Calls to Delay WordPress 5.0 Increase, Developers Cite Usability Concerns and Numerous Bugs in Gutenberg

Developers and business owners are waiting anxiously in the wings, as Gutenberg is 11 days away from its debut in WordPress 5.0. There is still a chance that the release could be delayed to the secondary date (January 22, 2019), but the decision has not yet been announced.

“I am lukewarm on the 19th, but not because of the number of open issues (which isn’t a good measure or target) — more that we’ve been a day or two behind a few times now,” 5.0 release lead Matt Mullenweg said during yesterday’s dev chat. He said that reports “from the field” continue to be good and companies that have already installed and activated the plugin haven’t reported a higher than normal support burden.

“My concern can be summed up as this,” Aaron Jorbin said. “There are approximately 400 issues that need either code or a decision to punt. Assuming five minutes per issue, that means there are about 33 hours worth of bug scrubs that need to take place between now and RC.”

“I don’t think we can make a decision on moving the date in the next 45 minutes,” Gary Pendergast said in response to concerns raised at the meeting. “I do think it’s fair to say that the Gutenberg and 5.0 leadership teams are hearing all the feedback, and are actively looking whether the timeline is still correct.”

Mullenweg said open issues are not a good measure of whether the release is on target but the numerous bugs the community is encountering has precipitated a flurry of posts advocating for the release to be delayed.

In a post titled “WordPress 5.0 needs a different timeline,” Joost de Valk, author of Yoast SEO, cites accessibility concerns and the stability of the project as reasons for a delay. de Valk identifies himself a strong supporter of Gutenberg and his team has already built compatibility and Gutenberg-first features into their plugin, which has more than 5 million active installs.

“It’s arguably one of the biggest leaps forward in WordPress’ editing experience and its developer experience in this decade,” de Valk said. “It’s also not done yet, and if we keep striving for its planned November 19th release date, we are setting ourselves up for failure.”

de Valk gave two reasons for why he believes the November 19th timeline to be untenable:

There are some severe accessibility concerns. While these aren’t new and a few people are working hard on them, I actually think we can get a better handle on fixing them if we push the release back. Right now it looks to me as though keyboard accessibility has regressed in the last few releases of Gutenberg.

The most important reason: the overall stability of the project isn’t where it needs to be yet. There are so many open issues for the 5.0 milestone that even fixing all the blockers before we’d get to Release Candidate stage next week is going to prove impossible. We have, at time of writing 212 untriaged bugs and 165 issues on the WordPress 5.0 milestone.

WordPress developer Mark Root-Wiley published a post the same day titled “WordPress 5.0 is Not Ready.” He outlined why he believes the release needs to be delayed and suggested the project pursue more auditing and quality assurance testing before shipping it out.

“WordPress 5.0 can and should be a positive change to WordPress, but if it is released in late November as planned, it won’t be,” Root-Wiley said. “There are simply too many bugs in the editor, and the experience is not polished enough. This is because the rate of development has prevented systematic quality assurance (QA) and user testing. Both types of testing are required to ensure the editor is ready and to increase the community’s confidence in the update.”

Root-Wiley describes a buggy experience when attempting to write blog posts with the new editor, which echoes many others’ recent experiences.

“I’m doing my best to give feedback, but it’s exhausting and there are so many little bugs that I struggle to isolate and replicate the one I’m reporting without running into another,” Root-Wiley said. “How is it possible for me to find so many bugs without trying from just writing 1.5 blog posts?”

Root-Wiley also suggested removing what he deemed to be unnecessary features in order to streamline the editing experience and focus on the fundamentals. These features include the tables block, paragraph background colors, spotlight and fullscreen mode, dropcaps, verse block, among others.

“The pace of development has been blistering,” Root-Wiley said. “That speed has been great for developing a lot of features and iterating on those features quickly, but it hasn’t allowed for sufficient testing. What’s needed now is more time for people to find and report bugs with the editor features in their proposed final state.”

Gutenberg criticism is often characterized as coming from people who are resistant to change, but these strong messages about delaying the release come from developers who believe the new editor is the future and have heavily invested in contributing to its success.

Both de Valk and Root-Wiley’s posts seem to have resonated with many who have had similar experiences with the editor. Other core developers and committers have also publicly lent their voices to the call to delay the release.

Opinions on Gutenberg’s readiness vary wildly depending on the person’s perspective and involvement in the project. Those who are working on it full-time have not publicly offered opinions indicating that it might not be ready for the November 19 timeline.

“The 5.0 milestone is in a very manageable place, but if the volume becomes more worrying in the next couple days or it becomes clear milestones won’t be made, we’ll revise as needed,” Gutenberg technical lead Matias Ventura Ventura said during yesterday’s dev chat. He confirmed that the fast pace of development will continue.

Regardless of when 5.0 is released, users can count on getting minor releases every two weeks to address bugs and issues that pop up after Gutenberg is in the hands of millions more users.

“Hopefully as people get used to the more regular cadence they can plan around it, much like they used to complain a ton about, but then got used to, 3 major releases a year,” Mullenweg said during the dev chat.

In 2016, Mullenweg began describing how WordPress could become “the operating system of the web,” with open APIs that others can build on. While that idea encompasses a lot more than just release schedules, WordPress seems to be moving in the direction of shipping updates that come more frequently and eventually more invisibly in the background, similar to how users update their browsers. Releasing Gutenberg in its current state, with frequent updates following, could prove to be a major testing ground to see if greater world of WordPress users are ready to embrace this new era of rapid iteration.

21 Comments


  1. I read both blog posts advocating for a delay, but I don’t think more time will fix the underlying problems with Gutenberg.

    Javascript-heavy writing interfaces tend to have a certain level of jank and wobble that can’t fully be fixed.

    Look at the help forum for Squarespace. Editor issues abound, and their project is several years old.

    With that said, if Gutenberg won’t break sites, then it’s ship-worthy.

    Report

    Reply

  2. I’m still at a loss as to understand why the November target date is being pushed so hard by Matt M. And admitting you’re releasing a major update to WordPress with a large number of bugs, many not even triaged, but suggesting they’ll be gotten to in minor updates later makes no sense to me. Who stands to benefit from this November deadline, knowing what we all know? Because it sure as hell isn’t the WordPress community, the visually and other impaired users, or Managed WordPress business owners and the like.

    Report

    Reply

    1. WordCamp US is December 7th. I suspect November 19 was picked to give 5.0 a couple of weeks in the wild and then call it “Mission Accomplished.”

      Its the only reason I can come up with.

      Report

      Reply

  3. The key quote from Mark’s post, was “This is a feature people won’t miss if it’s not there, but will complain about if it only half-meets their needs.”

    Having been involved with this on projects as small as a perl script to as large as Windows95 this is a big human-factors problem. Coming from a user-education and instructional design background this is the area I’m best able to address coherently. (Others are way better able to discuss the substantial coding issues, program management practices, and release-management concerns.)

    A truism in documentation is that users need a positive first five minute experience with a new software feature, and that after six weeks almost no one will start using a feature they overlooked or dismissed earlier. If something like moving blocks or adding columns doesn’t “work” for someone right away they’re likely to dismiss it — even if they don’t reject it out of hand the “later” in “I’ll try that again later” rarely materializes.

    I agree with Mark that an initial release with the four or five major blocks for writing with the writing flow clearly worked out would have been the way to go. Then roll out new blocks (with a little “try it” fanfare) when they’re ready. Even assuming the workflow was smooth (and the WPTavern folks, and Joost, and Mark, and I, and too many others agree it’s not) coughing out a bunch of half-baked blocks including “Hello Dolly” fluff like the Drop Caps block, is going to leave people going either
    – Wait, how’s that supposed to work? or
    – Wait, it’s broken

    You don’t want that in any new-product rollout. Just because something like Gutenberg is way overdue doesn’t mean it needs to be rushed. The team’s been making very good progress — there have been huge improvements in the last six weeks! If they could maintain the pace I’d expect things will be looking very good in January.

    As someone with a lot of experience in training and support as well as sales I’d much rather see it released with a splash all at once instead of dribbled out in two-week “ok, try it now” increments.

    Report

    Reply

  4. I don’t know where to answer this question, but wouldn’t having a full WordPress API that let anyone create their own headless WordPress that is feature complete and has its own administration panel in any language be ideal? What happened with the WP API? Gutenberg will eventually be part of many parts of WordPress and I sometimes wonder if we aren’t feeding things like ClassicPress by not allowing for a disconnection between WordPress as a platform and its management.

    Report

    Reply

  5. First, let me say that I really do like the idea of Gutenberg and think it has tons of potential. At the same time, I also believe in making sure that things are completely finished and tested before they see a full release.

    You just wonder if the break-neck pace of development is causing really simple usability issues from being fixed.

    From my outsider’s perspective, I don’t quite understand the hurry to put this out in such short order.

    I know the team is working hard and doing what they can. But they should be given more time to iron out the issues with accessibility, etc. and make sure the release is successful.

    Report

    Reply

  6. The accessibility concerns are well covered but the usability of this thing is still very odd. The interface doesn’t give the key block actions the prominence required.
    I’ve been testing Gutenberg since the start of its development and in my most recent tests I confess I have no real idea how to delete a block from a page layout! I guess the little bin was removed recently.

    As a relatively competent user I know I can click a block and then in its text-formatting bar I can click the 3 mysterious vertical dots and at the bottom of that resultant contextual menu select the “remove block” option, but that seems too buried to be the real primary method. Also the keyboard shortcut suggested is “shift alt Z” which is surely not anyones idea of a delete-block shortcut. I’d have guessed at “shift-del”

    Also, another example of the strange interface:

    The most important button to introduc to users is probably the new “add a block” button, but it’s a pale grey 15px mysterious + icon tucked up away in the top left. Completely anonymous. If the Add Block menu is found … this should now be the showcase of the new editors abilities, the grand presentation of the glorious new features. Look at the sidebars of Divi or Elementor. They lay out their wares. Meanwhile the Gutenberg block display … It’s about 300px high! A cramped, nervous, mysterious, awkward little thing. It’s nearly imposible to investigate potential through this tiny little scrollign window. Why is this not a glorious display with a height:90vh? Why is it so small? For what reason?

    That’s just two observations of about a hundred. I know I may as well throw these down a hole for all the good they’ll do – but as end users start to discover this interface they will be very underwhelmed.

    I agree with everything David Innes said earlier, in the first five minutes most users will not be excited and seduced by this UI introduction to the block metaphor.

    Report

    Reply

    1. Agree, After using any of the page builders you realise how weak Gutenberg is.

      Another thing to remember is that, while it can take a while to discover every nook and cranny of those page builders, they are used by people who have elected to use them and spend time mastering them.

      If you are going to subject people to a block builder system you would want to make sure that it is much much easier to use than any page builder already on the market.

      Report

      Reply

  7. Agree with the opinions. One thing I notice from WordPress.org is the Classic Editor plugin is now featured in the Plugins > Add New screen, which is a big plus since it lets users install it faster.

    I’d love to see Gutenberg is not active by default until the concerns are resolved. A warning message in the admin area or an option to enable/disable it will be a good option.

    Report

    Reply

  8. I can see only two groups who are absolutely desperate to get Gutenberg released as soon as they can:

    1. Matt and the people he pays to praise GB
    2. WordPress’ various competitors

    The second group can see where WordPress is heading with this GB thing and can’t wait for it to be unleashed on the world, with the end result that they’ll be getting more customers.

    The sad thing about GB is not the fact that it will be forced upon millions of people who don’t want it, it’s the terrible project management (no roadmap), non-existent transparency (a11y people being ignored) and Matts unwillingness to just be honest and say it like it is: “I’m doing this for the money, not the users”.

    GB reminds me of my time in the enterprise world, when small, human companies would end up growing into soulless, lying behemoths of bureaucracy. And if you didn’t drink the cool-aid then the HR-people would “let you go”. This makes me sad.

    Report

    Reply

  9. There’s a quite a number of WP theme / plugin vendors (again paid not free) and site builders that put off even getting started with Gutenberg until the 1st WP 5.x beta. That was a mistake IMO. Everyone gets busy but that’s no reason to flat out ignore and put off the inevitable. It’s now been many years since the phrase “Learn JavaScript Deeply” was uttered and nearly two years since Gutenberg development began.

    I can imagine that many ( but certainly not all ) need more time because they’ve already failed to keep up with the pace of their industry. Despite the valid number of issues still in Gutenberg, it seems to me that the v5 release likely could have been that much better off if they’d submitted their issues earlier. Even a year more may not make an appreciable difference as it would just allow them to procrastinate longer probably.

    With the classic editor plugin available and potentially even being bundled with upgrades, there should be no issues on existing sites for those that aren’t ready yet. The WP Community has seemingly held their breath waiting for Gutenberg and WP to drop and now it’s time to exhale.

    Report

    Reply

    1. .. that put off even getting started with Gutenberg until the 1st WP 5.x beta. That was a mistake IMO.

      It simply was and is not possible to seriously develop for Gutenberg. Way too many bugs, heavily changing UX over last year, many very breaking changes, deprecated features every few weeks, new APIs added besides “feature freeze” e.g. “Annotation API” today, and whatnot.

      All you can do today is testing a few small blocks, which is no serious development. Everything else is waste of time and sunk costs.

      It is not possible to offer “Gutenberg ready” themes and plugins, as long as Gutenberg is not ready.

      Report

      Reply

      1. I don’t expect “Gutenberg Ready” and I’d be somewhat surprised if anyone really does at this point. Many have managed admirably despite the bugs. UX changes. etc. “Gutenberg Tested” now that is another matter compared to the “we’ve not even begun / started review how to integrate with Gutenberg yet” I’ve been seeing routinely.

        Report


      2. Matt B: I don’t expect “Gutenberg Ready” and I’d be somewhat surprised if anyone really does at this point.

        Well then, surprise! https://www.google.com/search?q=“Gutenberg-Ready”

        Report


    2. Tbh, I think the delay is only natural.

      Last year this time WP 5.0 would come in April this year, that was moved to June, then it became August.

      If you run a plugin/theme house, would you waste your time with preparing for something that isn’t even sure itself when it comes?

      Of course not!

      Report

      Reply

    3. I wouldn’t doubt that most theme and plugin makers waited until the 5.0 beta to start working on Gutenberg compatibility (which in my opinion is reasonable). We didn’t wait, so we’ve had a significant head-start, but I’m still not sure we’d be able to make things what they ought to be by the 19th. Fortunately I don’t see the 19th happening as there are only five or six work days left and there’s no RC yet.

      Report

      Reply

    4. Ah hang on there for a minute!

      Yes I suppose that is one way of looking at it.

      Another way of looking at this would be to reflect on how long Apple allowed users to transition from OS 9 to OS X. About 5 years till OS X completely matched the feature set of the older OS and added extra functionality. You could dual boot between both operating systems and could even run OS 9 in the emulated true blue environment from within OS X.

      It’s not beyond the bounds of possibility that with the old editor UI still extant in WP 5 (just overridden by Gutenberg with REACT, requiring an extra plugin to disable) that you could have had a switch in writing settings to control the one liner add_filter( ‘use_block_editor_for_post_type’, ‘__return_false’, 100 ); for turning Gutenberg on and off.

      I think that Matt and the lads would have had better PR success if they had been more mindful, been a bit more considerate.

      As for the third party developers sitting on the fence. When you are observing something that from a distance that looks bonkers (a bit like Brexit), that’s where you remain until you see where the pennies drop.

      Report

      Reply

    5. A year ago I started developing a plugin to get a grip on Gutenberg. Every week or two there were new versions released, with breaking changes, even though the devs assured everyone that the editor’s API was stabilized. Frankly, I got tired of keeping up with the editor’s progress and I stopped developing my plugin. I’ll resume work on it when WP5.0 is released, thanks.

      Report

      Reply

  10. It would be better to release a stable, accessible product in January than to release a buggy, unaccessible one in November for the sake of market share.

    (Also, Twenty Nineteen will probably be finished by January, so 5.0 can come packaged with the new WordPress theme. Bonus!)

    Report

    Reply

    1. @Calum – In a ideal world, I would agree with that sentiment. :-)

      Report

      Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.