Open pnoeric opened 6 years ago
The major problem here is the same as outlined in https://github.com/justintadlock/members/issues/3#issuecomment-148420716
Essentially, shortcodes can be both inline and block-level elements. My guess is that most people use the shortcodes as blocks rather than inline. And, that was the original intention. However, there is some backward-compatibility concerns there. I'm not sure what problems it might cause to just start making the output always block-level.
Ideas:
1) One thing that I can definitely do is provide filter hooks for all of the shortcode output. That would allow you to add filters to do custom output.
2) With Gutenberg planned for integration into WordPress 5.0, Members should provide a "block" using the Gutenberg system for showing/hiding content. This would kind of give a bit of a reset on the problem with the shortcode by using a totally new system.
Sure, I get it. Obviously filter hooks would do the job fine (please pass both the content and the parameters entered with the shortcode, thanks!). And/or it could be a checkbox in the settings: "Wrap shortcodes in <div>
?" (or <span>
).
I would bet you could assume that a user advanced enough to turn that on understands the inline vs block issues and can tweak their CSS as they wish. The goal is just to general useful markup; inline vs. block is strictly styling, right?
Or hooks would work. :-)
Hi Justin, thank you for a terrific plugin! We use it here to put admin-only notes (for various levels of members) at the top of hundreds of knowledgebase posts. (Picture basic info/warning/similar CSS Bootstrap-esque alert boxes.)
These notes need to be styled. Rather than force the non-technical post authors to add CSS code in addition to the member_access shortcode, I would prefer to have the output wrapped in a div with classes that match the roles. Then I can just set up the CSS to style them nicely and they can add notes without worrying about HTML.
Justin and community, thoughts?
(Spoiler alert: I've already done a quick version of this in my fork of the code and am happy to send it now as a pull request, but am creating this issue first for discussion, per the contributing guidelines.)