WordPress 4.9 Beta 4 Removes ‘Try Gutenberg’ Call to Action

WordPress beta releases typically don’t generate controversy but in WordPress 4.9 Beta 3, a call to action was added to the dashboard that encouraged users to install and activate Gutenberg.

Try Gutenberg Call to Action
Try Gutenberg Call to Action

Members of the WordPress community raised concerns that clients may install Gutenberg and shared ways to hide the prompt from users. The negative reaction inspired some developers to create plugins that hide the prompt.

One of the primary concerns is that Gutenberg is in a high state of flux and encouraging users to create content inside of it on live sites may cause compatibility issues or adversely affect saved content in the future.

“Any change to the integrity of published content and its formatting that results from changes during continued development and evolution would be unacceptable from the point that we encourage users this directly to install it on live sites,” Nick Halsey said.

“On the other hand, this could require core to take on significant technical debt to maintain compatibility for earlier iterations of the editor as a plugin.

“There should be a make/core post addressing this issue, at a minimum, along with a compatibility plan for the next stage of development as a plugin. Before core encourages millions of sites to use the plugin and rely on it functioning a certain way.”

Other members of the community advocated for the call to action saying it would lead to a larger test sample.

The call to action was removed after the core team discussed it with Matt Mullenweg, “I like the idea of the Gutenberg promo, but want things to be a bit further along before we really open the doors to try to get as many users as possible,” Mullenweg said. “If we can flag off or remove the promo, we can always bring it back in 4.9.1 or another time when things are more ready.”

‘Try Gutenberg’ Dashboard Prompt Will Set A New Precedent

There have been many WordPress features in core that started off as plugins first, MP6 being one of the most memorable. However, to the best of my knowledge, there has never been a dashboard prompt encouraging users to install and activate a beta plugin on a live site.

Although the call to action is focused on raising awareness of Gutenberg, John James Jacoby suggests that the meta block be rewritten so that it can be recycled for other features or plugins to use in the future.

“My concern is that the current approach is not scalable to future feature developments beyond Gutenberg,” Jacoby said. “For example, when a new codenamed feature comes along for WordPress 5.2, cloning this same approach does not seem ideal.”

He suggests that the dashboard widget become a standard part of the dashboard. “This way, we can hype the new hotness on an as-needed basis, and plugins that want to hide it forever can reliably do so one time in a plugin,” Jacoby said.

When Is the Right Time to Hype Gutenberg to the Masses?

Gutenberg is actively installed on more than 3K sites with nearly half of installations running version 1.4. This is a far cry from the 100K active installs Mullenweg would like to see before merging it into core.

I don’t think advertising Gutenberg in the dashboard to millions of users as the new editing experience should be considered until a merge proposal has been published on the Make Core WordPress site. By this time, many of its quirks and how it handles meta data, meta blocks, and preventing data loss will likely be solved.

I am one of the people who raised their eyebrows at the idea of advertising Gutenberg at its current stage of development to the masses. My primary concern is that it’s not ready yet. At the same time, I wonder when or if there is a right or responsible time to advertise installing beta software onto a live site. What do you think?

20 Comments


  1. If installing Gutenberg results in a full-screen takeover of existing posts (as is the case with Gutenberg 1.5), then meta box support should be a requirement. My experience with installing 1.5 resulted in a blank editor that had previously been fully populated by ACF meta boxes. That cannot be the first impression for the general public.

    Report


    1. hmmm, I didn’t think of that angle but yeah, how important do you think the first impression is towards Gutenberg’s success or willingness to adopt it?

      Report


      1. First impression is critical, especially if Gutenberg is being encouraged in a production environment. Those users aren’t interested in testing experiments that break or otherwise inhibit the workflow they are used to. We have already seen negative reactions from developers who understand what a beta test involves; the public will be even less forgiving.

        Report


    2. An initial pass at implementing meta boxes was included in Gutenberg 1.5. A Firefox-specific bug preventing meta boxes from displaying correctly was later resolved by the 1.5.1 patch release. In my own test environment, ACF meta boxes are shown correctly in Gutenberg 1.5.1.

      Report


      1. ACF meta boxes with a High position (appearing after the title) do not display in Gutenberg 1.5.1.

        In addition, just because some of the meta boxes technically function in the Extended Settings drawer doesn’t mean that implementation is anywhere near good enough to ship in production. The first impression is still an empty editor where a full UI is expected.

        Report


  2. I think a better measure of Gutenberg’s progress won’t be the number of sites it’s installed on, but the number of posts made through it (even though that’s a bit harder to track). We’ll still target getting it on as many sites as possible, but the publishing activity is a better North Star metric.

    Report


    1. I think a better measure of Gutenberg’s progress won’t be the number of sites it’s installed on, but the number of posts made through it

      Changing our minds again aren’t we?

      Report


      1. Yes! I try to approach even my strongest held beliefs with a very open mind. I’m interested in the best results for WordPress and the broader web, not being “right” by sticking to an idea or belief just because I previously believed it.

        If we continue to learn, I’m certain there are other things we previously thought were true that we find a better answer to.

        Strong opinions, weakly held. :)

        Report


      2. I’m certain “try” would have resulted in a lot of “fail” for many of the sites my agency runs, because meta boxes and other custom field customisations are critical. Clients would have been drawn to the new shiny and been quite upset when it didn’t plan out as expected.

        So I’m very grateful that Matt and the core team have listened to feedback. I was a Gutenberg sceptic but the Ship of Thesus post I read recently convinced me that this it is the way forward – it just needs to be carefully rolled out once it’s really solid. One of the main reasons our clients love using WordPress is the ease of adding and editing content. Anything that changes that process needs kid gloves.

        Report


  3. Aren’t people discouraged from using beta releases of Core on production sites already? Adding a beta plugin to a beta Core doesn’t seem like that big a stretch, with appropriate additional cautions.

    Report


    1. I was thinking the same thing. It’s beta software not meant for live sites.

      Report


    2. I think you’re misunderstanding what was being proposed. The notice to encourage Gutenberg installation was added to WordPress 4.9 Beta 3 with the intention of it appearing in the official release of 4.9. That is the decision that was rolled back given it’s not ready for that kind of exposure.

      Report


      1. Ah, that does make a bit of a difference! That was not clear in the story, thanks for clarifying.

        I still think that it’s a good idea for beta testers to test Gutenberg now (as I’m sure you’d agree), but certainly don’t ask ordinary users to do that without several “don’t do this unless you know what you’re doing” messages.

        Report


      2. That said, I think Mike’s original understanding would be a good approach: only show the promo on sites that are running beta versions of the software. Although I suspect the number of WP core beta testers who don’t know about Gutenberg is low, people who are already testing beta software can and should be encouraged to test this beta software as well. The expectations of breakage and non-production use are similar for both.

        Report


  4. I wonder when or if there is a right or responsible time to advertise installing beta software onto a live site. What do you think?

    There isn’t. Naive clients should never be encouraged to try beta software just because and that goes 10x for something that takes over and drastically alters the core reason for WordPress’ existence, creating and managing content.

    This is actually the hardest part of this and why I think Gutenberg should live as a plugin for longer – it changes something so central to site operation that it’s unlike almost any other feature rollout. Training will be needed, possibly theme updates, etc. At the same time, we absolutely should want Gutenberg used in a wide variety of installs.

    That tension (“test widely vs make sure you aren’t breaking existing content process”) will likely slow any wide rollout. I’d never deploy this on a live site for a client until it was feature complete, stable, worked with things like ACF that I have widely deployed and I could spend some time with the client to walk them through the different workflow.

    Report


  5. It is a really ambitious challenge to improve on the content editing workflow of WP, however i just wonder how this whole Gutenberg story evolves. In my opinion it will be an epic fail. There are at least a dozen of other ways to improve on WP that are actually since long asked for…but no, no way will they get implemented. And here comes Gutenberg. Nobody ever asked for that again, has been hyped for months now and there are 3k sites using it…out of millions of WP sites.
    Maybe the idea of releasing it in 5.0 should be reconsidered as well. At least a usage statistic like that should really warn those who are responsible making the decision about merging it into core.

    Report


    1. You can’t use the 3k sites as a reliable number for Gutenberg. As beta software which is specified to not be used on live sites the number will be low.

      Gutenberg isn’t ready for core before there isn’t a good meta solution in place and the system for creating boxes hasn’t been ironed out and stabilized, but for now what they’re showing is great stuff.

      Report

Comments are closed.