summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Added cast to work around stupid type inference bug in MSVC. Apparently itdkf2008-11-031-2/+2
| | | | can't tell the difference between 'const char **' and 'char *const *'...
* Forgot to check this in due to local network woes.dkf2008-11-021-43/+52
|
* more internal -Wwrite-strings warning fixesnijtmans2008-11-028-58/+75
|
* Cleanup and document TK_MOVABLE_POINTSdkf2008-11-011-78/+167
|
* Implementation of TIP 97 - canvas imove and rchars methods.dkf2008-11-018-904/+1326
|
* Implemented the themed spinbox widget ttk::spinbox.patthoyts2008-11-0115-83/+933
|
* Minor formatting/usage correctiondkf2008-11-011-2/+3
|
* Temporary workaround for [Bug 2207435]jenglish2008-10-312-1/+6
|
* more internal -Wwrite-strings warning fixesnijtmans2008-10-3017-80/+91
|
* * tests/unixSelect.test: Revise the unixSelect-1.* tests so thatdgp2008-10-302-67/+49
| | | | | | | they test the ability of Tk's selection mechanism to faithfully pass valid Tcl values without corruption, and stop testing details of Tcl's internal encoding scheme. With this change, the Tk test suite no longer uses the identity encoding or [string bytelength].
* [Bug 2190619] Warnings due to Tk_SmoothMethodnijtmans2008-10-3010-83/+79
| | | | | | | | | | | | | | | | name constness change. Although dkf's solution (simply adding a type cast) is correct as well, changing the return value (as das suggested) has the advantage that all Tk_OptionPrintProc implementations don't need a type cast any more in its code. This makes all those functions robust against the -Wwrite-strings warning option. Further on, the customPtr field of Tk_ConfigSpec can be a constant, without any danger of source incompatibility.
* Remove incorrect 'const' qualifier.jenglish2008-10-292-7/+12
| | | | Remove useless 'register' declarations too, while we're at it.
* CONST updatedgp2008-10-291-2/+2
|
* formattingdgp2008-10-291-17/+17
|
* CONSTify TkPrintPadAmount, TkCreateFrame, TkCreateMainWindownijtmans2008-10-2820-97/+132
| | | | | and Tk_ParseArgv. move TkSelGetSelection to internal stub table (needed in tkWinTest.c)
* Expanded set of symbolic cursors.jenglish2008-10-287-37/+217
| | | | | Add correct platform-specific cursors for OSX [Bug 2054562] Use correct cursor for ttk::entry and ttk::combobox widgets [Bug 1534835]
* * win/tkWinTest.c: Revise [testclipboard] to form thatdgp2008-10-284-28/+23
| | | | | * tests/winClipboard.test: handles encodings. [Bug 2191960] * tests/constraints.tcl: [tcltest::bytestring] no longer used.
* Tweaking for better readability.dkf2008-10-272-71/+92
|
* Code style improvements, especially including making code less deeply nested.dkf2008-10-2710-1940/+2014
|
* Disable test ttk-6.3, it's not applicable [Bug 2175411].jenglish2008-10-243-16/+15
| | | | | | Use different Tcl_AssocData key for Ttk so the tile extension can be loaded into an 8.6 interp, in the off-chance that anyone wants to do this.
* Silence warning.dkf2008-10-242-4/+10
|
* * README: Bump version number to 8.6a4dgp2008-10-239-21/+33
| | | | | | | | | | | * generic/tk.h: * library/tk.tcl: * unix/configure.in: * unix/tk.spec: * win/configure.in: * unix/configure: autoconf-2.59 * win/configure:
* fix TkX11 builddas2008-10-231-2/+2
|
* Letting CONST die a slow and graceful death, since NO_CONST wasnijtmans2008-10-2212-896/+910
| | | | broken since 8.4 and no-one complained about it.
* add tkBusy.cdas2008-10-221-1/+7
|
* fix GenericEvent Tcl_Panic errordas2008-10-221-2/+2
|
* Reduce number of castsdkf2008-10-201-2/+2
|
* Correct mistake in [tk busy configure].dkf2008-10-201-27/+27
|
* fix compile errors and warningsdgp2008-10-201-9/+10
|
* Factor out the platform-specific bits of [tk busy]. [Bug 2180919]dkf2008-10-2010-428/+617
|
* msvc uses __inline, so added the necessary macro now that we are using thispatthoyts2008-10-191-3/+4
|
* Implementation of the [tk busy] command on non-OSX.dkf2008-10-1814-50/+1988
| | | | Adapted from [Patch 1997907]
* [Bug 1825353] To fix a problem with tiny fonts on Russian versions ofpatthoyts2008-10-182-3/+22
| | | | | Windows we will avoid removing the internal leading for fixed width fonts. See the bug for more details on why this change is as it is.
* Avoid const jiggling by using Tcl_SetObjResultpatthoyts2008-10-181-2/+3
|
* Add "const" to many internalnijtmans2008-10-1751-347/+397
| | | | | | | | const tables, so those will be put by the C-compiler in the TEXT segment in stead of the DATA segment. This makes those table sharable in shared libraries.
* fix missing "-Wl," prefix in HP-UX buildnijtmans2008-10-172-2/+2
|
* fix missing "-Wl," prefix in HP-UX buildnijtmans2008-10-172-2/+2
|
* Implemented keyboard bindings for ttk::scalepatthoyts2008-10-172-6/+46
|
* Add "const" to many internalnijtmans2008-10-1512-35/+49
| | | | | | | | const tables, so those will be put by the C-compiler in the TEXT segment in stead of the DATA segment. This makes those table sharable in shared libraries.
* Added new utility function: TkNewWindowObj()dkf2008-10-143-107/+141
|
* Correct error result generation of [$canvas select clear $bogus].dkf2008-10-113-4/+14
|
* *** 8.6a3 TAGGED FOR RELEASE ***core_8_6_a3dgp2008-10-102-1/+24
| | | | * changes: Updates for 8.6a3 release.
* * tests/entry.test: Restore test naming consistency with Tk 8.5.dgp2008-10-102-293/+297
| | | | | * tests/listbox.test: Remove some more dependency on precision in * tests/spinbox.test: test results.
* removed bogus commentdkf2008-10-091-3/+3
|
* * generic/tkListbox.c: Make literal return values consistent withdgp2008-10-093-74/+90
| | | | | | | | those generated by Tcl_PrintDouble(). * tests/entry.test: Restore test naming consistency with Tk 8.5. * tests/listbox.test: Remove some more dependency on precision in test results.
* Restored test naming consistency with Tk 8.5 tests.dgp2008-10-091-114/+123
|
* comment formattingdkf2008-10-091-4/+3
|
* * unix/tcl.m4: fix for bug [2073255]nijtmans2008-10-083-22/+27
| | | | * unix/configure: regenerated
* * tests/textDisp.test (textDisp-16.34): Update test that testeddgp2008-10-082-8/+18
| | | | | string equality of double values based on an assumption of tcl_precision==12. Test now does its own formatting.
* typodgp2008-10-081-1/+1
|