Jonathan Desrosiers 10:03 pm on February 28, 2017
Tags: browser support ( 4 ), editor ( 183 )

The New Editor and Browser Support

If you have not already, I encourage you to read through the Editor Technical Overview and check out the great work so far in the initial Gutenberg prototype. Progress on the new editor can also be followed on the GitHub project.

While the technical overview focuses on what obstacles are faced before forming opinions on the correct way to solve them, it does not focus on the actual technology requirements of the deliverables. As contributors start to create a new and improved editor experience, these technical requirements should be clearly defined and agreed upon so that decisions and progress can be made.

AccessibilityAccessibility Accessibility (commonly shortened to a11y) refers to the design of products, devices, services, or environments for people with disabilities. The concept of accessible design ensures both “direct access” (i.e. unassisted) and “indirect access” meaning compatibility with a person’s assistive technology (for example, computer screen readers). (https://en.wikipedia.org/wiki/Accessibility), for example, is one technical requirement. All new features must pass WCAG 2.0 AA.

Browser support is another technical requirement. A few lengthy discussions have occurred within #core on Slack and within an issue on Gutenberg’s GitHub. The discussions weigh the pros and cons of dropping support for older browsers in order to use features only supported in newer browsers to improve the editor. It is a tough decision to make because the editing experience will become less useful for a small percentage of users.

This post will attempt to summarize the discussions that have taken place so far and provide some data in an effort to elicit more discussion so a stance on this requirement can be determined.

Who Uses Older Browsers and Why?

Many users are still unable to choose which browser they use. Many companies and institutions have specific browser and version restrictions based on their own unique software, technical, and security requirements.

Here are some factors that contribute to older browser versions still existing:

IE8 is the highest version a Windows XP user can install.
IE9 is the highest version a Windows Vista user can install.
Windows 7 users are forced to update to IE11.

To determine how many people are still using these browsers, 3 data sources were considered. The following numbers are from StatCounter’s GlobalStats:

IE8: 0.41%
IE9: 0.26%
IE10: 0.26%
IE11: 3.79%

WordPress.comWordPress.com An online implementation of WordPress code that lets you immediately access a new WordPress environment to publish your content. WordPress.com is a private company owned by Automattic that hosts the largest multisite in the world. This is arguably the best place to start blogging if you have never touched WordPress before. https://wordpress.com/ numbers are nearly identical to these. W3Counter also shows similar numbers with IE8, IE9, & IE10 slightly higher (but with less precision as no numbers are available).

While these percentages are negligible in some contexts, it is important to remember how many people use WordPress. When these statistics are considered in the context of WordPress, they represent tens (if not hundreds) of thousands of people that could potentially be left behind if support for any of these browser versions is dropped.

Current Established CoreCore Core is the set of software required to run WordPress. The Core Development Team builds WordPress. Browser Support Policies

According to the design handbook, the current stance in Core for browsers is as follows:

Full support back to IE11, partial back to IE8.
Firefox, Chrome, Safari: Current – 1.
iOSiOS The operating system used on iPhones and iPads. Safari: Current – 1.
Android browser: Current major.
Chrome/Android: Current.
Everything else: Bugs fixed as reported.

Older browsers must continue to be functional, even if the experience does not match that of newer browsers.

Calypso only supports IE11 & Edge, dropping support for older versions.

Microsoft has even dropped support for IE8, 9, and 10 very early in 2016. This means that no bugbug A bug is an error or unexpected result. Performance improvements, code optimization, and are considered enhancements, not defects. After feature freeze, only bugs are dealt with, with regressions (adverse changes from the previous version) being the highest priority. patches or security updates will be released for any of those versions. Because of this, users should be encouraged to upgrade their browsers whenever possible.

Dropping support

In order to drop support for a browser, there needs to be a specific substantial benefit. Not dropping support for anyone, if possible, is desired.

The argument has been made that since this is an overhaul of a core feature essential to WordPress, it could be the perfect time to raise the minimum browser requirements.

Proposed Fallbacks

If a decision is reached to remove support for any of the browser versions mentioned, a proper fallback must be provided so users being forced to use an outdated browser can continue to enjoy using WordPress.

All potential options will be evaluated with the WordPress project’s core philosophies in mind. Please review them and consider them while brainstorming and discussing.

The following are some potential fallbacks that have been suggested in discussions.

Include two versions of TinyMCE in Core

This approach would be seamless for the user. It would, however, magnify the burden on Core’s editor component by requiring two versions of the library to be maintained. It would also ship code that a large percentage of the user base may not use. Every extra file shipped in core also has potential to reduce auto update reliability.

Move the current implementation of TinyMCE into a pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party.

This approach would allow site administrators to enable the current editor experience if their user base contains a large number of people using older browser versions. From a user perspective, this would be a very poor and frustrating experience, especially if the user is not allowed to install plugins, or does not have enough technical knowledge.

Use a simple textarea.

This is the current approach in the editor for users with JavaScriptJavaScript JavaScript or JS is an object-oriented computer programming language commonly used to create interactive effects within web browsers. WordPress makes extensive use of JS for a better user experience. While PHP is executed on the server, JS executes within a user’s browser. https://www.javascript.com/. disabled. However, this may alienate some non-technical users who, for example, may not know that <em> is used for italics. A basic HTMLHTML HyperText Markup Language. The semantic scripting language primarily used for outputting content in web browsers. knowledge would be required to effectively use this fallback.

Further Discussion

Do you have additional suggestions for an approach or fallback? Feelings on where the line should be drawn in the sand? Thoughts on why we should continue to maintain support for older browsers? Please mention them below.

#browser-support, #editor

Andrea Fercia 10:47 pm on February 28, 2017

I’m not sure I fully understand the reference to Calypso and how much value it adds to the discussion. This is WordPress.orgWordPress.org The community site where WordPress code is created and shared by the users. This is where you can download the source code for WordPress core, plugins and themes as well as the central location for community conversations and organization. https://wordpress.org/, I’d kindly suggest to remove references to products that are not part of the open-source project we love and contribute to.
- modemlooper 2:10 am on March 1, 2017
  
  But… the adminadmin (and super admin) is getting replaced with something like calypso so it makes sense to acknowledge it.
- Aaron Jorbin 4:28 am on March 1, 2017
  
  Calypso and what has done gives us insight into a sub-section of the broader WordPress community. Much like the decisions that plugins and themes make, we should be looking at the decisions that apps make and allow them to inform us in our decision making. Expanding our influences rather than solely inward looking will help make the best decision possible.
dmccan 11:39 pm on February 28, 2017

In addition to corporate restrictions, some people may not have access to newer computers. Mitigating those numbers is the possibility that some of those older IE users may have access to Firefox or another browser that they could use instead.

Is this about dropping support for IE 8-10? If so, then I think that is reasonable. There is likely no perfect solution when sunsetting a component. Having the current editor available as a pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party seems like an OK option, perhaps with a notice and link should a non-supported browser version be detected. Oh, I suppose you could offer to try to automatically download and activate it for them, to help the less technical users. The number of people who can only use IE 8-10 and who cannot install a plugin would probably be very small. How many of those people are running the latest WordPress? If you can update WordPress I imagine you can install a plugin. Also, a plugin provides an option for people who might not want to make the switch but who do want to be running the latest WordPress.
Andrew Roberts 1:48 am on March 1, 2017

I am not sure this fully summarizes the “pros” of dropping support. Most of the benefits revolve around reducing the cost of testing and implementation in newer versions of TinyMCE. We have basically made the decision to drop IE8-10 in future versions. It introduces a lot of extra testing and maintenance work and makes it more costly and time-consuming to implement new features. Polyfills are unlikely to help much as fundamental selection model and contenteditable APIs do not work as expected on these older browsers. We summarized just a few of the issues here:
https://www.ephox.com/blog/a-quick-guide-to-browser-selection-models/

We understand that this would impact a lot of people even with very very small percentages. Our CDN gets up to 1 million hits an hour :O

We (the TinyMCE team) are not trying to be antagonistic here and wouldn’t jeopardize our position in WordPress to make a stand but we are very very strongly in favor of deprecating IE8-10 and going with one the above solutions as a fallback (2 TinyMCEs, PluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party or Textarea).

As to which fallback solution? I think all three could work and would be interested what other people think. We will also have to have a fallback solution in TinyMCE for users outside of WordPress. I am guessing we will fallback to a textarea. We are keen to hear what others think!

-Andrew from TinyMCE
- Jonathan Desrosiers 2:31 am on March 1, 2017
  
  This is by no means an all-inclusive list of pros, cons or fallback approaches. There was a lengthy discussion about this in the last weekly Dev Chat and this post was meant to collect the thoughts from that discussion for reference so it can continue.
  - Andrew Roberts 2:51 am on March 1, 2017
    
    Agreed 🙂 and thanks for doing the post.
    
    Was hoping to nudge the readers here a bit towards the desired outcome 🙂
- James Nylen 1:45 pm on March 1, 2017
  
  As one of the developers who will be working on the editor I agree with and support this decision.
- Aaron Jorbin 5:52 pm on March 1, 2017
  
  > We understand that this would impact a lot of people even with very very small percentages. Our CDN gets up to 1 million hits an hour
  
  Do your browser version stats align with the ones posted here?
  - Andrew Roberts 8:28 pm on March 1, 2017
    
    Unfortunately CloudFront only gives us aggregate numbers. I can try and see what some of our other analytics give us.
    
    IE8-10 would fall under “Other” in these stats:
    https://cloudup.com/cVORSXFG8B7
    - Andrew Ozz 11:31 pm on March 1, 2017
      
      I’d imagine they will be able to give you a list of all browsers, similarly to the Download Data (.csv) link at http://gs.statcounter.com/ that was used for the above stats.
      
      Would be pretty good to compare.
      - Andrew Roberts 12:21 am on March 2, 2017
        
        Done!
        https://cloudup.com/caZ8koOnSNp
        
        Of the 25.6M unique visitors we had in the past two weeks 93,957 were from IE8-10… or 0.366% in total.
        
        To compare GlobalStats vs. cdn.tinymce.com:
        IE8: 0.41% vs 0.082%
        IE9: 0.26% vs 0.129%
        IE10: 0.26% vs 0.155%
        IE11: 3.79% vs 4.633%
        
        jteague 12:42 am on March 2, 2017
        
        4.7.3
        
        there might be a better option availalable
        
        jteague 12:44 am on March 2, 2017
        
        We’ll, that was a mobile fail at replying. Meant to say thank you for the stats update.
        
        Andrew Ozz 1:33 am on March 2, 2017
        
        Thanks! These look a lot lower than all other stats. Perhaps they are more “regional” to only North America + Western Europe?
        
        As far as I’ve seen most old IE users come from Asia.
        
        Andrew Roberts 2:00 am on March 2, 2017
        
        May be our stats are a bit skewed by regional differences in CDN usage.
        
        About 50% of our traffic is US and UK.
        
        India is the top Asian country and it is only showing 3.33%, Vietnam is 2.70%… etc. It does seem that Asia is under-represented.
Jon Brown 2:21 am on March 1, 2017

I too feel this summary focuses way to much on the cons on dropping support without giving more consideration to the benefits to the 99% that don’t use IE8-10. I’m very strongly in favor of deprecating IE8-10 and providing a textarea for the <1%.

The summary states that wp.com's numbers are very similar to global stats, but I'm guessing those are still the numbers for view platform wide? It would be MUCH more interesting to get the browser stats for visits to the _edit post page_, or the very least the dashboard rather than the platform wide numbers that include public viewers. If I'm guessing wrong, please clarify what those stats refer to exactly?
minionnz 4:15 am on March 1, 2017

Akismet is already included as a pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party on install, so why can’t we have a “Legacy Editor Support” plugin for the editor?
- djsteveb 4:40 am on March 1, 2017
  
  Wouldn’t it also be possible to keep a basic text editor as default and offer a shiny enhanced editor as an extra pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party option? On the update screen have a “wp suggests this cool new editor if you have a modern computer and want to add a small plugin crafted by the coreCore Core is the set of software required to run WordPress. The Core Development Team builds WordPress. team click here to install”
  
  or something like that. (?)
  
  take out the space of the video and put suggested enhancements right there in its place. Of course, you could still add a thumbnail svg of the video and offer a click to load / play.
  
  This would fall in line with some of the other suggestions from the tavern comments recently (https://wptavern.com/solving-the-mystery-of-how-people-actually-use-wordpress#comment-210743 )
- James Nylen 1:56 pm on March 1, 2017
  
  Automattic has a dedicated team working on the Akismet pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party. We shouldn’t really expect the same for an editor fallback for old versions of IE.
jteague 4:59 am on March 1, 2017

What it’s worth, I’m siding with the tinyMCE team. I’d prefer a programmatic fallback based on browser detection rather than adding to user workload/decisions with a pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party. We build Polymer based apps and package our own editor. Progressive web apps and web components are here now, and WordPress needs to advance and there’s so much coolness that can be done if we commit to stepping out beyond decades old tech. The TinyMCE team shouldn’t be burdoned with having to maintain legacy versions of the editor for future versions for .8% or fewer users. Thanks for listening.
- James Nylen 1:48 pm on March 1, 2017
  
  I’d prefer a programmatic fallback based on browser detection rather than adding to user workload/decisions with a pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party.
  
  Same here. Also, there is actually less risk of breakage and less maintenance burden of including the fallback in coreCore Core is the set of software required to run WordPress. The Core Development Team builds WordPress. as compared to a plugin. If the editor fallback code is in core, it’s more visible and more people can work on it. If it’s in a plugin, it’s more likely that it will break, no one will notice, and no one who has access to the plugin will be around to fix it.
Mike Nelson 5:45 pm on March 1, 2017

I vote fallback to plaintext editor. If there is a sufficient demand I suspect someone from the community will create a pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party for legacy TinyMCE for older browsers.
I’m thinking this approach follows WordPress’ 80% rule:

The rule of thumb is that the coreCore Core is the set of software required to run WordPress. The Core Development Team builds WordPress. should provide features that 80% or more of end users will actually appreciate and use.

from https://wordpress.org/about/philosophy/. This new version of TinyMCE will be breaking functionality, though, not adding. But still, it’s 0.8% of users benefiting from including two versions of TinyMCE, not 20%
Andrew Ozz 1:29 am on March 2, 2017

Most people here would really like to drop support for old IE. I agree. From a developer point of view, this will simplify many things and probably cut down testing/debugging time.

However from a user point of view things aren’t as simple. Looking at all stats, IE8 + IE9 + IE10 range from 0.93% (statcounter) to 2.5% (w3counter, looking at February 2017 in the IE trends chart here: https://www.w3counter.com/trends).

To properly evaluate the impact, we need to multiply the above percentages by the number of WordPress users. This is quite hard but a (very) raw estimate is:
27.6% (from w3tech.com) X 290 million (active websites from internetlivestats.com) gives us about 80 million active sites that run WordPress.

If we disregard users that have more than one WordPress site and sites with more than one user (consider them roughly equal), we have 744,000 to 2,000,000 IE8 + IE9 + IE10 users.

This is a huge amount of people. Don’t think we can afford to “leave them behind” with just a textarea. So if we want to drop support for old IE in the editor we will have to build a seamless fallback for it. This is not going to be easy/fast and will limit the functionality of the “new” editor to only inside the current iframeiframe iFrame is an acronym for an inline frame. An iFrame is used inside a webpage to load another HTML document and render it. This HTML document may also contain JavaScript and/or CSS which is loaded at the time when iframe tag is parsed by the user’s browser.. Everything outside of it will have to be IE8-compat.
- jteague 2:13 am on March 2, 2017
  
  I’m still of the mind that if there is enough need we’ll see community pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party options for those running legacy browsers this aged. Yes, it’s a significant number, but far below the 80 or the 20.
  - Andrew Ozz 2:21 am on March 2, 2017
    
    Don’t think we can apply the 80/20 rule here. We are talking about leaving 700k to 2m users “behind”.
    
    Also, don’t think a pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party will be a good option as a fallback. That fallback will have to be in coreCore Core is the set of software required to run WordPress. The Core Development Team builds WordPress. and “seamless” if we decide to drop old IE.
    - Jon Brown 4:12 am on March 8, 2017
      
      Right, but it’s not 80/20… it’s 99/1. 1% is a significant, but it’s wroth of a textarea fallback and nothing more.
      
      As I said above, we _really_ need to look at the status hitting the editor, NOT general browsing. Way too many old bots and such use IE8-10 as a user agent. We need to consider only the numbers that would actually hit the editor and I strongly suspect that’s way less than the global numbers suggest.
    - FolioVision 12:34 am on June 13, 2017
      
      Andrew, all of those old IE users (much less than 1% as I expect most people publishing in WordPress as opposed to just reading WordPress websites would not be using IE at all) can easily crossgrade to a relatively recent version of Firefox with all of the bells and whistles of modern publishing. It seems to me you are creating an accessibilityAccessibility Accessibility (commonly shortened to a11y) refers to the design of products, devices, services, or environments for people with disabilities. The concept of accessible design ensures both “direct access” (i.e. unassisted) and “indirect access” meaning compatibility with a person’s assistive technology (for example, computer screen readers). (https://en.wikipedia.org/wiki/Accessibility) issue where there is none. Plus those using old IE browsers are likely not to be updating their WordPress (as they very well may be using old PHPPHP The web scripting language in which WordPress is primarily architected. WordPress requires PHP 5.6.20 or higher too).
      
      Jon’s quite right. Text-field fallback with a hint about which alternative browsers would run in their OS would be more than enough.
- WraithKenny 2:48 pm on March 13, 2017
  
  If this are users, and 1% (according to the philosophy page) of users are “content creators” (who actually use the editor!), the number affected is much less, ain’t it? 7K to 20K users. “Design for the majority.”
  - WraithKenny 2:53 pm on March 13, 2017
    
    The “content creator” portion of the user populate almost certainly and heavily skews towards modern browsers. There isn’t any reason to support IE8-10. There is no real reason to support less than evergreen.
dmccan 1:50 am on March 2, 2017

There is another class of user to consider: those who may prefer the current editor over the new one. Might there be some features or abilities that one can do with the current editor that won’t be in the new one at launch? If so, the number of users who “must” have that feature will far outnumber the IE 8-10 users. Also, people are resistant to change.

It is a safe forecast that over the next two years the percentage of IE 8-10 users is going to further decline to the point that the fallback, if it is just of old versions of IE, can and should be removed. A pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party version that will be supported for 2 years (or whatever) might ease the transition.
- Jonathan Desrosiers 2:32 am on March 2, 2017
  
  If a decision is made to drop a feature, there will most likely be a very good reason or a replacement that serves as an improvement.
  
  If there are some things that are left behind and there is enough need, I think that would be something better suited for pluginPlugin A plugin is a piece of software containing a group of functions that can be added to a WordPress website. They can extend functionality or add new features to your WordPress websites. WordPress plugins are written in the PHP programming language and integrate seamlessly with WordPress. These can be free in the WordPress.org Plugin Directory https://wordpress.org/plugins/ or can be cost-based plugin from a third-party territory.
  
  There was a good example of this in 4.7 with the removal of the underline and justify buttons. There are several plugins in the repo that add those two buttons back, and more that allow users to further change the buttons in the editor.
gharr99 6:35 pm on June 12, 2017

We are using IE 11 here at my work. We noticed after upgrading to WP 4.8 the kitchen sinkKitchen Sink When using the WYSIWYG (What You See Is What You Get) editor in WordPress, you can expand the capabilities to allow more options. This expanded area is called the "Kitchen Sink." and the TinyMCE stopped being displayed.

I have no issue with dropping support for IE 8, 9 , 10. But I thought there would still be support for IE 11.