2003-06-16 ------ omnetpp-2.3 released ------ 2003-03-15 ------ omnetpp-2.3b2 released ------ 2003-02-15 ------ omnetpp-2.3b1 released ------ 2003-02-07 Andras Varga * MSVC .mak file replaced by Makefile.vc, generated by opp_nmakemake 2002-05-04 ------ omnetpp-2.2 released ------ 2002-04-07 Ulrich Kaage * Makefile.in (NEDCFLAGS): fixed substitution 2001-09-01 Andras Varga * const-related fixes 2001-03-10 Andras Varga * ------ omnetpp-2.1 released ------ * MSVC makefiles and project files now contain the OMNeT++ include and lib directories with absolute paths (previously they were relative paths, like ..\..\lib). This means the samples will compile even when they're moved to another location in the file system. * Borland C++ project file (*.ide) removed, Makefile.bc added (unsupported!) 2000-07-01 Andras Varga * ------ omnetpp-2.0p1 released ------ * ------ omnetpp-2.0 released ------ * calls to obsolete cQueue functions upgraded to new functions (insert(),pop()) 2000-06-02 Andras Varga * ------ omnetpp-2.0b5 released ------ * fix in MSVC makefile; warning eliminated in othermod.cc 2000-05-07 Andras Varga * ------ omnetpp-2.0b4 released ------ * char * --> const char * changes 2000-03-12 Andras Varga * readme expanded with advice on PVM setup 2000-01-25 Andras Varga * ------ omnetpp-2.0b3 released ------ * ssm_ifs.cc updated because of cArray interface change 1999-12-23 Andras Varga * BC3.1 support removed * rewritten msg->addPar( *new cPar(...)) lines (see change to cPar interface in the sim library) * minor cosmetics changes 1999-10-17 Andras Varga * Computer renamed to FDDIStation (clashed with Token's Computer when linking all sample simulations into one executable) 1999-10-10 Andras Varga * standardized indentation in ned files * fddi_net.ned split to fddi_nodes.ned, fddi_rings.ned, fddi_nets.ned * proper README written 1999-07-11 Andras Varga * ------ omnetpp-2.0b2 released ------ * ------ omnetpp-2.0b1 released ------ * bug in mystrlwr() [othermod.cc] fixed 1999-06-21 Andras Varga * all sources checked in under CVS * makefile.in modified for use with GNU Autoconf; also renamed to Makefile.in * version.log renamed to ChangeLog; some cosmetics on it (indents etc) * removed support for DOS with BC3.1 1999-06-21 Andras Varga older news follow this line FDDI80.arj * ? FDDI79.arj * simulation.error() calls corrected to error() FDDI78.arj * makefile renamed to makefile.in FDDI77.arj * added fddi.ide; doesn't run correctly under NT though (causes exception) [later discovered it was caused by the mystrlwr() bug, see Jul 11, 1999] FDDI76.arj * updated to the new histograms in sim161 FDDI74.arj * improved "display:" stuff in fddi_net.ned FDDI73.arj * `display:' phrases added to NED * bug fixed in othermods.h: FDDI_Address_Generator's constructor didn't initialize f * fddi_mac.cc: FDDI_MAC: setOwner calls for sync_buf/async_buf removed; * added a name string to sync_buf/async_buf FDDI70.arj * corrected a missing ':string' type declaration in a NED file FDDI69.arj * upgraded for JAR100/SIM158: parameter type declarations (string, numeric) added to NED files; * adopted to cancelEvent()'s changed behaviour ------maintenance completely taken over by Andras Varga------- FDDI68.arj * upgraded for sim151 FDDI67.arj * some beautifying FDDI66.tgz * inorder_epsilon value was corrected to 1e-10 * filenames were made DOS compatible FDDI65.tgz * separate files for different parts: The essential files for the FDDI MAC: fddi_mac.ned fddi_def.h fddi_mac.h fddi_mac.cc The files for the SSM Input and Output interfaces: ssm_ifs.ned ssmifdef.h ssm_ifs.h ssm_ifs.cc The file that contains the topology description of the FDDI network: fddi_net.ned Other parts of the simulation: othermod.ned otherdef.h othermod.h othermod.cc FDDI64.tgz * converted to be runnable with simulation kernel sim148e * the implementation of the FDDI MAC was placed into file fddi_mac.cc FDDI63.arj * bugfix: in case of empty histogram no packets to be generated by the IIF FDDI62.arj * syncpoint calls are added, runs parallel well. (Histogram dup() bugfix in sim. kernel was done.) * SSM-UI (time driven) produces similar results than SSM-UT (counter driven). * SSM-UI-1 and SSM-UI-2 produce slightliy different results, because SSM-UI-1 runs in one task and both rings use the same random number generator, but SSM-UI-2 runs in two tasks and the two rings use 2 random number generators. FDDI61.arj * the code for sync is written, the OIF's take the parameters from file: * "delete" and "send" commands (send is the default, can be omitted) * "every", "at" and "from - to - step" constructions the sync. function calls are commented out, so it compiles. Tested, seems to work. (except for the sync, of course) FDDI60.arj * the program runs, but the simulation times of the segments are not synchronized yet. (runs with Aug. 6. version of sim) FDDI59.arj * ported to PVM, just the code is modified, NOT tested yet! FDDI58.arj * bugfixes * histogram constructors always get the appropriate parameters * OIF and IIF classes have data members or constructors/destructors NO MORE FDDI57.arj * all kinds of bugfixes, runs with sim133, produces results! FDDI56d.arj * NR, SR : correct cable lengths, IDs, network description files: "*." FDDI56c.arj * bugfix: new cMsg(*msg) was replaced by msg->dup() FDDI56b.arj * RingID2prefix() function was introduced case 0 : return "nr" case 1 : return "sr" default: return 0 prefixes for load file names. eg. "./load/nr000728.trc" FDDI56a.arj * bugfixes FDDI56.arj * SSM OIF and IIF implemented * TUBw NRing and SRing connected with wires * TUBs NRing and SRing connected through OIF - IIF pairs FDDI55.arj * two interconnected rings, routing on the basis of the packet length statist. FDDI54a.arj * Bugfix in load filename, + linux will use also the 6 char. card addresses FDDI54.arj * Bugfixes * LoadMultiplier introduced * virtual char * FileNameEnding() introduced * FDDI_GeneratorPiSquare2x1D: implemented, not tested * directory "res" was renamed to "load" * version.log introduced FDDI53.arj A hierarcy of FDDI_Generator classes was introduced. The user can select the generator type of the SAS-s from the *.ini file. Possible types: * FDDI_GeneratorFromTraceFile * FDDI_GeneratorHistogram2x1D // not tested yet * FDDI_GeneratorPiSquare2x1D // not implemented yet * FDDI_GeneratorKSplit2x1D // not implemented yet * FDDI_GeneratorKSplit2D // not implemented yet FDDI52.arj Introducing the parameter "TrafficFromTracefile", taking the load from tracefile if yes and from previously saved statistics file, if no. FDDI51b.arj Introducing the parameter mentioned above, taking the load from tracefile if yes and from previously saved statistics file, if no. But the second is NOT finished yet! FDDI51a.arj Thinking about introducing a parameter "TrafficFromTracefile"... FDDI51.arj In the NeD file: * parameter max_idle deleted * time is given in ns and us * between stations in the same building, 10m cable length is assumed In the .cc file: FDDI_Sink and FDDI_Monitor::activity() functions... FDDI50.arj IdleTimes mechanism deleted. There is only one FDDI_MAC activity function. if ( findpar("StationID") < 0 ) my_station_id = parentmod()->index(); //in the case of a ring made by indexing else my_station_id = par("StationID") FDDI49.arj Runs correctly (without msg name loss) with the new simulation kernel FDDI48.arj: The FDDI_Sniffer is coded, but not tested yet. It is very much like its base class the differences are: two virtual member functions redefined. (and these are made virtual in the base class) ERROR: repeat loses the name of the FDDI_FRAME-s FDDI47.arj: An FDDI_Sniffer exists, but its simples moules have empty activity() functions. Compiles, but doest work for the reason above. FDDI46.arj: The last version before building a real "Sniffer". Here, sniffer is still just a SAS. FDDI45.arj: Rewritten for the new version of OMNeT++ (sim 103) FDDI42.arj: The real cable lengths are used: transmission delay is calculated from them using light speed of 200m/microseconds ... FDDI41.arj: The direction of the connections was reversed, because they're always "upstream neighbours" The address of Goliat was corrected to "Sun___1BD035" The sniffer is stll in its old place, it should be transferred, if the new 10 results would be used FDDI40.arj: The MAC algorithm was rewised according to the FDDI MAC standard. The algorithm used is described in: STATETR2.XLS This version seems to operate correctly FDDI38.arj: Runs under dos correctly too: the last 6 characters of the linux filenames are used under dos. ... FDDI37.arj: Takes the load of the fddi stations from files, runs only under linux, because the file names are 12 char long. (MAC address) FDDI33.arj: The topology of the Northen ring is decribed, everywhere the same cable is used compiles, but does not run: loop bounday problems with CON-s. ... FDDI31.arj: the token goes round amoung 3 stations of the N. ring FDDI30.arj: preparing for TUB Northen FDDI ring smulation, it compiles, but runs still the vector implemented ring as in fddi24.arj FDDI24.arj: more or less working FDDI ring version, implemented as a vector of computers