orbitcpp-commits Mailing List for orbitcpp (Page 3)
Status: Beta
Brought to you by:
philipd
Menu
▾
▴
orbitcpp-commits
You can subscribe to this list here.
| 2000 |
Jan
(32) |
Feb
(12) |
Mar
(40) |
Apr
(28) |
May
(5) |
Jun
(4) |
Jul
(2) |
Aug
(3) |
Sep
(4) |
Oct
(2) |
Nov
(11) |
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2001 |
Jan
(3) |
Feb
(4) |
Mar
|
Apr
(4) |
May
(3) |
Jun
|
Jul
|
Aug
|
Sep
(1) |
Oct
|
Nov
(9) |
Dec
(2) |
| 2002 |
Jan
(12) |
Feb
(5) |
Mar
(2) |
Apr
|
May
(1) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2003 |
Jan
|
Feb
|
Mar
(1) |
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Phil D. <ph...@us...> - 2000-11-16 14:47:11
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/11/16 06:47:10 Modified files: services/name : Makefile.am Log message: Use the ORBIT_IDL macro to get the correct path to orbit-idl |
|
From: Sam C. <edd...@us...> - 2000-11-13 01:35:31
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: eddiesam 00/11/12 17:35:30 orbitcpp/debian Update of /cvsroot/orbitcpp/orbitcpp/debian In directory slayer.i.sourceforge.net:/tmp/cvs-serv1186/debian Log Message: Directory /cvsroot/orbitcpp/orbitcpp/debian added to the repository |
|
From: John K. L. <jk...@us...> - 2000-11-12 08:13:38
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: jkluebs 00/11/12 00:13:36 Modified files: compiler : pass_xlate.cc types.cc types.hh Log message: I noticed a few problems with IDLArray after I made the last commit. Now, exceptions should work a little better, but there is a bug in ORBit's codegen for certain, really ugly exceptions, so I couldn't test every last detail (which shouldn't be an issue as you can't use such a construct with ORBit even). |
|
From: John K. L. <jk...@us...> - 2000-11-12 04:26:44
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: jkluebs 00/11/11 20:26:44 Modified files: orb : Makefile.am orbitcpp.hh orbitcpp_smartptr.hh orbitcpp_tools.cc orbitcpp_tools.hh Added files: orb : orbitcpp_any.cc orbitcpp_any.hh Log message: The array smart pointers now have a template parameter for the length of the array, to allow unique Array properties for arrays of different lengths. The support for Any is also being checked in. |
|
From: John K. L. <jk...@us...> - 2000-11-12 04:24:02
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: jkluebs 00/11/11 20:24:01 Modified files: compiler : pass.hh pass_xlate.cc pass_xlate.hh types.cc types.hh Log message: Support for Any (inserters and extractors are generated for everything less arrays which is in the works). Any types are marshalled and demarshalled. Bug fixes concerning arrays and sequences: No more seqfaults for arrays of sequences. No more segfaults for sequences in structs (without a typedef). Also arrays in structs (non typedefed) work correctly now. The properties functions for Arrays are only generated once for every unique (as far as CPP is concerned) kind of array so: typedef long Arr[5], Arr2[5]; does not cause multiple definitions. Arrays in interfaces work, with the mandatory functions made static and the properties generation deffered to outside the interface. Multiple definitions of sequences are also covered. TypeCode constants are generated except there is a // prepended until the typecode implementation is checked in by Sam. |
|
From: Phil D. <ph...@us...> - 2000-11-09 12:56:49
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/11/09 04:56:48 Modified files: compiler : pass_xlate.cc types.cc Log message: Added support for implicit _default() in unions |
|
From: Phil D. <ph...@us...> - 2000-11-09 11:57:54
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/11/09 03:57:54 Modified files: compiler : language.cc language.hh pass_xlate.cc pass_xlate.hh types.cc types.hh test/everything: client.cc everything-impl.cc everything-impl.hh everything.idl Log message: - Fixed CORBA::char constants - More idl union code. Not quite finished yet but the tests compile so I'm committing! (doesn't handle complex member types properly yet) - Added IDLUnionDescriminator abstract base class. Implemented by all IDLTypes that can be union descripters |
|
From: Phil D. <ph...@us...> - 2000-11-01 17:55:05
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/11/01 09:55:04 Modified files: compiler : language.cc language.hh pass_gather.cc pass_gather.hh pass_idl_it.cc pass_idl_it.hh pass_xlate.cc types.cc types.hh test/everything: client.cc everything.idl Log message: - Added an idl constants test - Added beginnings of union support |
|
From: Phil D. <ph...@us...> - 2000-10-26 20:02:28
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/10/26 13:02:28 Modified files: compiler : pass_skels.cc orb : orbitcpp_exception.cc orbitcpp_exception.hh orbitcpp_exceptionlist.hh orbitcpp_poa.cc orbitcpp_poa.hh test/everything: client.cc everything-impl.cc everything-impl.hh everything.idl server.cc Log message: - implemented POA::deactivate_object, POA::servant_to_id, POA::reference_to_servant - implemented CORBA compliant servant reference counting - replaced the 'delete self' in the fini method with a corba compliant '_remove_ref()' - fixed the system exception code to map on repoids rather than minor numbers |
|
From: Phil D. <ph...@us...> - 2000-10-15 17:22:08
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/10/15 10:22:06 Modified files: compiler : types.cc test/everything: client.cc everything-impl.cc everything.idl Log message: - Fixed a server side memory leak in the everything test - Altered the string packing code so that it's more obvious that the string is being copied. (rather than implicitly because the string is a const) |
|
From: Phil D. <ph...@us...> - 2000-09-26 06:28:49
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/09/25 23:28:48 Modified files: . : configure.in compiler : pass_xlate.cc types.cc types.hh orb : orbitcpp_smartptr.hh test/everything: client.cc everything-impl.cc Log message: - Added the array deep copy functionality - Added Array_var and Array_out types - Cleaned up array memory leaks |
|
From: Phil D. <ph...@us...> - 2000-09-21 09:08:40
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/09/21 02:08:39 Modified files: compiler : pass.hh types.hh orb : orbitcpp_exception.hh orbitcpp_tools.cc Log message: Added Mike Bond's gcc-2.96 fixes |
|
From: Phil D. <ph...@us...> - 2000-09-18 21:15:25
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/09/18 14:15:22 Modified files: . : acinclude.m4 configure.in compiler : types.cc orb : Makefile.am orbitcpp_constants.hh.in orbitcpp_orb.cc orbitcpp_orb.hh orbitcpp_poa.cc patches : 2000-03-09.orbit-patch test/everything: server.cc Added files: orb : orbitcpp_var_smartptr.hh Log message: - Added orbitcpp_var_smartptr.hh (sorry folks) - removed the configure test for the mem info block size - fixed the 'CORBA::OBJECT marshalling doesn't compile' thing - removed the ORB_instance caching stuff |
|
From: Phil D. <ph...@us...> - 2000-09-17 21:34:56
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/09/17 14:34:56 Modified files: . : HACKING compiler : pass_gather.cc pass_skels.cc pass_stubs.cc pass_xlate.cc pass_xlate.hh types.cc types.hh orb : Makefile.am orbitcpp.hh orbitcpp_object.hh orbitcpp_sequence.hh orbitcpp_smartptr.hh test/everything: client.cc everything-impl.cc everything-impl.hh everything.idl Log message: - Fixed the 'typedef'd strings don't initialise' bug which was making the name service test fail. - Made stub constructors protected to stop the compiler from printing warnings. (and commented this) - Commented out the const conversion operators - gcc doesn't use them and instead spews warnings telling you this. Has anybody tested with other compilers? - SequenceBase::operator=(SequenceBase const &src) was not returning itself. Fixed that. - Took out stub cast operators since (1) compiler (gcc 2.95.2) announces that won't be used by it and (2) even if they were, all casts are done with stub ptrs, not stub values so they'd never get called anyway - Created smart _ptr classes to handle implicit casts when multiple inheritence is involved to stop the compiler from thunking - Added better MI tests - Updated HACKING doc |
|
From: Phil D. <ph...@us...> - 2000-08-22 06:19:32
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/08/21 23:19:31 Modified files: compiler : pass_xlate.cc types.cc types.hh orb : orbitcpp_sequence.hh orbitcpp_tools.cc test/everything: client.cc everything.idl Log message: - Wrote some infrastructure to initialise compound elements without resorting to the operator new[] hackery. (check out IDLType::writeInitCode(), and look at the implementations in IDLString, IDLStruct and IDLArray) - Implemented corba 2.3 array initialisation (e.g. arrays of strings are initialised to "") - Removed the sequences operator new[] hack (hurrah!) in favour of this new code. (removed all the really dodgy complex stuff in orbitcpp_sequence.hh!) - Removed the thread-unsafe global variables surrounding the operator new[] hackage. |
|
From: Phil D. <ph...@us...> - 2000-08-20 09:45:21
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/08/20 02:45:21 Modified files: . : configure.in compiler : types.cc services/name : .cvsignore test/basic : client.cc test/everything: client.cc everything-impl.cc everything-impl.hh everything.idl Log message: - Added some code to fix the 'sequences maximum not marshalled in exceptions' bug in ORBit. - Added basic support for arrays with variable length elements - Bumped up the version number - Removed the orb->shutdown() code in the basic test (causes it to crash - is this supposed to work from within a corba method call?) |
|
From: Phil D. <ph...@us...> - 2000-08-06 11:00:33
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/08/06 04:00:32 Modified files: compiler : types.cc types.hh test : Makefile.am test/everything: client.cc constants.hh everything-impl.cc everything-impl.hh everything.idl Log message: basic support for fixed length type arrays |
|
From: Andreas K. <in...@us...> - 2000-07-22 15:12:19
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: inducer 00/07/22 08:12:18 Modified files: . : FAQ Log message: Fixed some CVS screw-ups. |
|
From: Andreas K. <in...@us...> - 2000-07-21 11:36:15
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: inducer 00/07/21 04:36:14 Modified files: . : FAQ README Log message: Updated FAQ and README to better support newbies. |
|
From: Phil D. <ph...@us...> - 2000-06-20 21:52:27
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/06/20 14:48:45 Modified files: compiler : pass_skels.cc types.hh test/everything: client.cc constants.hh everything-impl.cc everything-impl.hh everything.idl server.cc Log message: implemented polymorphism for attributes |
|
From: Phil D. <ph...@us...> - 2000-06-20 06:39:57
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/06/19 23:36:25 Modified files: compiler : language.cc pass_skels.cc pass_xlate.cc Log message: Compiler fixes: - Fixed indentation in cpp-common.hh files - Fixed bug which causes crashes when using inheritance and base interface has been forward declared. (if anyone's interested, the forward dcl was being found by the derived interface instead of the base interface proper, so none of the delegation methods were being written. The fix replaces the compiler forward dcl inteface with the full declared interface as soon as the compiler passes over it.) |
|
From: Phil D. <ph...@sl...> - 2000-06-11 10:02:59
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/06/11 02:59:57 Modified files: . : HACKING compiler : language.hh pass_xlate.cc types.cc types.hh orb : orbitcpp_sequence.hh orbitcpp_smartptr.hh orbitcpp_tools.cc orbitcpp_tools.hh test/everything-typedef: everything.idl Log message: - Brought orbitcpp memory management in line with new ORBit stuff. - Fixed the IDL compiler so that it compiles the name service. |
|
From: Phil D. <ph...@sl...> - 2000-06-03 20:21:42
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/06/03 13:19:19 Modified files: compiler : language.hh pass_gather.cc pass_skels.cc pass_stubs.cc pass_xlate.cc orb : orbitcpp_exception.cc orbitcpp_exception.hh orbitcpp_exceptionlist.hh test/everything: client.cc constants.hh everything-impl.cc everything-impl.hh everything.idl Log message: - Added support for attributes - moved _this() method into skels directly (i.e. instead of through an inherited 'mixin' class), since MI was making calls to _this() ambiguous. - Added MI polymorphism tests to the everything test. - Added a proprietary _orbitcpp_get_name method to the SystemException class |
|
From: Phil D. <ph...@sl...> - 2000-05-31 21:00:59
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/05/31 13:58:45 Modified files: compiler : pass_skels.cc pass_stubs.cc pass_xlate.cc types.cc types.hh orb : .cvsignore orbitcpp_exception.cc orbitcpp_exception.hh orbitcpp_object.cc orbitcpp_object.hh orbitcpp_orb.cc orbitcpp_orb.hh orbitcpp_poa.cc orbitcpp_poa.hh orbitcpp_sequence.hh orbitcpp_smartptr.hh orbitcpp_tools.cc orbitcpp_tools.hh orbitcpp_types.cc orbitcpp_types.hh test/basic : client.cc Log message: Removed _orbitcpp::c:: namespace for ORBit types. orbit-idl generated code still resides in _orbitcpp::c:: to resolve name clashes This was lots more work than expected! - mainly because the compiler now has to deal with 2 c namespaces (root and _orbitcpp::c) Compiler Changes: Added the virtual method IDLType::getNSScopedCTypeName() which returns the typename, scoped in _orbitcpp::c:: if appropriate (i.e. if compiler generated type). getCTypeName() just returns the C type name. Made loads of compiler functions call it. |
|
From: Phil D. <ph...@sl...> - 2000-05-31 06:09:39
|
CVSROOT: /cvsroot/orbitcpp Module name: orbitcpp Changes by: philipd 00/05/30 23:07:36 Modified files: orb : orbitcpp_object.hh orbitcpp_tools.cc orbitcpp_tools.hh Log message: Misc fixes to get things working with the new CVS ORBit. Basic test fails - need to investigate that. (Removed orbitcpp_get_repoid() - it's not needed and was breaking the build) |
