LVGL v6.0 Released - General Events and Multi-Display Support

Improvements#

General events#

In v5.3 you could add actions to only a few object types (like a button). In v6.0 you can do it for any object type using Events.

Use lv_obj_set_event_cb(obj, event_handler) to set an event callback for an object. In event_handler you will get the caller object and the event like LV_EVENT_CLICKED/PRESSED/VALUE_CHANGED etc.

Learn more in the Events documentation.

Better fonts#

A new font converter is created which supports kerning and some other features. It is written in Node.js so it should be easy to use offline.

You can choose more fonts and ranges to include them in the final font file.

• Online font converter
• Font converter source code

Multi-display support#

In v6.0 you can have multiple displays, each with its own drivers and objects.

To learn more about multi-display support visit the Display documentation.

MicroPython#

A lot of improvements happened to smoothly support MicroPython. Learn more about MicroPython with LittlevGL.

New forum#

Not really a LittlevGL feature but it happens along with v6.0. We have launched a new, more modern forum: https://forum.littlevgl.com

You can easily log in with your GitHub or Google account.

The old forum will be moved to http://forumold.littlevgl.com

Migrating from v5.3 to v6.0#

The most important things#

• Update your lv_conf.h from lv_conf_template.h
• Generate your fonts again with the new converter: https://littlevgl.com/ttf-font-to-c-array
• Read the new documentation: https://docs.littlevgl.com

Actions vs Events#

• lv_obj_set_event_cb(obj, callback) is used instead of lv_btn_set_action(btn, LV_BTN_ACTION_..., callback). Event callbacks are used to handle all the former actions (i.e. long pressed, released are handled by a single function).
• Objects now automatically detect if they've been deleted, no need for returning LV_RES_OK/LV_RES_INV in your lv_event_cb_t. For this reason, lv_event_cb_t functions return void.

Driver level#

• Driver function signatures now take an initial first parameter, the driver objects itself. This allows your driver functions to work with more than one physical device without requiring two separate drivers to be written.
• (i.e. indev callbacks now take an initial first parameter of lv_indev_t *)
• disp_flush is now flush_cb. Instead of directly taking x/y coordinates, it takes an lv_area_t argument which has x1, y1, x2, y2 fields
• The mem_fill (called gpu_fill_cb in v6.0) fills an area instead of a line
• Dynamically changeable display config. Instead of hardcoded in lv_conf.h.
• Need to explicitly initialize your display buffer and add it to your display driver. It replaces the VDB in v5.3. so there is no LV_VDB_SIZE in v6.0 but you should use these display buffers.
• LV_HOR_RES and LV_VER_RES are replaced with LV_HOR_RES_MAX and LV_VER_RES_MAX, which define the maximum horizontal and vertical resolution of your display, respectively. LV_HOR_RES and LV_VER_RES still work but always reflects the resolution of the lastly created display. (No difference with only one display).

Renames#

• All the macros like USE_LV_BTN now are like LV_USE_BTN.
• Rename symbols: SYMBOL_HOME → LV_SYMBOL_HOME
• Rename LV_GROUP_KEY_... to LV_KEY_...
• Several minor function API changes. Nearly all features are still present so you should find every v5.3 function but maybe with a slightly different name or usage

Animations#

• Renames in lv_anim_t. The ready_cb get's lv_anim_t * as argument.
• lv_anim_set_...(&a) functions are added to replace manual initialization of lv_anim_t
• lv_style_anim_t is removed. Use lv_anim_t a; and lv_style_anim_set_...(&a, ...) instead.

Objects and Styles#

• lv_obj_get_free_ptr and lv_obj_get_free_num are replaced by lv_obj_get_user_data. The type of the user data can be changed in lv_conf.h
• All padding and fits are left, right, top, bottom instead of hor and ver.
• lv_canvas: use lv_draw_... functions

Fonts#

• Font system is updated and now supports kerning. You can generate your fonts again with the new converter.
• Use the new online font converter to generate new fonts
• Built-in fonts have changed

Others#

• User data is now available for some other structures as well, like driver objects.
• Callbacks of lv_task now receive a pointer to the task itself instead of the user data. The user data can be accessed as task->user_data
• The img_decoder is improved to cache the opened images. So e.g. PNG images are not decoded all the time but left open. Learn more in the Image documentation

Credits#

• @embeddedt for tirelessly supporting users and professional help in the development
• @amirgon for creating and maintaining the MicroPython binding
• @puzrin for creating the new font converter

And for every contributor who helped with pull requests, ideas and suggestions, testing and feedback —

Thank you very much for all of you!