Extend Widgets UX and Capabilities

[davidakennedy]

After some great discussion in #37974 and at the recommendation of @folletto, I wanted to create a ticket to help foster the discussion and improvement of widgets in WordPress. As @folletto said:

Widgets are powerful. Widgets are also massively under-utilized in WordPress. We should totally work to make widgets easier to use, interact (drag'n'drop?).

Here are some critical questions to help guide this:

• How can widgets become more extendable within WordPress?
• How can widgets become easier to use for users?
• How can widgets be easier to work with for themers?

Let's figure out what the next version of Widgets looks like.

[joyously]

I think this plugin functionality should be put into core:
​https://wordpress.org/plugins/amr-shortcode-any-widget/

[westonruter]

@davidakennedy I agree 100%. Widgets need a lot of love in core, but they are very powerful and can be made so much more so if we modernize them and make them usable in more places than “sidebars”.

I think there should be a convergence between widgets and shortcodes in WordPress, with a goal of implementing content blocks. I wrote up a lot about this on the Shortcake project: ​https://github.com/wp-shortcake/shortcake/issues/585

Some related trac tickets:

• #33507: Allow widget controls to be JS-driven
• #35574: Add REST API JSON schema information to WP_Widget
• #35669: Store widgets in a custom post type instead of options
• #28580: Speed up customizer by lazy-loading controls and settings as needed
• #33473: Shortcodes + Widgets. Unified "component" API (aka Content Blocks)

[folletto]

WIDGETS IN EDITOR

amr shortcode any widget

I think there should be a convergence between widgets and shortcodes in WordPress, with a goal of implementing content blocks.

Agreed. That plugin is a great starting point! And yes to the shortcake bit too. I agree with you Weston that there's a definite convergence that has to happen.

The first place I can see where widgets could see use, and make a huge impact, is exactly in the Editor. This would mean:

• Shortcode to load any widget.
• A UI to insert the widgets that is both lean, incospicuous if one doesn't want to use it, and highly extensible.
• A UI to manage the shortcode in the editor. To be a proper first-class citizen it should be treated "as an image", not as a shortcode (moving, controls, editing, etc).

With these, I think we'll already be able to do a LOT more, and we'll finally have a usable form of "content block" (I hope this is the last time I mention it lol). I feel this has the potential to generate a new explosion in creativity and components in the future.

WIDGETS IN CUSTOMIZER

Once the editor part is "there", we can start elevating widgets to a higher "direct manipulation" experience straight in customizer. Adding them to the page (either in sidebars or content body, since it's then supported) with drag'n'drop and using the siderbar as an insertion tool, not as a navigate-and-set tool (think less of current customizer and more of Keynote's sidebar and top bar).

This is a little more blurry in my mind, and would probably require more exploration, both code and design wise. However I'd again split it in two parts:

1. A way to insert widgets that is more suited for direct manipulation's enviroments
2. A way to manage them directly on the page, and with the settings opening properly in the sidebar

This last bit is enabled by the other work currently happening in #27403, probably.

Also...

I think we partially know "why" we want to do this: widgets are a technical piece that can be elevate long-standing problems (different content types in pages, direct interactions, high flexibility) however there's a fair bit of speculation too, so it's probably something that has to be paired with strong research, including the two visions I tried to summarize above.

[desrosj]

I that where this is being explored as a focus during Phases 2 & 3 of the Gutenberg project, we could close this out.

If anyone disagrees, please feel free to reopen!