You can file this under “No one does proper systems design anymore.” And this is mostly so you can get a feeling for the pinball-like workings of my mind….
Earlier today, I was reading a posting entitled Short Circuit from the Institute for Justice. (How did I get there? Via The Volokh Conspiracy.) Anyway one of the stories was about a pizza company being sued, under the Americans With Disabilities Act, for not providing accessible websites or mobile apps. Which caused me to think immediately of the problems that WordPress is having with their Super-duper-new-and-improved-editor, Gutenberg.
I spent enough time with it to let me know I didn’t need any of its new and (not so much) improved features, but the folks who use accessibility software – like screen readers – can’t use it at all. (OK maybe they have improved it, but the next story makes me wonder.)
Rian Rietveld published the following article: I have resigned as the WordPress accessibility team lead. Here is why. In that article she quotes a former teammate, Andrea.
The main reason for this lack of overall accessibility is in the overall Gutenberg design, where accessibility hasn’t been incorporated in the design process.
Which threw me all the way back to my IBM database days. (How many people remember Information Management System, more generally known as IMS, or DB2?) There was a systems architect who gave a presentation at a bunch of the big conventions (IBM conventions = Guide, and Share). His point was that if you screw up the architecture, it is impossible to fix the problems by coding small workarounds. If you built a 3 bedroom house and then discover you need a 4th bedroom, you may need a bigger furnace/air conditioner. A bigger septic system. etc. While all of this can be done, it can’t be done in a hurry, or on the cheap. And if you aren’t careful, the addition is always going to look wrong when next to the rest of the house. The same thing applies in software. (OR why do you think that after a couple of DECADES of working on computer security, we still don’t have very much? If it was easy…)
There is much in Rian’s article about “We should have written the issues differently” in order to get the programmers to solve them. (Issues that were solved, were broken again in subsequent updates because of development chaos.)
I’m not sure how WordPress development is organized, but I have used it long enough, and dealt with their so-called support often enough, to believe that it isn’t very well organized. Requirements? Documentation? Locked code? Version control? Doesn’t sound like it exists in WordPress. And that’s before you get to issues of architecture, design, etc. (Programmers have to WANT to solve issues and fix bugs? I’m glad they aren’t in charge of supporting my brokerage house’s computer system.) For another perspective on things like User Interface (UI) and Application Programing Interface (API) chaos see the article from November of last year, titled, Pressbooks and Gutenberg.
The lack of clarity in Gutenberg’s development process has hindered us from integrating Gutenberg into our roadmap. We are now two weeks from its production release, and Gutenberg’s API freeze is not yet complete. We’ve been tracking blocking issues over the last year and a half and have tried to contribute where possible, but ongoing API and user interface changes have made it difficult for us to keep on top of things without neglecting Pressbooks core development, and have made us hesitant to invest our limited resources in building on a codebase that has not yet stabilized.
The UI should have been finalized in the design phase – or nearly so. And the API isn’t defined a few weeks before they are going to go live? That should have also, mostly, been defined in requirements and design, before coding started. (What functions exist? What functions do you want to expose via the API? Structure? Error handling? Security?) But then I never worked on systems where we just made it up as we went along.
Now I’m not using Gutenberg; I’m using the “Classic Editor” because it doesn’t suck and it does everything I need it to do, in a way I have been doing things since before there was WordPress. (I was using Generalized Markup Language long before it was extended/co-opted to become Hypertext Markup Language. And WYSIWYG is almost never what I want. Of course I can actually touch type, so that’s a help.) Gutenberg is actually their New, new-editor. Their old, new-editor (that didn’t have a fancy name) also didn’t offer me anything much, even though they tried for a couple of years to get me to use it. They are currently talking about “end of life” for the classic editor. Guess I need to start researching blogging software again.
Maybe the folks working on WordPress will get the accessibility issues ironed out. I hope they do, but I don’t really have much hope that it will get done the way things are going. Consider this: How is it, that in 2017 or 2018, when they started this project, that accessibility of the finished product was not a primary part of the design? It was not even a 2nd thought, but pretty much completely ignored, until they got a ton of bad press on the subject.
This all started because of legal problems a pizza company ran into over accessibility via the web. I’m no lawyer, but my guess is that WordPress and Co. won’t be in hot water as long as the Classic editor supports everything required. (Though making one class of customers go through the back door doesn’t look good in the 21st Century.) WordPress isn’t the first organization to ignore issues of accessibility, and the way programmers work on things, I doubt they will be the last.