this -- JörgenSamson 2007-12-14 15:55:19
11.2. Preparation of the next PRO version
- the digitalization chain is now complete in the new software framework and ready for testing
- a rather long lived new version of the ilcsoft core software should become available in it's group afs (v01-03-devel)
comment Sebastian: Why "devel" and not v01-03-nothing?
- important features: ROOT5.16
- Marlin with conddbmysql support
we will install a new "pro_test" version, which contains the new digitization chain (Sebastian), the EventSelector (Jörgen/Beni) (and possibly the ConditionMapping processor (Beni) )
Beni's ConditionMapping processor is already there (affects calice_userlib and calice_reco)
- The steps needed to do this installation and the "upload" procedure concerning the git repository will be documented in the wiki (Sebastian + Jörgen)
11.3. Problems, FAQ and Bugs
Please have a look at the FAQs and write down Answers if there are questions you are asked more than twice
- Please report bugs, you found and post solutions for them (maybe a link to a branch in your personal git repo) to prevent other users to stumble on the same problem
12. Minutes of Calice Software Meeting 7. Dezember 2007
offline comments by Niels
12.1. calice_torso CMakeLists.txt conflicts with ILCSoft core modules
New FindROOT.cmake confilicts with calice_torso 's CMakeLists.txt because both use the cmake variable libname
Can be fixed by moving the FindPackage(ROOT) or renaming the cmake variable libname in calice_torso's CMakeLists or FindROOT.cmake (Niels?)
comment Niels: change calice_torso!
12.2. Get Code into pro branch
changes in our pro branch of calice software need approval of (at least?) calice software meeting
comment Niels: I like the idea of a review for updates or newcomers. However, I would release the requirements for including old code from the old hcalsoft collection. Angela's o.k. should be enough here, if the code is tested to compile and run with the new setup. This applies to everything mentioned below.
to get something into the pro branch you have to provide a git branch which is a clean direct descendant of the official pro branch
12.3. new software in for pro
candidates for new software in pro:
- MappingIConditions (calice_userlib, calice_reco?)
- FastMappingMCProcessor (calice_sim)
OverlayProcessor (=Event Merge) (?)
EventSelector (calice_hh_proc) (!! ask Roman for a place to hold processors like this !!)
comment Niels: Roman already signaled that there will be a calice_analysis package comming up. Let's create a calice_analysis git repository and set up the CMake environment - there are chances we push this one through to the CVS. For the moment, fill it with things like the EventSelector, the OverlayProcessor and alike.
- Niels (added offline)
- started complete re-write of reco and digi code (userlib, raw2calohit and sim). First running version for tests expected end of year.
- have similar plans for the mapping complex
12.4. handling pro and stable
- stable will only be touched, when Roman creates a new tag in the official repository
- a mail will be sent to the users on the "CALYSIS" list, when pro is changed
comment Sebastian: I can't find the list on https://lists.desy.de/sympa/lists ?
12.5. Coding conventions
- Sebastian will place a link to Frank's coding guide lines into the wiki
comment Niels: there is already a talk linked on http://www-flc.desy.de/flc/flcwiki/HCAL_Documentation (behind external link 'the general documentation'). Is there more?
comment Sebastian: Now, as far as I know there isn't. I created the page http://www-flc.desy.de/flc/flcwiki/Software_Guidelines for our guidelines.
12.6. sofware meeting
- short weekly meeting on friday at 14:00
- Beni will reserve the new seminar room in the container for this slot.
The functionality of the "OverlayProcessor" could be covered by the "JoinEvents" processer of the "Overlay" package in the core software. If not this processor can be put into the "Overlay" package.
comment Niels: this is somewhat delicate, since we require merging hit amplitudes. If I remember correctly, the Overlay package aims at tools to add events and to merge collections (i.e. daisy-chain collections of the same name). I fear that a general implementation of the next level, e.g. merging hits with the same ID, has too many degrees of freedom. IMHO there's no problem with a dedicated CALICE implementation for this special task.
-- JörgenSamson 2007-12-07 15:29:19