Parent Directory
| 
Revision Log
| Links to HEAD: | (view) (download) (annotate) |
| Sticky Revision: |
USe PNG image for splash, instead of SVG. Change installer to use Tcl 8.5.8.2. Fix URL for Tcl/Tk download.
Merging slvreq functionality into trunk.
Mask out some files from the tarball.
Fix exclusion of 'testipopt' binary from prereqs for tarball
Working on bug #439.
Fixes bug #440.
Fixes bug #438.
Fixing CONOPT_PATH problem again (Linux)
Fixing apparent problem with CONOPT_PATH on Linux (needs checking!)
Bump trunk to 0.9.8. Some work on detection of SUNDIALS on Linux (will need to be reviewed on Windows). Update DistTar for smoother building of tarball, less stray deps.
Remove 'nsis' from tools to avoid 'scons dist' errors on Linux.
Fix SWIG detection using Python 'subprocess' module.
More work sorting out the pixbuf loaders, not right yet.
Working on fixing up pixbuf loaders, probably need to flatten directory structure.
Some progress on copying GTK into the Mac installer.
Add IPOPT to the installer for Windows.
Remove corrupt analyze.allterms.c file.
Switch to 'subprocess' in config tests, stop using deprecated os.popen4.
Changes made for 0.9.7 (?) release against latest GTK, PyGTK.
Changes for detection of GTK on Mac. Changes for building .dmg disk image on Mac. Ignore flags for Mac generated files.
Update support for SUNDIALS 2.4.0. This fixes a bug with boundary crossing detection, as was hoped!
Almost there... ASCEND.app runs OK from command line now, but errors from Finder.
Some more efforts on Mac ASCEND.app. Need to add DYLD_LIBRARY_PATH stuff for libascend still.
Fixed plist installation, but file still seems not to be detected.
More work on ASCEND packaging for Mac.
Fixed building of 'app' for Mac... needs testing.
Some efforts towards a Bundle for an installer for ASCEND on OS X.
MERGE Jose's branch [2274]:[2584] into trunk
Few minor changes to fix build with SWIG missing.
Fix for OSX linking against Python.
Empty message
Moving SIZEOF_* vars into config.h.in.
issues resolved: 295 390 301 cmslv.c: unused var cleanup. lsode/SConscript: fortran flags bugs-- may not work with 0.9x scons. works with 1.2+. when adding -w, or any special flags, be sure to add them and not replace the original flag. system/var.c: 64bit clarity. system/discrete.c: 64bit clarity. system/analyze*: g_reuse declared in wrong place. 64bit clarity system/diffvars: missing prototype function, 64bit clarity. compiler/numlist.*: changed from int to glint. compiler/simlist.c: missing includes needed for 64bit clarity. compiler/instance_io.c: missing includes needed for 64bit clarity. compiler/initialize.[ch]: const clarifications. compiler/packages.c: const clarifications. compiler/module.c: const clarifications. compiler/statio.c: unused var cleanup. compiler/procframe; const clarification. memory deallocation bugs. compiler/notequery.c: repaired multiple casting and 64bit issues. compiler/importhandler.c: const and free issues fixed. compiler/type_desc.c: ridiculous if constructs clarified. compiler/createinst.c: casting stupidity repaired. linear/ranki2.c: missing includes needed for 64bitness. solver/solver.c: const issues clarified. utilities/ascConfig.h: added GLint typedefs for dealing with gllist 64bit portability. utilities/ascPanic.c: removed extraneous const. general/ospath.c: safer,quieter handling for string pointer difference. integrator/integrator.c: const issues clarified. packages/sensitivity.c: missing includes needed fo 64bit sanity. tcltk/interface/Integrators.c: 64 bitness. tcltk/interface/SimsProc.c: const errors. tcltk/interface/Driver.c: fixed env var handling wrt ascend-config (295) models/test/z*a4c: fixed meters -> m conversion; someone never ran the test suite after teasing the default units to ambiguous abbreviations. SConstruct: added sizeof checks; output might be better put in a ascend system-wide header.
Fixed a few problems with ABSOUTE_PATHS that emerged during packaging. Updated .spec file. Forced use of ABSOLUTE_PATHS in deb package.
Bump version to 0.9.6. Documented changes since last release in CHANGELOG.txt
Fixed up Tcl/Tk GUi to work with both ABSOLUTE_PATHS=1 and =0.
Fixing up Tcl/Tk build with renamed dirs.
Fixed compile for new header file locations <ascend/compiler/xxx.h> etc.
Fixed building of test suite after code reorg. Test suite may still be broken though: seems to segfault and give lots of errors.
Trying to fix up ASCEND for change of model library location to /usr/lib/ascend/models.
Fixed 'preview' window in canvas GUI. Add missing link to libm and libdl in Linux.
Break ASCEND into dev package and main package. Fix up gtksourceview syntax rules, add more keywords, remove C cruft. Add tools/scons/ascend.py tool for detecting ASCEND when building software using SCons. DTAR now fails if dpkg-buildpackage fails, keeps temp files hanging around for inspection. Plots generated by 'plot.a4l' no longer add legend if no labels specified. Committing python bindings for FPROPS. Much testing remains to be done there.
Fix for Ubuntu 9.04
Added dependency of installer on the 'solvers' directory.
Fixed support for SCons 0.97 on Linux (Ubuntu 8.04). dtar checks for fakeroot before proceeding. Debian package doesn't depend on g77, we prefer gfortran instead.
Fix for SCons 1.2.0: problem with detection of g77 on MinGW. Still need to test on Linux. Modified ascend-config for Windows to output paths with double backslash, seems to fix a problem with ParseConfig in SCons 1.2.0.
Example exergy calculation. Added delta_entropy to atoms.a4l. Cleaning up some library dependencies issues in build script. Increased default block size to 64×64 in canvas-based modeller.
Update copyright date of manual. Added ignore .deb,.dsc,.changes in tarball.
Fixed detection of and linking to Graphviz on Windows. Still problems with graphviz output due to file passing from Python to ASCEND (msvcrt version issue).
Fixed detection of Graphviz on linux where pkg-config is available (should be always). This will break detection of Graphviz on Windows though.
Fixed dtar to support separate debian.tar.gz file.
Add 'debian' folder into tarball once again. Add new 'dtar' tool (will describe it on wiki). Update debian/changelog.
Change to MUMPS from test of IPOPT. Tested with F9.
Fixed against Tcl/Tk on Fedora 9. Removed lots more env.Copy(), change to env.Clone() for newer SCons.
Fixed for missing libpython (DistutilsPlatformError). Fixed several env.Copy() warnings. Removed swig 'SideEffect' statement for current SCons swig tool.
Updated for IPOPT 3.4.0. Support for IPOPT 3.3.x is broken due to new path structure for #include files.
Fixed for case where LyX missing. Checking all with Ubuntu 8.04.
Bump version!
Fixed bug #361. Moved a number of vars out of SUBST_DICT to prevent config.h changes from spawning a rebuild of libascend. Eliminated use of ASC_WITH_CONOPT from C code (still need ASC_LINKED_CONOPT for the moment).
Fixing up DOPRI5 as standard integrator. Fixed bug 357 preventing repeated registration of integrators (in integrator.c). Cleaned up SConstruct file use of WITH_SOLVERS a little bit.
In process of moving DOPRI5 solver to 'solvers' dir.
Hack to correct treatment of GraphViz on Ubuntu 7.04 (evil RPATH).
Fixed 'lmodern' test.
Fixed detection of 'boolean' definition in GraphViz graph.h. Newer versions of GraphViz (2.16) include this definition, although it's not included in the older versions (2.12). Older versions still (2.8) are in use (Ubuntu 6.10 and Ubuntu 7.04).
Added test for GraphViz-defined 'boolean'.
Fixed detection/linking with GraphViz in Windows (GraphViz 2.16). Added support for linking against MSCVR71.DLL but still needs some work.
Fixed RPATH thing for Ubuntu 7.10
Fixed RPATH nonsense with GraphViz.
Add RPATH to env, not just libascend_env.
Adding RPATH to handle strange GraphViz on Ubuntu 7.04 and 6.10
Fix for different GraphViz include paths on Windows.
Added comment about GRAPHVIZ_CPPPATH.
Added graphviz-devel build-time dependency for RPM build. Fixed treatment of *_LIBS flags from command line in SCons.
Changed fundamental units to the standard SI abbreviations. Added new solver parameters for IPOPT (still needs testing). Added context-aware UnitsDialog that by default uses the type selected in the browser. Added attempt at detection of 'lmodern' (for bug #351). Changed base units in compiler/units.h.
Removed GtkSourceView stuff from SConstruct (it's not the right place for it), fixed ascend.spec.in to include the necessary code.
Fixed installation of language definition file for GtkSourceView-2.0. Removed debug output in cond_config.c, used by CMSlv.
Added 'Incidence Graph' functionality for PyGTK interface.
Removed references to MFGRAPH, seems better to use GraphViz directly. Added configuration test for GraphViz. Added configuration test for strdup, fixed mmio to not create its own strdup if other form is available already.
Removed references to <malloc.h> (replaced by <stdlib.h>).
Comment out CheckDLOpen, as it breaks Windows build.
fixed CheckSigReset
changed conf.CheckIPOPT
Added tests for 'malloc' and 'dlopen'/'LoadLibrary'.
Suppressed LSDOE warnings. Reverted some of the 'is False' changes that weren't necessary.
Fixed to support 'scons -h' in current scons (see thread data 1/11/07 on scons-users list).
created typedef ExprSymFilterFn. updated GtkSourceView lang definition. removed execvp call in test.py.
fixed bug 307
Added CUnit test suite for 'compiler' module. Added a couple of *very* preliminary tests. Removed some debug output from packages.c Fixed a #include missing from module.h. Added a couple of missing #ifdef __cplusplus from test header files. Added little bit of test code into relman_hessian_count (still active).
fixed detection of missing C++ compiler (tested on Ubuntu 7.04)
Added 'asc4dev' script.
Added initial code for 'accumulate builder' based on SCons Wiki code.
Adopt freedesktop.org method for mime association on Debian. Add postrm and postinst scripts.
Debian uses a didfferent format for the MIME files. Bumped version.
Fixed building of bboxtest for new naming convention.
All platforms will now use the lib<name>_ascend.so convention for ASCEND external libraries. (it makes it simpler for people writing such things).
New external library naming scheme on Windows: 'name_ascend.dll' instead of 'name.dll'. This applies to solvers as well as external methods, extpy, import handlers, etc.
Modified Debian build log. Added creation of debian.tar.gz for use with OpenSUSE Build Service. Added notes on building packages using the Build Service.
removed redundant comment
Fixed python executable path on Windows.
Fixed default web help URL. Fixed handling of web links on Fedora (needs checking on other Linux)
Bumping version. Will attempt to do this immediately *after* a release from now on, before new work is started.
More work on Windows installer
Changed default fortran to G77 on MinGW/MSYS.
Added 'UPDATE_NO_YACC_LEX' option to SCons to stop _no_* files being updated all the time. Minor commenting in asc_ipopt. Fixed ascend-config for case where -I/usr/include or -L/usr/lib.
Added lots of svn:ignore tags
Fixed detection of web browser on Windows (no need for Win32 python package now). Removing SUNDIALS_* flags where not needed. Removing debug output about not having YACC or LEX. Fixing typelex for case of LEX missing (tested with Windows).
Adding some windows target files to ignore list. Fixing build on Windows with Python 2.5 library. Fixing for NSIS not present.
Fixed build of asc_ipopt.c. Small comment added in relman. IPOPT_LIB replaced by IPOPT_LIBS due to multiple linking requirement of that library.
Some work on IPOPT detection (more required)
Updated config for CONOPT_LINKED
Little more on IPOPT. Fixed ascdev script to work with configured solver list.
Making build of IPOPT dependent on config test.
Fixing for missing yacc/lex on Windows, auto-update of them when yacc/lex *are* avail. Fixing 'test.py' operation on Windows (ASCENDSOLVERS env var) Adding missing ida.dll to installer. Adding some missing export symbols in libascend.
Adding support for ASCENDSOLVERS env var, for defining location of the ASCEND solvers. By default this will be /usr/share/ascend/solvers.
work on fixing Windows for the external solvers changes
Bumped version to 112.
Dependencies of shared libraries on 'libascend'.
Fixed external loading of integrators, at least on my system. Needs testing with/without fortran, sundials, etc. Changed little thing kn d1mach.c to make default behaviour correct on Linux.
More on relocation of solvers
Moving solvers to their own directory
More of the TRON stuff
Fixed default detection of Tcl/Tk on Debian-based distributions (Ubuntu 7.04 specifically)
fixed WITH_DOC stuff
Fixing LyX detection? (this commit equals 0.9.5.111)
More work on Windows installer detection of PyGTK, PyCairo etc at install-time.
Add python version to installer name in Windows
Move 'create.nsi' to top-level
Added 'alert' option to 'loading' module. Added start menu shortcut to Model Library Bumped version to 0.9.5.111. Fixed loading of fourbarplot.py example when matplotlib not available.
Removed 'ascend.bat' hack from installer, changed to saner 'pythonw.exe' for starting ASCEND on Windows. Still needs testing on Linux.
DOPRI5 is running... but is it correct? Still needs lots of testing
Added tests for presence of CC and CXX.
Fixed problems with installation of documentation on FC6
Build documentation by default, but detect when LyX is not available.
Fixed problem with tcl/tk in 'scons install' for SCons 0.96.92
Fixed build for SCons 0.96.92. Needs testing.
Fixes for the windows installer, start menu, splash screen.
Working on fixing for SCons 0.97. There is a problem with SCons 0.96.92 on Ubuntu 6.10, waiting an answer on that.
Some more files excluded from the tarball
Some experimenting with crossmingw cross compiling. Removed shebangs from files that shouldn't have them.
More work following Debian package feedback from 'lintian'
Almost there with the debian package now
Fixed problem with Driver.c in ABSOLUTE_PATHS case. Updated changelog and spec file for new 0.9.5.108 release.
Added <nvector_serial.h> in config tests.
Fixed support for SUNDIALS 2.2.0 on Ubuntu Feisty
Working on support for SUNDIALS 2.2.0 (ongoing)
Added support for standard DEbian package for SUNDIALS (ongoing)
Fixed default CONOPT_PATH (now /usr/lib:/usr/local/lib on Linux)
Fixed FORTRAN flag?
Updated ipython_view from Accerciser. Allowed SCons version 0.96.96 (works OK on FC6)
Added plot support in Integrator output tabs. Some other minor debugging for pylab integration and idaanalyse output.
Fixed build problem on MinGW GCC 3.4.2 Fixed problem with font in IPython console on Windows Some attempts to fix problems with FILE* support in Python bindings gfortran is now the preferred FORTRAN compiler. Comment out failing FILE* tests for the moment.
Added SCons version check so that Krishnan stops wondering why he can't build ASCEND on brand-new shiny version 999 of obsurre linux platform X.
Removed ASC_IDA_NEW_ANALYSE condition from the codebase. Cleaned up some debug output.
Fixed problem with logrel instances in PyGTK GUI. Removed faulty INSTALL_DOC directory -- docs just go to INSTALL_ASCDATA for now.
Added bindings to UFSPARSE sparse matrix library, which provides matrix algebra operations via the CXSparse library which it includes. Tested with the UFSPARSE that comes with Ubuntu 6.10. Added simple test case (no validation)
Added test suite for linsolqr (very early days still) for checking by Ben
Another one of these big refactorings. Split out 'system' directory from 'solver' directory. This will need more work but it's a start at getting the non-solver-specific code out separately.
Added -fPIC to SHFORTRANFLAGS on Linux platform (attempt to fix problems of M Fothergill on 64 bit linux)
On Windows we don't need the libm test.
Fixed prob with HAVE_IEEE. Added self_test method to bvp.a4l bvp_test.
Fixing 'exit' to 'Exit'. Bumping version to 0.9.5.107.
Turn on signals again by default
Some disgraceful hacking to obtain SIGFPE details in ida.c. Working next to fix the proper ASCEND ascSignal implementation (which I broke at some point).
Added getParameter by name in SolverParameters. Fixed up refences to signal-related header files. Fixed up various warnings from -pedantic compilation. Identified an apparent bug with RelationCalcGradient (unsafe version). Added support for compilation of a standalone C++ program (idatest) in SCons build.
Defaulting to IDA_NEW_ANALYSE
Huge reorganisation -- all mtx and lin files moved to base/generic/linear. Required many #include statements to be changed.
Shifted everything integration-related out into a separate directory.
Fixed 'ascdev' script to work correctly in MSYS (now needs retesting in linux). Removed 'make_matrix' and 'free_matrix' from sensitivity code and reused densemtx.c instead. Fixed linking issues in finitediff.dll.
Fixed escaped paths in SConstruct. Added some more keywords to gedit syntax definition.
Got paranoid about turning off signal handling... added '#ifdef ASC_SIGNAL_TRAPS' anywhere that signal handling referenced. Signal handling is OFF by default now.
Another small fix for mfgraph linking
Small type relating to mfgraph
Fixed mfgraph binding
Added some more user error messages from the solver. Added detection for the MFGRAPH library, which will be used to output GraphViz files based on the incidence matrix.
Catch regex prob in MinGW buildbot build
Catch regex prob in MinGW buildbot build
Catch regex prob in MinGW buildbot build
Fixed up problem with IDA version in SConscript. Identified a bug with findIndependentVar in PyGTK GUI, ongoing.
woops
Tidying up MinGW stuff
Adding more diagnostics for MinGW buildbot build
More on fixing MinGW build for buildbot
Bumping version to 0.9.5.106.
Adding dump of env vars when stdio.h is not found.
Fixing MINGW detection
Bumped version to 0.9.5.105.
Added scons switch IDA_NEW_ANALYSE for the new code that implements a better DAE analysis procedure. Minor tidying in list.[ch] Removed some DIEDIEDIE stuff from analyze.c. Added 'odeid' to the 'varip' structure in same. Collapsed a bunch of repetitive stuff with those fancy #defines I like so much. Flipped some return codes from functions integrator.c to return 0 on success. Added #include for sensitivity.h in lsode.c. Catch exceptions from sim.build() in do_integrate() in PyGTK GUI. Attempted to fix tcl/tk for errors from integrator_analyse as well (not tested).
Added SCons test for 'lex'.
Some work on fixing error with test_ascSignal. Breaking down into smaller test cases. Removed some debug output from detection of ASC_RESETNEEDED. Changed all calls 'signal' to 'SIGNAL' macro that includes optional debug output. Removed 'libasctest.so' (made part of libasctestsuite.so FWIW) Fixed big in test.c wrt CUEA_ABORT. Added 'print_stack' and 'Asc_SignalPrintStack' and 'Asc_SignalStackLength'.
Added Matrix Market export routines (from math.nist.gov). Added necessary build commands for above. Work on Jacobi preconditioner for IDA (ongoing) Set integrator_analye to make a call to slv_block_partition. Not sure if that's a good idea or not. Tidied up comments in linsol.h Moved unnecessary #defines from model_reorder.h into model_reorder.c. Cleaned up codedocs in mtx*.h Added WITH_MMIO config.h flag.
Bumping version (checking buildbot)
Fixed for nonexistent libpython
Fixed detection of libpython on Windows. Fixed TestLSODE.testlotka test.py Removed a sundials_dense.h assertion under windows.
Added SCons tests to check SIGINT and to replace ascresetneeded (need replacement for this in Autoconf as well).
Removed debugging from createinst.c
Typo (text) in evaluate.c
Commented out redundant code in importhandler.c
Added signal handling in ExecuteCASGN.
Added missing ospath_free in ModuleSearchPath.
Exported InitSymbolTable, DestroySymbolTable in symtab (dubious)
Moved FPRESET macro out of ascConfig.h and into ascSignal.h
Added Asc_SignalHandler{Push,Pop}Default.
Added ASC_RESETNEEDED and HAVE_C99FPE macros in config.h.in.
Found the bug causing the SIGFPE in idakryx.a4c (raises a question about int/float division in modelling, I think)
Added system_destroy call in Simulation::~Simulation (dubious).
Fixed OSPATH typo Other minor doc changes.
Able to compile now on Solaris. Fixed a problem with datareader.c returning non-void. Remains a problem with std::logic_error not being linked to _ascpy
static linking to python, again
Bit more on building against static libpython
Fixed for static linking to python where libpython2.x.so is not available.
Modified SConstruct so that no assumption is made about the version of Python being used.
Added test for C99 FPE handling Fixing mess-up of ChildByChar in arrayinst.h header. Added 'safeeval' config option to IDA. Changed 'SigHandler' to 'SigHandlerFn *' in line with other function pointer datatypes being used in ASCEND. Moved processVarStatus *after* 'Failed integrator' exception (ongoing issue).
Added support for dmalloc (http://dmalloc.com) Added Integrator::setEngine(string name) Minor other stuff.
Adopted suggestions wrt Py_None on Windows platform (extpy) Fixed up default values of CUNIT_PREFIX (needs testing for case where missing)
Expanded the 'ascend-config' mechanism to be supported in Windows installations (On Windows, ascend-config is installed in $INSTDIR and uses the registry instead of hard-coded file paths).
Added test for SUNDIALS version (2.2.1 and 2.3.0-pre are preferred) Fixed silly warning message about ignored return values from CONSOLE_DEBUG on GCC.
Bumping version for fixed IDA integrator
Some success with IDA: fixed up the indexing dilemma and was able to integrate 'johnpye/thermalequilibrium.a4c' for a short time span (but through to 3000 s as with LSODE). I would blame lack of jacobian routine in the first instance. Added 'more properties' button in Properties dialog for a variable, to allow values of ode_id, ode_type etc to be queried (but not changed).
Bumping version to 0.9.5.102 after checking windows build.
Better plot points in moody diagram model. Attempt to get rid of some compiler warnings in extmethod. Better handling of LD_LIBRARY_PATH and ASCENDLIBRARY in ascdev. Changed default documentation path (to avoid warnings, seems like a scons bug?)
Trying to fix installation of the ASCEND manual, still not quite right :-(
bumping versions to 0.9.5.101
Some work on the IMPORT handler functionality. Mostly just placeholders at this stage.
Improved detection of Python, GTK, Tcl/Tk, Glade in Windows installer. Bumped version to 0.9.5.100
Added new script 'ascdev' to launch ASCEND in 'developer mode' which means that files in the source hierarchy will be used instead of the installed files.
Updated installer to allow separate choices of PyGTK and Tcl/Tk GUIs at install-time. Installer won't allow GUIs to be installed unless ActiveTcl and/or Python are located. Bumped version to 0.9.5.99.
Bumped version to 0.9.5.98. Added support for 'setSymbolValue' from C++ interface, and allow symbol variables to be edited via the PyGTK GUI. Added some limited data about structural analysis of beams in johnpye/aiscbeams.a4c. Updated johnpye/beam4.a4c to use this data. Added some currency conversions to measure.a4l. Added some units related to Mechanics of Solids to atoms.a4l.
Renamed testTrig to testfunc.a4c. Removed some buildtime messages. Added self-test to testfunc model. Some windows-specific changes to SConstruct (need testing?)
Added an attempt to fix the long-command-line-on-windows problem, then disabled it pending assistance from the scons user's list
bumped version to 0.9.5.97 added error message when datareader file not found in search path fixed datareader linkage for Windows.
Adding the 'sun.c' and 'sun.h' files back into ASCEND so that we don't need dependency on libradtran. Added detection of Scrollkeeper, the freedesktop.org documentation manager. Added placeholder SConscript file for building the user's manual.
Added vim 'syntax=python' comments to many SConscript files.
Lots more work on the datareader extfn.
The testtmy.a4c file requires the file obtained here ('gzip -d' before use)
http://rredc.nrel.gov/solar/old_data/nsrdb/tmy2/unix/23161.tm2.Z
In slv9, made changes to keep CONOPT from complaining. CMSlv still not working as expected though. Decreased required version of Bison to 2.0. Switched search path for CONOPT on Linux to the CONOPT_PATH env var (LD_LIBRARY_PATH was being interfered with when ASC_DEV=1). Some other debug message changes.
Removed incorrect copyright in extfntest.c. Fixed CONOPT detection on linux in SConstruct. Tidied up comments in instquery.h.
Fixed for CONOPT under Windows.
Added 'testcmslv' model. Fixed error in CONOPT build test.
Bumped version number 0.9.5.96
Enabled the LRSLV solver, so that CMSLV will be able to function correctly.
First work towards reinstating the CONOPT solver. So far, I've got it building and linking without errors, but haven't tested the resulting runtime.
Radu Serban sent me a preview of the new version of IDA which has new header file layout. This patch updates for the new layout. Also couple of minor fixes for gcc warnings in numlist and library.cpp.
Added 'docs' target to run doxygen for libascend.
Added necessary changes to allow 'johnpye/extfn/libextfntest.so' to be built automatically as part of main SCons build. This will make testing of external relations a little easier.
Lot of svn ignore tagging. Little more work on DSG model.
Few small changes to fix for Linux after the recent MSVC++ oriented changes.
I think this almost resolves all the Visual C++ issues. Just an outstanding issue with the use of the option struct in LibraryProc.c eg use of &g_compiler_warnings in initialiser.
Fixes for MSVC++
Fixing build for Microsoft VC++ 8
Working on fixing for MSVC on Windows.
Added '--models' option to ascend-config. Fixed a bug with packages.c wrt generating absolute paths to library files. Added SCons directives to install header files in INSTALL_INCLUDE subdirs. Changed rel.c so that init function is only run if provided in the package. Added 'ospath_getabs' to convert relative paths into absolute paths by adding prefix of fully-resolved '.'
Monster commit! Lots of recommenting and reorganising of external relations-related stuff. Replaced a lot of ascmalloc and asccalloc calls with the new ASC_NEW* macros. Fixed (?) the problem Art is having with icons in PyGTK. Turned on -Wall in SConstruct and fixed up a stack of warnings. Removed the redundant exit(2) from after Asc_Panic calls and added __attribute__((noreturn)). Set doxygen to create callgraphs to level 2, updated doxyfile to version 1.4.7. Fixed up building of extfntest.c.
Improved detection of math library. Fixed bug that was affecting detection of CUnit library.
Changed default SCons option, was TCL=~/activetcl now TCL=/usr. This means that by default we assume that the user will provide suitable Tcl/Tk from their on-root libraries.
Added 'ascend-config' scriptlet to be used when building external libraries and other software that must link to ASCEND. Also some minor updates to the INSTALL instructions.
Adding explicit dependency of tarball on 'ascend.spec'. Use automated creation of SRCS lists in base/generic/general makefile.
Fixing up incorrectly merged SConstruct (merging r702-820 from dae branch to trunk)
Small changes to CUnit detection.
Bumping version number
Fixing up for buildbot problem with null release ID
Removed dependency on ccache. Added RELEASE flag to allow BuildBot to control RPM versions.
Bumped version to 0.9.5.93. Added Tcl/Tk GUI to the GNOME applications menu.
Turning off MSVC extensions when compiling
Minor changes to bison/yacc detection, hoping this will make MSVC happier...
Fixing up Yacc/Bison detection a little.
woops
Setting default Tcl path to one that MSVC likes as well as MINGW.
...
Fixed up some other small probs with MSVC2005
Changed the way that installer is created on Windows: now you must use 'scons installer' to build the installer. This will only work on windows. The WITH_INSTALLER option has been removed. Switched the expected version of Tcl to 8.4 under Windows for builds with SCons.
Added changes to allow autotools build on ubuntu. There were some wrong assumptions about file locations that I was able to correct using information in the tclConfig.sh and tkConfig.sh files. This will need testing on Fedora/Unix/etc.
Fixed up standard paths stuff to use 'INSTALL_ASCDATA' as the place where tcltk/python/models etc will live. The confusion here came from the fact that I was taking 'INSTALL_SHARE' to be be assumed to mean 'the location of /usr/share' -- hence the resulting '$INSTALL_SHARE/ascend' that was bothering Ben. The 'INSTALL_ASCDATA' fixes this, and behaves as he will like, I hope. Also note that I have switched the default situation in config.h.in to use relative paths by default, which is the preferred CMU way. SCons builds will default to absolute paths on non-Windows systems.
Fixing the windows build. Added full support for Tcl/Tk version to the installer.
Summary: most of the recent complaints about autotools build fixed, but the buildbot might find a problem with the scons build/install/rpmbuild. Details: configure,Driver.c,utilities/config.h.in: The choice of INSTALL_DATA as a configure variable name has been repaired (renamed INSTALL_SHARE); INSTALL_DATA is the unix canonical name for the program 'install' handling a data file. Both autotools and scons builds have been updated to reflect this, in the process correcting the definition from prefix/share to prefix/share/ascend. autotools build: Added --with-quiet option to make the install of already installed and unchanged files less verbose. Put in a workaround for the '' problem in tkConfig.sh seen by KC/JP. Added --datadir support to autotools configure. If you configure the autotools --datadir=PREFIX/share/ascend you get the same behavior as the Pye packaging; by default you get the CMU packaging convention instead. Added utilities/env.c to build. compiler: Turned off AWAL in anontype.c. Noodling away at blackboxes continues...
Completing missing part of RELATIVE_PATHS configuration for SCons, to address bug #262.
Still working on a way to add options to config.h.in such that it doesn't break autotools.
fixed for scons, can't work out a way to allows @MISSING_SOMETHING@ to fallback safely, unfort.
fixing so that Ben won't need to add this macro to autoconf, hopefully.
Adding configuration option to turn off xterm colours in output.
Bumped version to 0.9.5.91. Changed WITH_CUNIT_TESTS to WITH_CUNIT. Added GCOV scons option. Fixed up 'test' target for SCons. Added lots of export symbols to libascend.so.
Added GCOV=1 option to SCons build, working on integrating LCOV tool with Buildbot.
Fixing some malloc bugs.
Working on improving handling of paths.
Fixed for Ubuntu 5.10 (linking to libstdc++ when building _ascpy.so) Tested with MinGW/WinXP, still OK.
Customisable installer .exe names. Moved NSIS tool into the scons subdirectory, reverted to standard NsisSconsBuilder from the SCons Wiki.
Bumping the version number for NSIS and RPM pre-releases
Changing from 'tcltk98' to 'tcltk', in progress.
Remove dependency on perl5
Manual override to set name of created tarball (for use with buildbot testing).
RPM builds now, but has a dependency on perl(!?!)
Changes to detection of TkTable. TkTable detection was not possible on FC5 due to the provided tktable package not coming with header files, and no tktable-devel being offered. Hence static linkage to Tktable is not possible. Maybe it's not a great idea anyway. Also modified and updated the DistTar builder for SCons. Modified SConstruct to allow Tcl/Tk 8.4 to be used. Modified ascend.spec.in to use system-standard version of Tcl/Tk 8.4.
Fixing some dependency issues with the building of the SWIG wrapper. Added automatic SWIG dependency scanner (from the SCons Wiki).
Some fixes to the 'disttar' thing for 'scons dist'
Some more on the RPMs. The tarball still missing some files.
Fixed building of source tarball using 'scons dist'. Will next add automated building of RPM targets. Started work on building of separate RPMs for tcltk and python interfaces.
Created 'scons install' command that also works for Tcl/Tk interface. These changes required some modifications to the default values of the environment variables. Installed Tcl/Tk version now works with only the LD_LIBRARY_PATH env var being required.
Added LICENSE.txt to root directory Fixed a 'const' problem in ospath Fixed up some variable names and key errors in SConstruct Modified License-Warranty.tcl in accordance with the recommendations in the GPL.
More on RPM
Adding 'gcc' to runtime requirements of RPM Adding gccvisibility switch.
This is a patch to allow statements like "IMPORT johnpye/extfn/extfntest". It modifies the methods in packages.c. Plan is to reuse the same code for "REQUIRE xxxxxx" statements as well.
More work on the ospath.c file, trying to iron out some windows-specific issues Fixed McGraw-Hill URL in TCL files. Created platform-specific TCL_LIBPATH so that c:\Tcl\bin is used instead of c:\Tcl\lib for .dlls.
Working on platform-independent pathnames for the IMPORT command. Added 'ospath.c' to base/generic/general for this purpose. Patched kvalues and sensitivity to use the 'IMPORT "libname";' syntax instead of 'IMPORT registerfn FROM libname;'. Fixed pathnames in create.nsi. Added GPL header to coupla files. Added quoting to 'IMPORT' syntax in ascParse.y. Removed 'PackageOption' from Tcl/Tk related Scons options so that missing Tcl/Tk doesn't cause breakage.
Removed 'interface' directory in trunk/pygtk/interface (moved everything up a level) Made corresponding changes to SCons* and spec files.
Working of debugging the mysterious PyGTK bug. Seems to be coming from a wrong char *filename in an error_reporter call from the DEBUG_RELS sections of the compiler code. Also renamed the WSEM and WSNM macros for clarity, removed unused file pointer.
Fixed up a problem with ordering of headers with Python.h. Removed psyco initialisation, pending evidence that it actually does something :-) Disabled python 'director' error message callbacks, trying to debug a segfault. Added ability to disable GCC Visibility, scons WITH_GCCVISIBILITY=0. Turned off runtime signal tests for SIG_INT and SIG_FPE (makes using GDB a pain), see base/generic/utilities/ascSignal.[ch].
Getting the modular version the Tcl/Tk GUI working. In progress.
SConstruct for python2.3
SWIG < 1.3.28 can't do GCC visibility, so turn off for python interface in that case.
More fixes for compiling and running on ubuntu 5.10 with GCC: gcc version 4.0.2 20050808 (prerelease) (Ubuntu 4.0.1-4ubuntu9) Added ASC_SHLIBSUFFIX and ASC_SHLIBPREFIX which are used in packages.c in preference to platform-specific #ifdefs.
Fixing for ubuntu GCC 4
The Visual C++ 2005 version now builds without linker errors. But I ran the installer and it crashed during the first C++ call, so still problems.
woops, fixing typo that showed up with GCC visibility on MinGW.
Added check to see that YACC runs OK. Fixing build with MSVC++ 2005. All builds now except for some problems in the SWIG-generated code.
Managing GCC visibility is the job of the build too, not ascConfig.h
Reduced some runtime debug messages from Python and C++ and base/generic/compiler/type_descio.c. Fixed up support for GCC Visibility and add SCons 'sniffer' for this. Fixed a bug with 'fileopenpath' in PyGTK interface (File->Open location is remembered from last time). Fixed a bug with missing includes in C++.
Working on Linux again. For some reason you need to run /sbin/ldconfig though. I'm going to ask Ben about this.
MinGW build fixed and working again.
Changed SConstruct so that if you run SCons from MSYS, SCons won't try to use MV++ tools.
Wrapped extern "C" {} around #include <utilities/ascConfig.h> in pygtk/interface/config.h.in.
Added 'HAVE_LEX' and 'HAVE_YACC' env vars in SCons so that we can be smart about using _no_yacc source.
Changed tcltk98 SCons script so that _no_yacc is switchable.
Huge cleanup of file header comments (GPL notice etc). Changed from ASC_DLLSPEC to ASC_DLLSPEC(TYPE) -- like python Fixed up some badly export symbols (missing 'extern') Fixed exporting of symbols from error.h Removed ASC_DLLSPEC from .c files-- only in .h files now (hope that's right...) Changed the way bintoken does its exporting, to take advantage of ASC_BUILDING_LIBASCEND (needs testing) I wonder if the 'extern' keywork should be a part of the ASC_DLLSPEC macro?
Fixing build for SWIG 1.3.24 (no -O flag) Adding more debug info to loading output in PyGTK interface/config.py.in Adding SWIGVERSION flag to SCons env. Commenting in ascDynaLoad.c and ascParse.y Adding VERSION to config.py
Rolled back revision 451, since the DLL loading issue on Windows is now resolved. Fixed up some #ifdef brackets in slv*.h headers. Added NSIS functionality to SConstruct file Renamed default installer to 'ascend-setup.exe' pending a more clever approach to version numbers.
Renamed 'WITHOUT_TCLTK_GUI' option to 'WITHOUT_TCLTK'. Hard-wired path to c:\Python24 and c:\GTK\bin in Windows Python GUI installer.
Building Tcl/Tk with dynamic linking to shared libascend.dll. Still some problems.
Converting to the new shared-object configuration (on email discussion with Ben) First stage: PyGTK interface on windows. More to come.
Fixing for 'STATIC_TKTABLE' flags for Fedora Core 5
Fixing default PackageOption values Adding --library option to gtkbrowser for ASC_DEVELOPING override of 'models' path.
Tidying up some debug output Fixing up use of icon filenames on Windows Removing reference to .svg file in glade/ascend.glade
Remove default CUnit locations for failsafe 'scons -h' on Windows
Got the test suite to compile with SCons. Seems that there are some serious problems still though.
Adding 'models' directory to SCons install
RPM build works now. The model library isn't being installed but the PyGTK interface opens up correctly.
Added ascend.spec (for 'rpmbuild -ta' command). Fixed VERSION substitution in ascend.spec.in.
Installation instructions and an RPM SPEC file (untested)
Implemented installer version of PyGTK interface. Use 'scons install' to install. See 'scons -h' for new build options.
Starting work on linux-based install process
Much progress on the MinGW build of the Tcl/Tk interface. Builds and links now, just having some problems getting the DLLs for Tcl/Tk to resolve at runtime.
tentative: Disabled ascDynaload functions when DYNAMIC_PACKAGES is not defined. Altered error.[ch] so that MS VC++ variadic macros are used when available. Disabled Windows MessageBox in ascPanic, so that compilation of base engine is possible with the Platform SDK. Added SConscript files to build FORTRAN components. Removed some autoconf-related stuff from the pygtk/interface directory.
Debug output for MSVC build
Small correction to keep MinGW build working
Some progress on getting a MSVC build to work with SCons. Issues with linking still.
Improved SWIG detection and version checking.
Trying to fix problems with the parser 'yy' and 'zz' stuff so that it works with SCons and hopefuly autotools at same time.
Woops
Added Tk detection, improved reporting
Added Tcl detection. Still to do: check that Tk version matches Tcl version.
Fixed bug #251
More work on tracking down Krishnan's problem with SWIG detection.
Fixing up message about bad SWIG version
Almost there with getting Jerry's tests to run. Just sorting out the linking.
Added SWIG version detection. Added some ignore patterns for files in the repository.
Fixing python interface on FC4. Still a problem with detection of the full path to swig under Windows.
Successful build of Python interface on Windows with MinGW/MSYS.
Fixed SCons build to do output of config file 'config.py' require by PyGTK GUI. Removed requirement for 'config.h' from the C++ file and instead supplied required flags via compiler commandline.
Some fixes to the SCons build: - adding flags to enable QRSLV and CMSLV by default. - adding commands to build SWIG library correctly for Python interface.
A little bit of progress on getting the windows build to work
Successful build of the PyGTK interface with SCons on Linux (FC4)
Added SConscript for PyGTK interface. It's not finding the lib files, something wrong with the -L options?
Added SConscript to Tcl/Tk directory. Still having some problems with linking.
First attempt at SCons build. This will build the static libraries for me on Linux with GCC 4.0.2. Will work now on getting it to build the Tcl/Tk GUI.
This form allows you to request diffs between any two revisions of this file. For each of the two "sides" of the diff, enter a numeric revision.
| john.pye@anu.edu.au | ViewVC Help |
| Powered by ViewVC 1.1.22 |