summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* RISC OS: Bitmap modified flag is not used.Michael Drake2022-03-241-1/+0
|
* Include: Bitmap: Fix comment typo.Michael Drake2022-03-241-1/+1
|
* Include: Bitmap: Drop get_bpp from bitmap operation table.Michael Drake2022-03-241-7/+0
| | | | Core bitmaps have always been 32bpp.
* BeOS: Drop bitmap get bpp function.Michael Drake2022-03-241-15/+0
|
* Atari: Drop bitmap get bpp function.Michael Drake2022-03-241-12/+1
|
* Amiga: Drop bitmap get bpp function.Michael Drake2022-03-241-16/+1
|
* GTK: Drop bitmap get bpp function.Michael Drake2022-03-241-17/+0
|
* Framebuffer: Drop bitmap get bpp function.Michael Drake2022-03-241-7/+0
|
* Monkey: Drop bitmap get bpp function.Michael Drake2022-03-241-7/+0
|
* Windows: Drop bitmap get bpp function.Michael Drake2022-03-241-6/+0
|
* RISC OS: Drop bitmap get bpp function.Michael Drake2022-03-241-15/+0
|
* GUI: Bitmap bytes per pixel callback is not required now.Michael Drake2022-03-241-4/+0
|
* Image: ICO: Update for new libnsbmp API.Michael Drake2022-03-241-1/+0
|
* Image: BMP: Update for new libnsbmp API.Michael Drake2022-03-241-3/+1
|
* PNG: Bytes per pixel is always 4.Michael Drake2022-03-241-1/+1
|
* GIF: Update to new libnsgif API.Michael Drake2022-03-231-1/+2
| | | | Set required colour component order.
* GIF: Tolerate errors during scan if at least one frame.Michael Drake2022-03-031-5/+9
|
* GIF: Update to latest libnsgif API.Michael Drake2022-02-261-206/+160
|
* Amiga: Don't open/close dos.library ourselves on OS3Chris Young2022-01-261-2/+2
| | | | Suspect this is the cause of crash on exit
* GTK: Add support for word left/right keybindings.Michael Drake2022-01-231-0/+4
|
* textarea: always clear selection on NS_KEY_WORD_{LEFT,RIGHT}Pranjal Kole2022-01-231-6/+8
| | | | | | | | | | | | | This bug can be seen by selecting some text starting from the beginning of a textarea (so that caret is 0) and then pressing the NS_KEY_WORD_LEFT binding. NS_KEY_WORD_LEFT was breaking early when caret was 0. So, to always clear the selection, the clear selection code has been brought above the break statement. NS_KEY_WORD_RIGHT did not have such a break statement, so one has been added for consistency, and because string operations are expensive.
* GTK: Add support for word delete left/right.Michael Drake2022-01-221-0/+4
| | | | These are ctrl+delete and ctrl+backspace.
* textarea: implement NS_KEY_DELETE_WORD_{LEFT,RIGHT}Pranjal Kole2022-01-222-6/+95
| | | | | | | | | | | | | | | | | | | | | NS_KEY_DELETE_WORD_{LEFT,RIGHT} have been added to include/netsurf/keypress.h and implemented in desktop/textarea.c An unsigned int, caret_copy, has been added since both of these require a temporary variable to hold the original position of the caret. The LEFT one deletes separators towards the left till it encounters a non-separator and then deletes the non-separators until it encounters a separator. The caret is moved towards the left by the number of characters deleted. The RIGHT one does the same towards the right, but the caret is kept at its original position. These are intended to be mapped to Ctrl+Backspace and Ctrl+Delete by most frontends. Additionally, some style and typo fixes have been made.
* Amiga: Add ARexx scripts to play videos via YTChris Young2021-12-173-0/+61
| | | | Credit: Chris Handley
* Update ca-bundleChris Young2021-10-041-673/+361
|
* CSS: Selection callbacks: Update to latest LibDOM API.Michael Drake2021-07-051-0/+3
|
* Revert "Amiga: do not force 32-bit bitmaps"Chris Young2021-06-286-13/+14
| | | | This reverts commit 793f514220c15c6bb1b0db31bf0b6d0cfea0fc7b.
* Amiga: do not force 32-bit bitmapsChris Young2021-06-286-14/+13
| | | | This appears to be unnecessary
* Amiga: report correct depthChris Young2021-06-281-3/+0
|
* AMIGA: Use RTG calls on OS3Chris Young2021-06-281-10/+7
| | | | Enabled for experimentation
* RISC OS: GUI: Clean up alpha sprite support test.Michael Drake2021-06-251-12/+10
|
* RISC OS: Image: Tinct workaround: Use OS alpha sprite rendering.Michael Drake2021-06-252-6/+67
| | | | | | | | | If NetSurf is configured to use OS for image rendering, and the OS supports Alpha sprites, avoid going via Tinct completely. Going via Tinct loses the alpha channel. However, with this workaround, we lose Tinct's pretiling optimisation for tiling tiny sprites.
* RISC OS: GUI: Add check for sprite alpha channel OS support.Michael Drake2021-06-252-0/+31
|
* RISC OS: Image: Add support for tiled plots with OS renderer.Michael Drake2021-06-251-6/+15
|
* RISC OS: Image: Remove forward declaration.Michael Drake2021-06-251-71/+63
|
* RISC OS: Image: Use #define to enable use of mask.Michael Drake2021-06-251-1/+1
|
* RISC OS: Image: Use #define for unspecified sprite area.Michael Drake2021-06-251-3/+3
|
* Switch to new libcss API for unit conversion.Michael Drake2021-05-3024-883/+542
|
* CSS: Update to latest libcss: Remove weird units.Michael Drake2021-05-303-46/+2
| | | | The 'rlh', 'ic' and 'cap' units were never implemented by anyone.
* Amiga: BitMap always alloced with the ami_rtg_ function, so should be freed ↵Chris Young2021-05-171-5/+1
| | | | with it too.
* Windows: Use pkg-config when building on Windows.Michael Drake2021-05-131-1/+1
|
* RISC OS: Unify both browser window redraw call paths.Michael Drake2021-04-261-67/+54
|
* RISC OS: Constify redraw region through buffer API.Michael Drake2021-04-262-2/+2
|
* RISC OS: Corewindow: Fix autoscroll segfault when pointer leaves window.Michael Drake2021-04-252-0/+13
|
* RISC OS: Fix EX0 EY0 rendering glitches when scrolling.Michael Drake2021-04-254-12/+42
|
* RISC OS: Plotters: On RISC OS y1 is top and y0 is bottom.Michael Drake2021-04-251-11/+16
| | | | When storing RISC OS coordinates, use y0 and y1 consistently.
* Amiga: Fix codesets.library conversion when string specified as zero lengthChris Young2021-04-121-6/+5
|
* Amiga: The return value of Codesetsfind was not being assigned to a variableChris Young2021-04-111-2/+2
|
* hoist the Referer header generation logic up to the low level cacheVincent Sanders2021-03-255-104/+152
| | | | | | | This removes the need for the fetchers to have any interaction with the Referer header. It has not been completely removed from the fetch interface as fetch.c:fetch_set_cookie() still uses it for unverifiable cookie decision logic. (There is an anchient todo here)
* extend teh implementing a new frontend documentVincent Sanders2021-03-231-9/+70
|
generated by cgit v1.2.3 (git 2.39.1) at 2024-06-17 08:46:07 +0000