I) Library Versions A) Version 1.1 released as announced in 12/10/99 Telecon on 1/4/00. B) A few bugs fixed and released in Rev 1.1a - FSeries bugs fixed in Power() and operator= - FrWriter no longer brain dead (new copy functions, specify time span in for frame in FrameWriter::buildFrame()) - A few variable types assigned for compilation under linux. - Bug fixed in TPlot, Bode and Spectrum macros (data shifted 1 bin) - FrameCPP build kludged (see below) II) Installation Problems A) New web page - list of packages/versions needed for compiling software - Step-by-step procedures for Solaris installation provided by Ed Daw. - Pointers to all distribution versions. B) Continue to fix source to fix compilation incompatibilities as they are discovered (thanks to E. Daw and A. Ottewill for reporting bugs and compilation glitches as they are noticed). C) Unfortunately, LDAS started to remodel the FrameCPP build procedure (autoconfig/automake). Result: Latest fixes to FrameCPP not included. Remodeling should be finished today (2/14/00). III) LHO Status A) Bug found and fixed in Frame Broadcaster - No data errors since 1/14/00. - All missing frames seem consistent with DAQ start/stops. B) Fortress set up for writing designer datasets - Fortress machine+disks swapped with stone - now has two 70GB RAID arrays - Tape robot connected to fortress (hand mounting - robot S/W not working on Solaris 2.7) - Receives data from frame broadcaster - FrWrite program selects channel subset, writes out frames. - Now in use by LHO scientists: Mini-NDS server data subsets to Data Viewer and GDS S/W. C) Minor operational changes on DMT machines - Good reliablity. 4) Future developments A) Data Conditioning API library - LDAS & LSC members (S. Finn, J. Romano, P. Charlton) are putting together signal processing library for use in LDAS Data Conditioning. - May be able to incorporate into DMT structure - Modified lowest level data vectors to use STL containers (e.g. valarray) to facilitate connection with DataAPI library. B) Communication with Background processes (model discussed by D. Sigg and JGZ) 1) Small data server incorporated into monitors - UDP socket based communication - User defines objects to be served (name, type, data pointer) - Serving requests transparent to monitor code (handled by the base class). 2) Scripted display manager - Standard displays defined by parameter file or scripts (implementation not defined yet, may evolve) - Script tells display manager which data to get (process.dataname) and how it is to be formatted on a screen. 3) Data Display - Probably will be based on Daniel's GDS data displayer.) 4) Status - Data communication package mostly ready (translation between different platforms only partially finished, Server API missing) - Asynchronous data handling partially implemented. C) Other stuff 1) TSeries::getMinimum(), TSeries::getMaximum(), TSeries::getNLess(), TSeries::getNGreater() (or would an iterator + STL be sufficient). 2) Test macros for various DMT components. 3) Incorporate Alex DAQFrame code (Fast input of partial frames - assumes all frames alike).