Core Minutes 6/14/2005
Present: Joanne Bogart, Toby Burnett, David Chamont,
James Chiang,
Richard Dubois, Dan Flath, Warren Focke,
Tom Glanzman, Navid Golpayegani,
Tony Johnson, Heather Kelly, Matt Langston, Julie McEnery,
Chuck Patterson, Igor Pavlin, James Peachey, Dirk Petry, Leon Rochester
- Data handling: (Tony)
- JIRA/Confluence administration Matt and Tony plan to take down
JIRA and Confluence to install newer versions. This may address the problems
we've been seeing, requiring us to restart these applications every couple
weeks. If there are no objections, they'll do this tomorrow (Wednesday)
from 5-7 PM.
- Naming Matt has proposed adopting a naming convention so that
Java code is easily distinguished from C++ in our repository
- Archiving Dan and Navid decided to manually archive existing
data, then set up automatic archiving for new data. Dan has had trouble
getting certain jobs to run under certain accounts.
- Schedule Tony is working on a master timeline; not
ready quite yet.
- Data catalog Julie has asked that something be ready for
DC2. This should be possible. Input is requested on just what's needed,
particularly from Julie and Richard.
- ROOT Review: (Heather) It took place as
planned yesterday. Heather, Tom, Anders, Matt and Richard attended from
our side, as did BaBarians Peter Elmer, Eric Charles and Stephen Gowdy.
Each group described its use of ROOT. Main point that Heather came away
with was that we ought to go over our classes, looking into ways to make
things more compact, especially tracker recon. Another item that
should perhaps be pursued is a method they have of keeping related files
logically linked. Finally, the consensus is that it should not be a big
deal to move to ROOT 5; it's billed as being backwards-compatible with
ROOT 4.
- Navid report:
(Navid)
- Windows RM builds are still suspended pending installation of
new disks. Toby asked if it would be possible to, for now, resume just
the creation of Windows source zip files. Navid said it should be possible,
since they go to an nfs disk, but might be a significant amount of work.
He'll look into it. [Instead, since there is no ETA on the installation of new
disks, he's decided to just re-enable Windows RM. ed.]
- Archiver has been rewritten; it now handle many
concurrent requests.
- Tag collector The new version, integrated with RM, is up. He
fixed some bugs; none have been reported since. Soon he expects to
pull the plug on the old tag collector.
- Automated Systests were invoked for the last EM tag, but
there was a problem because the wrong tag was picked for reference. Will
try again with the next EM tag.
- When to bite the Apple™
Richard put together a
confluence
page about the possibility and implications of supporting Mac
as a third platform. There was an extended discussion, touching on at least
the following points.
- We need a proof in principle: a successful build of GlastRelease or
ScienceTools. Some work is in progress on this by Navid and Pat. James
will have a new Mac soon and will join the fray.
- Endianness. Existing Macs are big-endian, but this may change. Total
amount of work might be reduced somewhat by waiting for the change to
happen, but most of what we need to learn about support doesn't depend
on endianness.
- Will the Workbook need to support a new browser? (Safari). Mac end-users
(as opposed to developers) are likely to use it as their principle browser so
yes, it probably should be supported.
- I & T: (Anders)
6-tower testing will probably start tomorrow. Calibrations will be the next
week, with data-taking to follow later that week or at the start of the
next week. Also will recalibrate 4 towers since there were problems
last time.
- EM tags: (Heather)
There was a tag late last week and will probably be another one today. Don't
be surprised if it looks different. Form of EM tags is changing to make
it easier for RM to organize them sensibly.
- Calibration: (Joanne)
reorganization of rdbModel is the main item in
progress at the moment. The xml component has been revamped to include
more information. Now have to finish propagating changes and additions through
various layers of code, expected to take some weeks.
- CalRecon status: (David)
-
The head of RootIo now requires the head of the new package RootConvert,
where we will move progressively the low level conversion code, to be shared
between RootIo and RootCnvSvc (as planned a long time ago).
- In CalRecon head, the new CalCluster has been made persistent, and Tracy
has started to import the mip finding code. Yet it seems some files to be
added have been forgotten. Under investigation.
- DC2 status: (Julie) She'll talk about it
at some length on Thursday. Right now she needs to know who is coming
to the workshop.
- GLAST Citrix: (Matt) It still doesn't let in
Linux users. SCS thought they had already confirmed this was ok, but they
hadn't and it isn't. It probably has their attention now.
- Announcements: (Richard)
- A 2-1/2 day
DOE
HEP Program Review is in progress at SLAC.
- Robin Staffin from the DOE Office of High Energy Physics spoke at
the SLAC Colloquium yesterday. It was not a cheery talk.
J. Bogart, Last Modified:
01-Jun-2010 15:45:47 -0700