ROOT Weekly Meeting

Friday 22 Jul 2005, 11:00 → 12:15 Europe/Zurich

32/1-A24 (CERN)

Rene Brun (CERN)

11:00 → 12:15 Tour de Table

Minutes of ROOT weekly meeting 22 July 2005 =========================================== Rene ==== Proposal to change the frequency of meetings to bi-weekly meetings. Produced the LCG quaterly report (sorry if I did not consult the work package managers) "Defended" ROOT position at yesterday AForum. Request from CMS and Atlas in particular to remove all possible dependencies from CINT and TObject (Linear algebra in particular). Keep plug-in manager in SEAL. Develop in ROOT the new plug-in manager as proposed by Fons. 3 experiments do not accept the work plan otherwise. Agreement that ROOT will support the existing SEAL libraries for some time (until experiments decide that it is deprecated). Quite demotivating meeting. As usual help from users is expected. Discussion with LBL about CopyRights with TBitmapIndex. Philippe ======== New Fermilab Root Team member: Paul Russo New member function TDirectory::GetDirectory. This can be used (amongst other things) to test for the existance of a directory without cd-ing to it. CINT's source files all (but one) moved to C++ but still exposing an extern "C" interface. Guenter ======= The Resource discovery (Apple Bonjour based) is now working. Waiting more info from IT to test it on CERN LAN in general. Started to work on extensions of Xrootd OLBD system for Proof. Jakub ===== working on web interface for PROOF based on Carrot. Difficulties contacting Valeriy O. Bruno ===== Continued to compare results on the Qt layer to the ones on the X11 layer. Several bugs and not documented features and limitations of the Qt layer were identified and communicated to Valeri Fine. On the Graphics level documented the following feature: * Line Style + Line Styles don't look the same as on the X11 layer. This is a Qt constrain. And identified the following problems: * Text + The Text Vertical Alignment seem not to work properly. * Contours + In the $ROOTSYS/tutorials/ContourList.C example, the contours are not drawn. On the GUI interface, continued to exchange impressions about the Scroll Bars appearance: * Scroll Bars + Scroll Bars look strange with strange colors and patterns. And closed the Color Widget problem, which is now solved and working properly: * Color Widgets [2] + Color selector dialog shows slightly different RGB values. For example 254,254,254 for white color, 254,254,0 for yellow, etc. Took reponsability of the Qt layer bug/feature documentation and actualization. This can now be viwed on the link: http://belbute.web.cern.ch/belbute/QtToDo.html Was envolved on the determination of the OpenGL (XFree86 version and the DRI (Direct Rendering Interface) which is packaged with it) problem which caused the machine to hang. Will still be working on this issue with Olivier Couet and Richard Maunder. Pere ==== Discussion about Savannah When to close a bug? Need to add new items and people assigned to tasks. What about the standalone mathcore installation? wait for Lorenzo Olivier ======= * Complete rewrite of TArrow::PaintArrow. The previous version used the pixel coordinates. This method generated many rounding errors. The visible effects on arrows were: 1) distorted arrows' heads 2) misplaced arrow compare to a TLine at the same coordinates. The new TArrow::PaintArrow version fixes these two problems. Instead of using pixel coordinates it uses "true NDC" coordinates. The "ROOT NDC coordinates" are not appropriate because X an Y axis are not normalized the same way, therefore angles are not kept during a rotation. * CMYK color model requested in EPS files. CMY and CMYK models are subtractive color models unlike RGB which is an additive. They are mainly use for printing purposes. CMY means Cyan Magenta Yellow to convert RGB to CMY it is enough to do: C=1-R, M=1-G and Y=1-B. CMYK has one more component K (black). The conversion from RGB to CMYK is: Double_t Black = TMath::Min(TMath::Min(1-Red,1-Green),1-Blue); Double_t Cyan = (1-Red-Black)/(1-Black); Double_t Magenta = (1-Green-Black)/(1-Black); Double_t Yellow = (1-Blue-Black)/(1-Black); CMYK add the black component which allows to have a better quality for black printing. PostScript support the CMYK model. I recently got a request to support this color model in ROOT. The preliminary tests I did seem to work but I would like to have confirmation from the user that it is really what he wants. I don't know any printers around requiring explicitly this color model, so it is difficult for me to tell the difference between a RGB file and a CMYK one. * Fixes in 2D errors painting: - Hoption.Error is now set by default to 2 only for 1D histograms - In Paint2DErrors: fH->GetBinCenter was used to get Y position. That's obviously wrong. Now fYaxis->GetBinCenter() is used. * Denis is making progress on the TStyle editor definition. We now are looking are the various sections to put in place and their content. * Timur is back at CERN for 2 weeks (arrived last Monday). He is working on the GL version of TPad. Yesterday he did big progress in that area being able to mix X11 and OpenGL drawing in the same window. Andras ====== Making progress with MathMore. Difficulties with rootcint should be fixed now in CVS (thanks Philippe). Here you can find the comparisons for the special and statistical functions of mathcore: http://seal.web.cern.ch/seal/root/comparisons/ComparisonROOTMathematica.html Anna ==== Improvements in TF1::GetMinimumX. Generalize this algorithm to TF2 and TF3. Using Minuit. Marek ===== Finished with Friend trees and indices for Proof. Wrote a specification of the GUI for Proof. Stefan ======= Investigating rootcint/makecint generating reflex compliant code. will get CVS rights for Reflex Gerri ===== o Commit of two rather big patches: 1. Re-organisation of the authentication-related code in such a way that is not any more part of libCore but confined into a new plugin library (libRootAuth) loaded by TSocket via plugin manager. Any reference to authentication related stuff has been dropped from TROOT, TProof, TSlave and TProofServ. 2. Import of new TXNetFile using the xrootd posix interface. Allows to get rid of several somewhat duplicated classes now mantained in the XrdClient module of XROOTD. o Preparation of the new PROOF structure: 1. Factorization of "technology"-related calls in TProof, TSlave and TProofServ. I am preparing implementations able to talk to both proofd and xproofd. This wil allow to have the present system always running and at the same time to start using the new system. 2. Cleanup of the new classes TXPSocket and XrdProofdProtocol . o Support for 'rootd' deployment in CERN central services 1. Following a nasty bug found by Fons in rootd, triggering the need of a re-deployment, I had some iterations with Jan Van Eldick (IT) to fine-tune the RPMs. Jan started the deployment of the new RPMs yesterday. Richard ======= Optimize the scene rebuild in GL. Not yet satisfactory. Problems with some GL implementations (hangs). Working on the clipping planes and box. Fons ==== Setting up Alice/Apple/Proof cluster Discussion about copyrights complexity.