And it’s not just the aggravation of having the cursor come out of a text field after a typo and backspace taking you away from the page altogether.
There is an inevitable back and forth in software: users pull one way, IT the other. In the 80’s users pulled away from the mainframe / minicomputer paradigm, putting in PC’s (or Macs) so that they could do the sort of playing with numbers (or images, or text) that people who run transactional machines could never allow.
IT wasn’t too concerned initially, until the lowered cost of Ethernet made it practicable to replace the mainframe screens with green screen windows in operating environments like Windows 3.x or OS/2 2.x. Suddenly it became an issue for IT, an issue that was based around a perceived loss of control, and the result was to pull everything back to a refurbished mainframe, aka a web application server.
The problem that created was seen as a lack of ‘dynamism’ in the early web UI’s, when in fact it was a lack of control. Anyone could do what they liked with a PC or Mac without bothering anyone else. Web applications, on the other hand, only allowed you to do what they were built to do, exactly like the mainframe. The screens were prettier, though nowhere near as efficient for data entry.
It may have gotten better, but HTML5 is still not as good as HyperCard, and it’s 30 years later. And writing markup using a text editor is inane. Word has a far more flexible and complex markup than HTML, but you don’t see secretaries writing Word documents in Sublime Text (‘the night where all cows are black’ — Hegel).
As far as I can tell, it makes ‘developers’ who are actually ‘bourne-again secretarial hacks’ (er, BASH, for short) feel like ‘real’ developers.
After years of developing more capable protocols to work around HTTP, along comes the inventor of HTTP, converts it to an interface and announces that stateless, resource-oriented architecture is the way of the future (although it’s 100% dependent on the very stateful internet underneath it), thus smashing any arbitrarily capable protocol by putting HTTP (re-bourne as an interface) in front and back of it.
And if the representation being transferred includes derivations or aggregations, trying to update changes is akin to trying to update 100 accounts condensed into a simple bank statement, when the numbers are edited in Excel and uploaded back to the bank.
The oversimplification results, as usual, in over-complication elsewhere. An application is, after all, a bloody state-machine.
Enter AJAX to make things more ‘dynamic’. But that’s solving a different, largely unimportant problem. The problem is, again, control.
So we’re back to the fact that many people, not only me, don’t like web applications. In fact they’d rather use Excel, despite its limitations.
If you disbelieve me, then you need to come up with a reason so many companies that began writing purely web applications are desperately trying to create desktop versions (using the same web-oriented technologies), i.e. #Slack, #LinkedIn, #Messenger, #eBay etc.
But it doesn’t solve the problem of control, it merely avoids it. IT wants control, users want control. At some point the battle has to come to a head.
Who will win?
I’ll give you a hint, it won’t be IT. And the reason is that user-based control is really not control, since each user will use technology differently.
We’ve been trying to ‘control’ or ‘manage’ technology in one form or another for thousands of years. We’ve never succeeded, because technology defines man as much as man invents technology. The ‘being’ of human being is always techne.
Claude Ciborra’s idea of ‘cultivating’ technology rather than ‘managing’ it is a promising start, but one that leaves IT out in the cold.