-Lorenzo presented the progress with mathlibs.
Markus made the remark that it would be desirable to have
small data objects only for the physics vector classes
complemented by other classes with functions accepting
the data objects as arguments. Lorenzo will rethink along
these lines and will post an update once it is ready.
Once the document will be more mature, it will be posted
to roottalk to get more comments.
-Andras presented the progress in interfacing the version
of Minuit (was called GMinuit) by Matthias Winkler.
A hot discussion followed where Andras was pushed to present
a table with pros and cons of the new class.
It is likely that the new system will have to be substantially
restructed to simplify the API.
Is this new set of classes easier to maintain than the current TMinuit?
What are the added advantages?
-Anna presented the progress with the specialized new fitters
and her work in adding the new robust fitters. The TLinearFitter
shows impressive speedup compared to teh normal Minuit fits.
-Rene presented the extension of the Double32_t data type. He also
presented slides from Philippe showing some new features in the I/O
and trees including mention of problems still to be resolved.
-Jose presented the status of his work in adapting the RuleChecker
to the ROOT classes. Jose will upgrade his interface such that
it can run daily with results on a web page.
All elements in the generated table should be clickable.
-see what RN7 means
-show the list of violations for a given rule and package
-Round Table discussion
-Fons said a few words about the progress in merging the ROOT
and SEAL plugin-managers. He will present the status at the next
week team meeting.
-Richard reported about his progress with the 3-d viewers.
He should be able to show the first results in about one week.
-Markus reported about a discussion aiming at removing
gDirectory/gFile from TKey.
-Stefan is preparing documents and benchmarks comparing CINT
and Reflex dictionaries in view of the coming workshop in May.
He is using the TGeo classes and stressgeom.cxx as a benchmark.
-Guenter reported about new features in CRTL/C in PROOF and
his new UDP socket class.
-Olivier reported on some enhancements/bug fixes in some
drawing options.
-Gerri gave a short report on the use of xrootd as a base
for the control of the new PROOF. During this week PROOF
phone conference meeting, Gerri has shown a very impressive demo
of the new system that looks very promissing.
-Marek commented about the progress in implementing Tree features
like Tree friends in the context of PROOF.
-There is a lot of progress with PROOF since a few weeks.
Marek will present teh status of this work just before ACAT05
where he will give a talk.
Rene will be away next week. The weekly meeting will be run by Fons.