summaryrefslogtreecommitdiffstats
path: root/generic/tkCanvLine.c
Commit message (Collapse)AuthorAgeFilesLines
* size_t for Tk_ItemDCharsProc(), with Tcl 9 headersjan.nijtmans2020-06-021-6/+6
|
* Convert Tk_ItemIndexProc/Tk_ItemInsertProc to use size_t as index type, when ↵jan.nijtmans2020-05-261-16/+16
| | | | compiled against Tcl 9 headers. Various more fixes in using size_t for Tcl >= 9
* Formatting and even more size_t usage.jan.nijtmans2020-05-251-2/+2
|
* More progressjan.nijtmans2020-05-221-9/+9
|
* Fix [3134d697c8]: Reduce unnecessary code checks in canvas postscript code ↵fvogel2020-03-291-25/+8
| | | | (patch from Emiliano Gavilan).
* More usage of (internal) TkGetIntForIndex() function, and give this function ↵jan.nijtmans2020-03-201-1/+1
| | | | | | an additional parameter "lastOK". This way, "end" doesn't need to be handled especially in other code places, all is centralized in this single functions. Only Ttk's "treeview" not done yet, since that requires handling a linked list in stead of an array.
* More internal use of TkGetIntForIndex() function.jan.nijtmans2020-03-181-25/+25
|
* TIP 569 remove lint commentsdgp2020-03-181-5/+0
|
* Merge trunkjan.nijtmans2020-02-171-2/+1
|\
| * Clamp output of listbox indexes to be >= -1 always. Many other's (like ↵jan.nijtmans2020-02-171-2/+1
| | | | | | | | menu's, entries) already do that. Doesn't make really a difference, as -2 is just as invalid as -1.
* | Much more WIP: Appears to be working on UNIX, and for a bit part on Windows too.jan.nijtmans2019-12-201-16/+32
|/
* Merge 8.6jan.nijtmans2019-10-061-1/+1
|\
| * Minor simplification: DEF_CANVLINE_FILL and DEF_CANVTEXT_FILL are each only ↵jan.nijtmans2019-10-061-1/+1
| | | | | | | | used once, and always equal to DEF_CANVITEM_OUTLINE. So better use DEF_CANVITEM_OUTLINE directly.
| * Too quick conclusion: all lines and text were gone :-). Reverting most of ↵jan.nijtmans2019-10-041-1/+1
| | | | | | | | previous commit
| * Remove DEF_CANVLINE_FILL/DEF_CANVTEXT_FILL, since the default fill value ↵jan.nijtmans2019-10-041-1/+1
| | | | | | | | | | (according the documentation) is {}. This has no visible effect. Modify canvas items demo, so it can be run on wish8.5 (for visual comparison)
* | Fix Tk_Offset() -> offsetof(). Tk_Offset() was inherited by merging ↵fvogel2019-09-291-1/+1
| | | | | | | | chavez-canvas branch that originally targeted core-8-6-branch.
* | Merge chavez-canvas branchKevin Walzer2019-09-291-1/+2
|\ \ | |/
| * Branch to test Christopher Chavez patch in ticket [2beda17141] to remove ↵Kevin Walzer2019-09-241-1/+2
| | | | | | | | some hard-coded values from canvas, allow canvas to vary apperance in Dark Mode on macOS and also on Windows
* | Merge trunkjan.nijtmans2019-05-101-1/+54
|\ \
| * | Document. Factor out basic rotation function.dkf2019-03-131-16/+3
| | |
| * | More item typesdkf2019-03-111-1/+67
| |/
* | Merge 8.7jan.nijtmans2019-01-131-6/+6
|\ \ | |/
| *-. Merge 8.6 and all changes from original [bug-9e31fd9449] branchjan.nijtmans2019-01-081-13/+13
| |\ \
| | | * Fix [9e31fd944934e269121fa78ff56b7b86f33e6db6|9e31fd9449]: X11/X.h and ↵jan.nijtmans2019-01-081-7/+7
| | | |\ | | | | | | | | | | | | | | | | | | | | | | | | | Windows.h have conflicting symbols. *** POTENTIAL INCOMPATIBILITY *** on Windows only: gcc/clang/MSVC will generate new warnings in extensions when the "None" symbol is used incorrectly. Those warnings are all fixed in the core, that's what most of this commit is doing.
| | | | * Bring back more original "None" usages, and fix other warnings which ↵jan.nijtmans2019-01-031-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | gradually slipped in. Wherever possible, pragma's are used in MSVC to silence useless compiler warnings.
| * | | | As requested by Christian Werner/François Vogel and others: Undo many None ↵jan.nijtmans2018-12-311-19/+19
| |\ \ \ \ | | | |_|/ | | |/| | | | | | | | | | | | -> 0 changes, in order to reduce the probability of merge conflicts with other branches. Remark: Many usages of "None" in Tk are wrong, "NULL" should be used in many places where pointers are referenced in stead of XID's. Those places are corrected.
| | * | | As requested by Christian Werner/François Vogel and others: Undo many None ↵jan.nijtmans2018-12-311-27/+27
| | | | | | | | | | | | | | | | | | | | | | | | | -> 0 changes, in order to reduce the probability of merge conflicts with other branches. Remark: Many usages of "None" in Tk are wrong, "NULL" should be used in many places where pointers are referenced in stead of XID's. Those places are corrected.
| * | | | Fix [9e31fd9449]: X11/X.h and Windows.h have conflicting symbols jan.nijtmans2018-12-201-23/+23
| |\ \ \ \ | | |/ / / | | | / / | | |/ / | |/| | *** POTENTIAL INCOMPATIBILITY *** for Win32 only: On X11 and Mac, "None" can still be used as before
| | * | Fix [9e31fd944934e269121fa78ff56b7b86f33e6db6|9e31fd9449]: X11/X.h and ↵jan.nijtmans2018-12-201-29/+29
| | |/ | | | | | | | | | | | | Windows.h have conflicting symbols. Also fix a few newer (harmless) gcc warnings.
| | * [Bug 3175610]: Incomplete refresh of line items.dgp2011-06-101-2/+3
| | |\
| | | * [Bug 3175610]: Incomplete refresh of line items.dgp2011-06-101-2/+5
| | | |
* | | | Minor tweaks, mostly formattingjan.nijtmans2018-12-011-1/+1
| | | |
* | | | Experiment: let's see if we can deprecate Tk_Offset() in favor of offsetof()jan.nijtmans2018-10-061-22/+22
|/ / /
* | | some more int -> size_t changesjan.nijtmans2013-02-011-3/+2
| | |
* | | Working towards making the error codes more consistent.dkf2012-07-301-3/+3
| | |
* | | Postscript generation now uses Tcl_Objs more extensively; still nasty in places.dkf2012-07-271-32/+79
| | |
* | | consistent error formatting, added error codesdkf2012-07-211-13/+22
| | |
* | | Working towards adding all the Tcl_SetErrorCode calls that should be there.dkf2012-07-161-50/+48
| | | | | | | | | ** WORK IN PROGRESS **
* | | Purge RCS Keywordsdgp2011-06-081-2/+0
|\ \ \ | |/ /
| * | Purge RCS Keywords.dgp2011-06-081-2/+0
| |\ \ | | |/
| | * Purge RCS Keywords.dgp2011-06-081-2/+0
| | |
| | * * changes: updates for 8.4.14 release.das2006-10-161-1/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * macosx/tkMacOSXDraw.c: fix numerous issues in CG and QD drawing procs so that they now match X11 drawing much more closely [Bug 1558051]; use Tiger ellipse drawing API when available; fix comments & whitespace. * macosx/tkMacOSXInit.c: set default linewidth limit for CG antialiasing to 0 as thin horizontal/vertical lines look good now. * generic/tkCanvLine.c (ConfigureLine): on TkAqua, pass outline * generic/tkCanvPoly.c (ConfigurePolygon): linewidth in gc even for * generic/tkRectOval.c (ConfigureRectOval): fills (as it controls AA).
| | * Fix [Bug 1252702]dkf2005-08-111-4/+4
| | |
| * | * generic/tkCanvLine.c (LineDeleteCoords): [Bug 2900121]: Backport ofdkf2010-08-111-23/+34
| | | | | | | | | | | | fix to sense of test.
* | | Remove casts from uses of ckalloc/ckfree/... now that Tcl declares them to bedkf2011-03-121-41/+33
| | | | | | | | | using useful casts internally.
* | | [Bug 3175610]: Incomplete refresh of line items.ferrieux2011-03-031-2/+3
| | |
* | | Revise Tcl_Panic() calls ending with a newline removing the newline, because ↵nijtmans2010-11-191-2/+2
| | | | | | | | | | | | Tcl_Panic() outputs a final newline already
* | | Eliminate many unnecessary (ClientData) type casts.nijtmans2010-06-151-9/+9
| | |
* | | CONSTify everything related to Tk_ConfigSpecnijtmans2010-02-171-2/+2
| | |
* | | fix more gcc warnings: missing initializernijtmans2010-01-181-13/+13
| | | | | | | | | | | | | | | Apply patch [#2932808]: Canvas items not updating on widget state change
generated by cgit v0.12 at 2024-09-23 02:19:56 (GMT)