Slashdot Mirror
Research Proving People Don't RTFM, Resent 'Over-Featured' Products, Wins Ig Nobel Prize (improbable.com)
An anonymous reader writes:
Thursday the humor magazine Annals of Improbable Research held their 28th annual ceremony recognizing the real (but unusual) scientific research papers "that make people laugh, then think." And winning this year's coveted Literature prize was a paper titled "Life Is Too Short to RTFM: How Users Relate to Documentation and Excess Features in Consumer Products," which concluded that most people really, truly don't read the manual, "and most do not use all the features of the products that they own and use regularly..."
"Over-featuring and being forced to consult manuals also appears to cause negative emotional experiences."
Another team measured "the frequency, motivation, and effects of shouting and cursing while driving an automobile," which won them the Ig Nobel Peace Prize. Other topics of research included self-colonoscopies, removing kidney stones with roller coasters, and (theoretical) cannibalism. "Acceptance speeches are limited to 60 seconds," reports Ars Technica, "strictly enforced by an eight-year-old girl nicknamed 'Miss Sweetie-Poo,' who will interrupt those who exceed the time limit by repeating, 'Please stop. I'm bored.' Until they stop."
You can watch the whole wacky ceremony on YouTube. The awards are presented by actual Nobel Prize laureates -- and at least one past winner of an Ig Nobel Prize later went on to win an actual Nobel Prize.
"Over-featuring and being forced to consult manuals also appears to cause negative emotional experiences."
Another team measured "the frequency, motivation, and effects of shouting and cursing while driving an automobile," which won them the Ig Nobel Peace Prize. Other topics of research included self-colonoscopies, removing kidney stones with roller coasters, and (theoretical) cannibalism. "Acceptance speeches are limited to 60 seconds," reports Ars Technica, "strictly enforced by an eight-year-old girl nicknamed 'Miss Sweetie-Poo,' who will interrupt those who exceed the time limit by repeating, 'Please stop. I'm bored.' Until they stop."
You can watch the whole wacky ceremony on YouTube. The awards are presented by actual Nobel Prize laureates -- and at least one past winner of an Ig Nobel Prize later went on to win an actual Nobel Prize.
Research like this is causing software to be increasingly dumbed down to a point where it is extremely difficult to use. In the past you could configure software to work in a way you found desirable and productive, but now all the sophisticated is being removed and you're forced to work the way the UX designers dictate. Take Firefox:
I could go on all day with Firefox, but dumbing down of the browser by removing features and options has turned it into a nightmare to use. The same is very much true of Windows 10 which is an absolute train wreck. I've found myself increasingly moving away from commercial software produced by UX designers to FOSS produced by programmers simply because I want software that works.
The thing is, it's not just technical users who hate what UX designers are doing to software, and casual users I speak to also hate the constant UI changes, the hiding of features and the removal of options. Now here we have some worthless 'intellectuals' being given a Nobel Prize for telling people to fuck up their software.
It's little wonder we're moving to a world where computes are becoming less sophisticated and turning into machines for running 'apps' that you can only get from a curated store that bans anything remotely useful. With the direction computing is headed, I think I'll just go and live in a cave.
I do, or at least try to, read the manual. But then I'm a sysadmin. Even so, I typically don't read the entire thing because it's just too tedious, annoying, incomprehensible, or whatever. Most manuals suck giant balls through tiny straws. I try to get a general sense of the thing and fill in the needed-now blanks until it functions enough to make do, and maybe will get back to making it do more later. I typically notice soon enough when I need more power, at which point I'll re-check the manual and if unsatisfactory might have to conclude I'm in need of a different tool to get the job done.
Even those manuals ostensibly written for professional manual-readers like me tend to uselessness: Take almost any "reference manual" for networking equipment. It'll be chock-full of "foo bar <parameter>: Set the bar parameter to the foo subsystem", in the case of CLI and lots of pictures on how to click your way around the equivalent GUI setting. But no word on what it means, how it relates to the rest, what you can do with it, when you might need it, gotchas and other considerations. All conspicuously missing.
This was one big large selling point for Unix, in that it came with honest manpages. Written in troff so they're the same whether nroff puts'em on your screen or you look'em up in the paper version (or print them yourself). That typesetter, incidentally, was the killer app for Unix within AT&T. But I digress. Those manpages are written for technicians to get shit done. Good ones tell you what the thing does, how to make it dance, what not to do, and what to watch out for. They work, for a very specific audience.
It doesn't surprise me that people trained from birth that everything is supposed to be "intuitive, no training needed" will not bother with the manual, certainly not seeing the quality of said manuals. It doesn't really bother me that people don't read manuals, since most people are followers--they'll only learn if someone shows'em. So maybe it's a good idea to bring back those classes "learn to use your computer" you used to get with your (professional get-shit-done hardware) purchase in the microcomputer era. It does bother me when I need to make hardware dance and the manual turns out to be unreadable and wrong.
In my experience, applications that have enough different features an complexity to really be a problem often support doing several different tasks. Often it can be made much more usable by designing a UI around the distinct jobs one can do using the software, a UI based on workflow. I'm not going to say wizards exactly, but UI flows designed for specific tasks the user wants to accomplish.
The UI often represents the underlying underlying data rather than the tasks. You know for sure that you have this kind of UI if it has different sections for different kinds of objects. Of these different kinds of objects map to different table in your database, you definitely have a data-centric UI instead of a task-centric one.
My server backup software originally had a data-centric UI. It was basically alot like managing hosting accounts. It had a page for servers - listing, adding, removing, and editing them. It had a page for DNS names - adding, removing, and editing them. There were a couple other pages like that, for manahaing the objects in the application. That worked great for me. Users didn't like it.
We added another UI that started with this page:
Add a new server
Restore file or server
Manage billing
Other tasks
Clicking "add a new server" took the user through the steps of adding a new server - including any domain names related to that server. Clicking "restore a file or server" took them through the steps to do that. At each step, the only saw the options relevant to that step.
The task-centric model is a great way to manage complexity. The data-centric UI can also be useful at times, but it's inherently more difficult to learn and use in many cases. Some applications warrant having both options. We kept both for the server backup system. Customers used the task-centric, wizard-like UI for common tasks. For less-common tasks, the data-centric UI was more flexible.
Yes, well... Is it a surprise when they've tossed all that we've learned about UI making?
Clickable objects are no longer clearly marked as such. Different kinds of content are no longer clearly distinguished from one another. Contectual information on mouse-over or right click or even F1 is hit and miss, but usually miss.
And be honest, how often do you go "What kind of moron from Bizarro Land would name this function that and put it there??!!".
Or lists... on one frame, it's exportable on the second it's sortable and on the third you can do multi-selection. But not one of them can do all three.
Not to forget that if whatever you are using happens to have multi-platform apps, GUIs or whatever, you can bet your ass they won't have been developed by the same team. So you not only need to handle each and every single app with a lot of TLC for them to even remotely do what you want, you'll have to learn to do it differently on your MacOS laptop, on your Android phone and on your Windows desktop. If you're especially lucky, the online interface will behave differently depending on the browser too.
And let's be exceptionally frank here, writing a good manual is an artform few have ever mastered. 90% of my use cases I find my answers on some message board online and certainly not in a manual.
OR, as it's the case with our current backup software, the manual is easily 1000 pages. Now, if I were tasked solely with pampering our backup software, you could argue that that is doable and you'd be right. But I also have to pamper the storage environment (with several products, of course), the Cisco UCS, SAN infrastructure and Vmware Virtualization as a cherry on top.
And in order to not kill motherfuckers daily, I strictly adhere to my 8 hour work days, except for emergencies and maintenance tasks that cannot be done on hours.
So imagine this new-fangled dohicky coming along expecting me to forego everything I thought I knew about gadgets and do it their way now because reasons. Yeaaah, no. Go fuck yourself, would you kindly?