summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* * Improve socket.test by checking the latency on the loopback address and ↵max2011-06-012-40/+63
| | | | | use that for some of the tests instead of fixed "big enough" times. * Improve correctness of [socket -async] in some error cases.
* * Fix setting up of [fileevent] while an async socket is still in progress max2011-05-302-44/+96
| | | | * Cache async socket errors for later use by [fconfigure -error] * Add tests for the above
* Fix [socket -async] for DNS names with more than one addressmax2011-05-273-98/+163
|
* fix a timing issue in socket-12.3max2011-05-271-1/+2
|
* committed missing tzdata/Africa/Cairo from venkat's last commitKevin B Kenny2011-05-251-178/+0
|\
| * committed missing tzdata/Africa/Cairo from venkat's last commitKevin B Kenny2011-05-251-178/+0
| |
* | Bump to msgcat 1.4.4dgp2011-05-255-6/+13
|\ \ | |/
| * Bump to msgcat 1.4.4.dgp2011-05-255-30/+23
| |
| * silence compiler warningdgp2011-05-251-1/+1
| |
* | Bump TclOO version.corresponds_to_TclOO_0_6_3dkf2011-05-256-5/+7
| |
* | Implementation of TIP #381: Call Chain Introspection and Controldkf2011-05-2511-15/+1313
| |
* | Update to Olson tzdata2011gvenkat2011-05-241-0/+4
|\ \ | |/
| * Update to Olson tzdata2011gvenkat2011-05-241-0/+4
| |
* | Undo mistaken commitdkf2011-05-241-5/+2
| |
* | Remove some useless code from mcset and mcmset: [dict set] builds dictionary ↵dkf2011-05-243-16/+10
| | | | | | | | levels for us.
* | Merged core-8-5-branch, fixup of TIP 280 location mapping.andreask2011-05-173-0/+78
|\ \ | |/
| * * generic/tclCompile.c (TclFixupForwardJump): Tracked down and fixedandreask2011-05-173-0/+78
| | | | | | | | | | | | | | | | | | * generic/tclBasic.c (TclArgumentBCEnter): the cause of a violation of my assertion that 'ePtr->nline == objc' in TclArgumentBCEnter. When a bytecode was grown during jump fixup the pc -> command line mapping was not updated. When things aligned just wrong the mapping would direct command A to the data for command B, with a different number of arguments.
* | Merged 8.5 following http fix that is not needed on trunk.pat2011-05-120-0/+0
|\ \ | |/
| * [Bug 2715421] fix backported from 8.6 to remove extra newline from POSTs.patthoyts2011-05-124-7/+6
| | | | | | | | | | Package version incremented to 2.7.6. Signed-off-by: Pat Thoyts <patthoyts@users.sourceforge.net>
* | If your compiler already defines _WIN64, assume --enable-64bitjan.nijtmans2011-05-122-0/+73
|\ \ | |/
| * If your compiler already defines _WIN64, assume --enable-64bitjan.nijtmans2011-05-122-0/+73
| |\
| | * If your compiler already defines _WIN64, assume --enable-64bitjan.nijtmans2011-05-113-21/+71
| | |
* | | * unix/tclUnixSock.c (TcpWatchProc): No need to check for server sockets ↵max2011-05-113-18/+35
| | | | | | | | | | | | | | | here, as the generic server code already takes care of that. * tests/socket.test (accept): Add tests to make sure that this remains so.
* | | New internal routines TclScanElement() and TclConvertElement().dgp2011-05-1011-300/+829
|\ \ \ | |/ / | | | Rewritten guts of machinery to produce string rep of lists. [Bug 3173086]
| * | New internal routines TclScanElement() and TclConvertElement().dgp2011-05-1010-291/+820
| |\ \ | | | | | | | | Rewritten guts of machinery to produce string rep of lists. [Bug 3173086]
| | * | Completed patch with mucho comments. Merge 8.5.bug_3173086dgp2011-05-1095-1607/+3076
| | | |
| | * | merge trunk + partial improvements to commentsdgp2011-03-077-50/+96
| | | |
| | * | mergedgp2011-03-063-2/+45
| | |\ \
| | * | | Correct flaw in the rewrite handling &#91;list \\\\\}].dgp2011-03-032-1/+5
| | | | |
| | * | | Significant rewrite of the Tcl*(Scan|Convert)*Element() system, and revisionsdgp2011-03-035-242/+394
| | | | | | | | | | | | | | | | | | | | to the callers. Needs more work on comments, and testing to check for any performance impact in either direction. Fixes reported bug.
| | * | | Simplified the logic of the Scan/Convert pair of routines for generating dgp2011-03-031-21/+5
| | | | | | | | | | | | | | | string reps of lists to stop using the unnecessary flag BRACES_UNMATCHED.
* | | | | Reduce use of Tcl_AppendElement, which is not (and can't be) a Tcl_Obj-aware ↵dkf2011-05-094-37/+81
| | | | | | | | | | | | | | | | | | | | API.
* | | | | Revise empty string tests so that we avoid potentially expensive string repdgp2011-05-092-61/+16
|\ \ \ \ \ | |/ / / / | | | | | generations, especially for dicts.
| * | | | Revise empty string tests so that we avoid potentially expensive string repdgp2011-05-092-61/+16
| |\ \ \ \ | | | | | | | | | | | | generations, especially for dicts.
| | * | | | Remove the old implementation.dgp_list_simplifydgp2011-05-091-75/+0
| | | | | |
| | * | | | A different technique to more precisely identify the optimization case.dgp2011-05-091-14/+35
| | | | | |
| | * | | | Disable all the special case code in place to prevent allocation of a Listdgp2011-05-051-10/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | struct for empty lists as much as possible. The value of this avoidance is not clear to me. Hoping it can be demonstrated (or refuted) with both variants side by side in the code history tree.
* | | | | | Convert TclGetLoadedPackages to use Tcl_Obj API for result generation.dkf2011-05-072-40/+47
| | | | | |
* | | | | | fix USE_TCLALLOC so that it can be enabled without editing the Makefilemig2011-05-073-4/+13
|\ \ \ \ \ \ | |/ / / / /
| * | | | | fix USE_TCLALLOC so that it can be enabled without editing the Makefilemig2011-05-073-4/+13
| |\ \ \ \ \ | | | |_|_|/ | | |/| | |
| | * | | | fix USE_TCLALLOC so that it can be enabled without editing the Makefilemig2011-05-073-4/+13
| | | | | |
* | | | | | comment fixjan.nijtmans2011-05-061-3/+3
|\ \ \ \ \ \ | |/ / / / /
| * | | | | comment fixjan.nijtmans2011-05-061-3/+3
| |\ \ \ \ \ | | |/ / / / | | | / / / | | |/ / / | |/| | |
| | * | | comment fixjan.nijtmans2011-05-061-3/+3
| | | | |
* | | | | Stop generating string rep of dict when converting to list.dgp2011-05-052-9/+47
|\ \ \ \ \ | |/ / / / | | | | | Tolerate NULL interps more completely.
| * | | | Tolerate NULL interps more completely.dgp2011-05-052-5/+9
| | | | |
| * | | | Stop generating string rep of dict when converting to list.dgp2011-05-052-5/+39
| | | | |
* | | | | Fix typo spotted by Emiliano Gavilan.dkf2011-05-051-1/+4
| | | | |
* | | | | no longer depend on MODULE_SCOPE being definedjan.nijtmans2011-05-051-0/+3
| | | | |
* | | | | merge 8.5dgp2011-05-036-208/+116
|\ \ \ \ \ | |/ / / /