I have waited for a while before I decided to write this review of the Gutenberg, I wanted to test editor that has come far in the development process. Now, with version 1.1 Beta, I think that editor is becoming ready for testing.
Gutenberg is a proposed replacement editor for the WordPress core. Currently, it is in development, and available for testing in form of a plugin. And, you should not use it on the live websites, it is highly unstable, far from complete, it doesn’t work with any theme as it should and it is in conflict with just about every plugin currently available for enhancing WordPress editing.
For the last several months, you have most likely stumbled on an article about the Gutenberg, and maybe even participated in fierce discussions about the goal of the Gutenberg editor, current development course and many other things, with the main focus related to backwards compatibility, data format for the content and few other things directly related to the way Gutenberg works and how it affects the complete WordPress plugins and themes ecosystem.
I can’t remember of any issue in the WordPress recent history that has polarized WordPress user base so much as the Gutenberg. And, there is a very good reason for that: Gutenberg will completely change how the WordPress work and how the future development will unfold for the years to come. Gutenberg has the power to shape the future of WordPress. And that is why, it has to be done right, or it can have severe consequences on the future of WordPress and the economy built around it.
If you want to check out some of the articles on Gutenberg and read the lengthy discussions, the best starting point is the WP Tavern website.
There are so many subjects I want to include in this article, and also, to reference some of the most important articles from other authors in an attempt to bring you all relevant information to help you start with the Gutenberg testing and maybe include you in the discussion that will lead to getting the best possible editor we can have for WordPress core.
I will start with…
Overall, I like how Gutenberg is coming along. For writing text and basic editing, it works OK. I like the idea of using blocks to format the content. It is easier to understand what is going on with the content, it makes the editing process easier when you have controls for each block that limit what the block can do.
And, I understand how this type of editor can help regular users with writing content. With more development, this type of editor can be extended to support templates for the content where you can predefine areas for users to work with and many other things. In the long run, this type of editor is a good thing to have in the core.
But, for now, there is a lot more to be done before all this is realized and before the Gutenberg is ready to be a part of the WordPress. So, we come to…
Here are some of the issues related to the current plugin version:
- There are way too many blocks. Many of them can be merged into one because they are more or less the same: paragraph, shortcode, blockquote, verse (why we need such block is also an interesting topic).
- Why do we need a block for shortcode??? And why show such a small input field to insert it? Most shortcodes can have arguments, extra content, and making all that fit in there is going to be a big problem.
- Why do we need to have each embed as a different block? Each embed block takes too much space, and the only difference is the icon displayed. The oEmbed object is smart enough to recognize which URL is embedded, there is no need to select what is embedding.
- Paragraph block doesn’t allow you to have some words in a different colour from the rest of the text in a paragraph. This is a highly illogical approach, and paragraph must allow multiple colours.
- Gutenberg doesn’t work on small screens and mobile devices. Scrollbars are everywhere, bringing block sidebar makes it impossible to preview changes. Even with desktop screens when the browser is not maximized, scrollbars are everywhere.
- And for the large screens with touch capability, the editor is again very hard to use, because some of the controls and buttons are very small, hard to click on with any precision.
- Many of the block controls are confusing and behaving erratically: changing the font size, block and text alignment confusion, media selection dialogues showing wrong messages and button labels…
- Columns block is completely useless. If you want to have such control, it has to behave as a container, and you must be able to insert all other blocks inside the column cells. As is stands right now, it is better to remove it.
- Selection of blocks is also very confusing, and in many cases, you need to click just right to select some of the blocks. It is very easy to click the button for deleting the block.
- Image block allows you o resize the image, but aligning image inside the block is not possible. The image caption is limited in editing option with only basic styling allowed.
- Keyboard navigation is very inconsistent, and sometimes it works as you expect it, sometimes, you need to use the mouse to click on the block or select something. This is very annoying when you work with long text.
- Display of blocks that are aligned left or right is awful, truly awful. It is completely unusable, especially if you try to use embed, gallery or image blocks around other similar blocks.
There is much, much more for this list, but I am not going into more details. This list is just to show you how much work remains to bring Gutenberg from beta to stable status. I expect that these things will be fixed in the future versions, making editor much smoother. But, there are several issues that will make or break this project…
- Data structure: One of the worst decisions made when Gutenberg development started is to invent a new way for storing content using very ugly and highly limited format based on using HTML comments and parsing content using regular expressions. This approach is one of the biggest issues I have with Gutenberg because regular expressions are terrible for the task they have to perform here: there is no way to have nested content (there goes the flexible columns block where you can add any content inside the cell) and parse it reliably, it is prone to errors and with addition of new types of blocks, it will definitely break at one point. If this remains the way Gutenberg works, it will be the biggest spent opportunity to store content in a hierarchical structure that will be easier to expand, will allow nesting and be easier to render. JSON comes to mind and there are many existing implementations available to reuse without reinventing the wheel.
- Meta boxes: are gone as it stands right now. Gutenberg has a narrow sidebar for adding something resembling meta boxes we all are used to and love. Imagine 3 or 4 more boxes there, it would be completely unusable. And even worse, most metaboxes I have seen are huge, they will never fit into narrow sidebar area. Imagine monster of a metabox in WooCommerce, that can be expanded by more plugins to even bigger element, and it might take few pages when you take into account tabs it has. No matter how much time WooCommerce developers spend, that will NEVER fit into Gutenberg (as it is right now) and have the same functionality. Meta boxes must be supported as they are right now, or Gutenberg will be only a failed experiment.
- Blog orientated: right now, Gutenberg is viable for blog use only. Lack of support for meta boxes renders it completely useless for eCommerce websites, communities and pretty much everything else (except simple blogs).
- Expensive transition: from current code to Gutenberg. If all the developers have to rewrite all the plugins and themes relying on the current editor to use Gutenberg, while at the same time remain compatible with older versions of WordPress, it will crush the WordPress ecosystem. I can’t even imagine how much work would be needed to do that. I have several plugins that integrate meta boxes, that depend on the saving process, and realistically, I would need 3 to 6 months to do all that work, while putting everything else on hold. Imagine now how many plugins rely on meta boxes, on a very complex meta boxes. I don’t see a realistic way to do all that work in time for when the Gutenberg is ready for the core.
What I propose
Well, I have written enough criticizing Gutenberg as it stands right now. But, what can be done to make Gutenberg good or even great? I have few suggestions.
React has to go. Facebook and their anti open source license (masquerading under BSD license) are not in the spirit of WordPress. There are many other alternatives to choose from, and some of them are even compatible with React code, but they are under real open source license.
The decision is made, and React will be removed from the Gutenberg. This is a great news, and you can read more about it here. Very soon, we will know what will replace React.
- Gutenberg has to be backwards compatible. All the filters and actions current editor has, must be working with Gutenberg editor too. And, the same goes for meta boxes. There has to be a way for all the current code adding meta boxes to work without changing a single line of code. If that is not done, Gutenberg will crash and burn.
- Rethink and remake the data format used to store content. The current HTML comments based system is bad, severely limiting and if not changed right now, a few years from now, it will have to be revisited to allow for more flexibility. Why wait, do it right now!
- On the day it is decided that Gutenberg should be merged into WordPress, we need to have full documentation ready, examples for building new blocks, rendering control and anything else developers might need to extend it and integrate with it.
- Make Gutenberg optional for one or two major versions. Include options in the WordPress settings to choose between old TinyMCE based editor and new Gutenberg editor. Proposed plugin to disable Gutenberg is a stupid idea, just include the checkbox in main WordPress settings, and by default leave Gutenberg disabled.
- Include new options in the post types registration to choose which editor interface to use.
Here is the list of some great articles about Gutenberg you should read:
- You called it Gutenberg for a Reason. That Doesn’t Make it Revolutionary.
- The Gutenberg that could have been
- We Called it Gutenberg for a Reason
- Diving Into the New Gutenberg WordPress Editor (Pros and Cons)
- Stop the Presses: Gutenberg is What WordPress Needs, but It Doesn’t Go Nearly Far Enough
Final words (for now)…
Well, I am looking forward to next updates and the decision about the library that will replace React. I would like to see the Vue.js as a replacement, and a lot of developers are vocal about including Vue, but we will see what happens soon. I will revisit Gutenberg releases in the future, and keep you up to date with the progress.
14 thoughts on “Gutenberg: The Good, The Bad and The…”
Thanks for your in depth article about this. Since Gutenberg was announced I migrated my website away from WordPress and to IPS. I hate that I had to do this and maybe it was a knee-jerk reaction to the announcement. Still, I miss my WordPress and the community that supports it. I’m not against change and I fully support change that will foster improved WordPress performance overall. I had hoped that this would drive WordPress towards PHP7 with all of the performance improvements I expect. From reading your article, it seems like it is being forced as a way to end commercial plugins such as Visual Composer and even yourself. I hope that the final product will be great and will help grow development of great plugins like bbpress Toolbox, but I fear it’s going to cause so many problems when they make the final push to core.
Thanks for the comment Erich! Gutenberg has come a long way since the version I reviewed in this article, and I plan to revisit the Gutenberg soon. Many things are still a problem, but a lot of other things are much better now. I am trying each version as it is released, testing with mine plugins, and it is getting better. I will wait for 2.1 version to do a new review.
Thanx for your words on Gutenberg. With your huge back-catalogue of plugins the community really ought to listen to your warnings. I really agree on your overall call for caution, if not necessarily a 100% backward compatibility. But at least some alternative for existing solutions. And I think it would be sensible to make Gutenberg a general settings option for a good long time.
Now, we know that Matt M & co have made some really sensible strategic decisions over the years, so I’m sure this time will not be different. I think everybody knows how much is at stake here. We should focus on the irreversible changes. The ones that will be hard to work around in future updates. The very basic choices (like regex vs JSON) are really important. So a good long roll out period to pick up experiences in the real world would be wise.
I’m looking forward to your update article. Maybe not so much on the nitty-gritty bugs that are not yet fixed, but the larger strategic scope.
PS: You might want to edit your final words about React, as you have done earlier in the article. 😉
Thanks for the comment Jesper!
I am currently testing and updating my plugins that have metaboxes, and with the latest Gutenberg 2.4, mostly there are styling issues with metaboxes, and they work fine. There are breaking things in metaboxes that will affect plugins that need to work with the edit screen and editor area, and I don’t see a solution for such plugin, other than radical code changes to support Gutenberg. At least, for now, major WordPress plugins will be updated for the Gutenberg in time, but I still think that a huge number of plugins will not be, and that will cause websites to break.
I will publish follow up article later this month, including my experience with updating plugins to work with Gutenberg.
I really feel your pain, MillaN. :-/ Will the Gutenberg API be the way to integrate these things. At the risk of exposing my huge ignorance on the matter of core coding, could you perhaps elaborate a little on the hurdles of closing the plugin coding gap between Gutenberg and the classic WordPress environment? If not, I understand.
First thing is metabox styling, that has 100% different with Gutenberg, so plugins with metaboxes relying on the previous metabox styling will look bad. That is easier to fix with additional styling. But, metaboxes that rely on interaction with the editor, other elements on the edit page (other metaboxes) will be completely broken, and what is worst, in some cases unfixable, because we no longer have one editor, each block, in essence, is an editor. Plugins that add items to the TinyMCE editor toolbar will need 100% rewrite to support toolbars in each editor block (or some of them).
Thanx for enlightening me! That helps to understanding the inherent obstacles in the whole project.
And BTW, MillaN, what do you know about and how do you see the possibilities of Gutenberg in bbPress? I think you would be qualified to comment on that better than most people?
Right now, I don’t think that there will be any development for bbPress to add Gutenberg support. bbPress development is slow, and upcoming 2.6 version is few months late already, and supporting Gutenberg would require a lot of changes, that might be considered for 2.7 (or more realistically, 2.8) version. But, it is possible that someone will make a plugin for bbPress to integrate Gutenberg.
For the most part, forums don’t need complex editing, and I know that most forum users are not using even TinyMCE editor, so I don’t think that most websites will embrace Gutenberg in the forums on the front end.
Sounds right. I’m concerned about a forum that has a lot of photo sharing, and I could envision much simpler interface if we at some point allowed a matured Gutenberg with a reduced number of blocktypes. We might not have to do BBCodes then. 🙂
I think that would be possible, limit the available blocks to specific post type, or make it easy to disable blocks using filters. That would make sense with bbPress. But, Gutenberg should be able to work in the limited space, not only full screen, I would like to see how that turns out.
Yes. I have yet to test it on a mobile phone. I’ll do that. Thanx.
Well, I put in the necessary stops to prevent my websites from updating to wp gutenberg. I will be using a fork of WordPress, which is now in alpha release phase. I tested Gutenberg multiple times and each time proved to be a lesson in futility and complete frustration, not to mention wasting of my time. That’s when I decided that removing WordPress from my servers was the best option. Once the fork (ClassicPress) is ready to go live, that’s what I will be using.
I am also looking forward to test ClassicPress, but, for now WP with Classic Editor plugin is the best way to get rid of Gutenberg and still get normal WP. I don’t expect ClassicPress fork to get stable release for at least 6 more months.