* add temporary compile test shell script
* Extended support of SKIP_VERSION to make invariant compile results during testing.
* build_keyboard.mk, tmk_core/rules.mk: add LIB_SRC, QUANTUM_LIB_SRC support
Support compiled object enclosed in library.
e.g.
```
LIB_SRC += xxxx.c
xxxx.c --> xxxx.o ---> xxxx.a
```
* remove 'ifdef/ifndef USE_I2C' from quantum/split_common/{i2c|serial}.c
* add SKIP_DEBUG_INFO into tmk_core/rules.mk
When SKIP_DEBUG_INFO=yes is specified, do not use the -g option at compile time.
* tmk_core/rules.mk: Library object need -fno-lto
* add SKIP_DEBUG_INFO=yes
* remove temporary compile test shell script
* add '#define SOFT_SERIAL_PIN D0' to keyboards/lets_split/rev?/config.h
* quantum/split_common/serial.c: Changed not to use USE_I2C.
* Refactor fourier to use split common code
* Refactor fourier to use split common code - fix handedness when not using EE_HANDS or MASTER_LEFT
* add SOFT_SERIAL_PIN config
* Add delay in Tap Code to avoid issues
I think a few people have reporting issues with it working properly, and it may be a timing issue. The 'register_code' uses this sort of delay in some of the functions, and
this is probably why.
Adding the 100ms delay should hopefully fix any issues with it.
* Make tap_code delay configurable
* Update documentation
* Bring tap_code16 inline with changes
* Fix type for tap_code16
Bad copy-paste job
* Just use the value check for the define
* Clarify timing in docs
Co-Authored-By: drashna <drashna@live.com>
* Wordsmithing
Co-Authored-By: drashna <drashna@live.com>
* handwired/CMD60: refactor
- renamed layout macro KEYMAP to LAYOUT
- removed K2C, K3B and K3C locations (indications are these locations were unused in hardware)
- keymap
- now uses #include QMK_KEYBOARD_H
- converted keycodes to short format
- deleted TMK fn_actions and action_get_macro blocks
- white space changes (readability/alignment)
* handwired/CMD60: Configurator support
* handwired/CMD60: readme update
- updated to conform to current QMK template
- added image and build gallery links
* handwired/cmd60: rename all files and folders to lowercase
* handwired/maartenwut: readme cleanup
- Fixed typo (Maarten's last name was misspelled)
- Updated Docs links
* handwired/maartenwut: Configurator support
* S7 Elephant: Configurator bugfix, part 1
Reformatted the JSON so each key is on its own line:
This makes the keys easier to rearrange.
* S7 Elephant: Configurator bugfix, part 2
Deleted the extra key from LAYOUT_iso_1u:
JSON contents stated this layout used a split Backspace, but the matrix is configured and intended to have a 2u Backspace.
Extended the Backspace to 2u width and deleted the extra key. (This fixes the error that broke LAYOUT_iso_1u's rendering in the Configurator.)
* S7 Elephant: Configurator bugfix, part 3
Fix ISO layouts:
Noticed JSON object for Enter on ISO layouts was immediately after Right Bracket. Inspection of s7_elephant.h revealed the expected and appropriate location to be after ISO Hash.
Moved Enter key to the proper location on all ISO layouts.
* Add standard ALGR defition, remove (re)definitions from language files
* Use ALGR(kc) consistently in ALTGR(kc) aliases
* Non-Nordic keymaps should not use NO_ALGR
* Add standard KC_ALGR definition
* Update docs with ALGR and KC_ALGR
* Update SS_ALGR and ALGR_T aliases
* handwired/numpad20: refactor
- layout macro no longer auto-prepends keycodes with KC_
- keymaps for this keyboard will now compile in QMK Configurator
- keymap now uses #include QMK_KEYBOARD_H
- deleted unused fn_actions code block
* handwired/numpad20: Configurator support
* Preonic: refactor layout macros
Unify layout macro names across AVR and ARM.
- all layouts and matrices have moved/duplicated to their appropriate revisions
- preonic.h now includes the appropriate header file from the selected revision
- revision header files only contain layouts appropriate for that revision
- Previously, all layouts were available in source for all revisions.
- removed KEYMAP layout macro alias (was unused)
- macros LAYOUT_preonic_1x2uC and LAYOUT_preonic_grid for AVR and ARM revisions
- LAYOUT_preonic_grid has alias LAYOUT_ortho_5x12 for all revisions
- deleted LAYOUT_ortho_4x12 alias (not appropriate for this keyboard)
- ARM-based revisions have 1x2uR, 1x2uL and 2x2u matrices
* Preonic: updated revision headers to #pragma once
* Preonic: update Configurator files
Updated the Configurator files to reflect the previous changes in this branch.
- Preonic revs. 1 and 2 will use keyboards/preonic/info.json
- Preonic rev. 3 will use keyboards/preonic/rev3/info.json
- This file will override keyboards/preonic/info.json when preonic/rev3 is selected in the Configurator.
* handwired/practice60: Configurator Support
* handwired/practice60: refactor
- layout macro rows did not have an equal number of arguments in all rows of the matrix
- white space changes for readability/alignment
* handwired/pilcrow: refactor
- layout macro renamed from KEYMAP to LAYOUT
- keymap now uses #include QMK_KEYBOARD_H
- layers reformatted for readability
- removed unused and deprecated fn_actions and action_get_macro blocks
- keymap config.h
- updated to use #pragma once
- removed redundant config.h include
* handwired/pilcrow: Configurator support
* handwired/pilcrow: readme update
Updated readme.md file to use modern template formatting.
On a laptop with god knows what mandatory security software (Cylance?), running up-to-date Windows 10 with msys2 mingw-64, attempting to install the AVR toolkit results in the following error:
```
1 [main] 7z (13316) C:\msys32\usr\lib\p7zip\7z.exe: *** fatal error - cygheap base mismatch detected - 0x612A5410/0x2375410.
This problem is probably due to using incompatible versions of the cygwin DLL.
Search for cygwin1.dll using the Windows Start->Find/Search facility
and delete all but the most recent version. The most recent version *should*
reside in x:\cygwin\bin, where 'x' is the drive on which you have
installed the cygwin distribution. Rebooting is also suggested if you
are unable to find another cygwin DLL.
```
This appears to be related in some way, based on my research, to ASLR functionality in security software. Since I'm unable to override whatever is enforcing ASLR on my system, after trying several other approaches (removing other copies of msys-2.0.dll, which is what this is apparently actually referencing, rebasing that file in Windows to address 0x61000000, a few other things) I simply edited the installation shell script to use `unzip` instead of 7zip; `unzip`'s binary does not provoke a mismatch error and the installation proceeds as it should.
I'm not aware of the reason why some parts of the install script use `unzip` (e.g. `install_arm`) and others use 7zip, but it seems that for broader compatibility and sparing users on locked down machines the 120 minutes or so of futzing this took me to fix, it might be better to just use `unzip` in all cases.
Note: There is another function that uses 7zip, `extract_flip`. The line is `7z -oflip x FlipInstaller.exe`. I'm not sure what this is doing, or whether it's possible to do it with `unzip`, but it produces the same error. I haven't attempted to fix that in this PR, but it might be good to fix it for the same reason.
Takeshi ISHII