Which is especially unforgivable, given that responsive design is already built in to HTML. But the modern design paradigm is for the web designer to take control of everything, which means that they need to know everything about how your device displays, and basically come up with a different design for every possible device.
Yesterday I was trying to cancel a subscription to a music streaming service, using my 13" laptop. After clicking on the ‘Cancel’ button, I was taken to a survey page where I was asked why I wanted to cancel. After making my selection, I couldn’t locate the ‘Next’ or ‘Continue’ button. I finally had to maximize the Chrome browser page, close the Downloaded files bar at the bottom of the page, and only then could I see just the top of the button that allowed me to continue. I still don’t know what the button actually said.
“I changed my mind, I’d like to start a monthly subscription to this product instead. Prepaid for one year.”
(I realized I posted my “Stupid Software” rant in the “Stupid Product” thread… my bad. Here it is, in its proper place.)
Just got home from the orthopedic specialist. Upon check-in they handed me a tablet that was running the absolute worst software in kiosk mode, allowing me to fill out those endless medical questionnaires electronically. That’s a win, right?
The initial questions were spoon fed a page at a time and they provided only a tiny text field for things like lists of medications and prior surgeries, all to be fat-fingered in using the terribly laggy keyboard.
If you click the “back” button because you made a mistake, the entire thing resets to the beginning and forgets all of the information you entered.
Best of all, when it gets to the actual list of questions, it shows something like “Mother” then two questions like “Heart condition?” and “High blood pressure?” with a big “Next” arrow…
What’s the catch? It turns out that the list of “Mother” questions is actually 20 or 30 questions long, but there is no indication at all that the list scrolls. The questions are shown in large font in kiosk mode with no scroll bars or indications that there is more, but with a big arrow to go to the next page. Unless you use your finger to scroll down, you will never see the remaining questions, and they are optional, so the app will happily let you go to the next page.
If anyone actually cared (or could do anything about it) I would have explained to the staff that it might be worth looking at why every patient only answers the first two questions on every single page, leaving the rest blank.
If anything, this kind of device is not very usable by many people with temporary or permanent disabilities–like patients at an orthopedic specialist facility.
As an IT professional in the pharma industry, this kind of laziness in application design is inexcusable. I would not allow an application like that out the door on my watch!
USPS self service kiosks. If you want to print a shipping label for a package it walks you through the process step by step – enter the recipient’s zip code, then their street address, then apartment number if necessary, then their name, then the return address, weigh the package, measure the package, select what level of service you want, then you get to a screen with all the information you entered to confirm everything’s correct.
If you get to that confirmation screen and realize you need to correct something, you can’t simply re-enter that one thing. You have to tap the “back” button multiple times to go back to the step where you enter that information, in my case the apartment number. But that’s not the worst part. Then it makes you go through all the subsequent steps again, and it doesn’t remember the information you entered previously. So after correcting the apartment number I had to re-enter the recipient’s name, and my return address, weigh the package again, measure it again, and so forth.
Is it posdible that the tenths digits are changing too fast to see? The digit “8” is, after all, what you get if all the “segments” are on because the indivdual numbers have blurred together.
Of course, you’re right in that you don’t need that amount of precision for what you’re doing. This is the actual issue: Why even design it with tenths of a second?
Interesting thought, but I don’t think so. There are enough cleared segments between numbers that at a tenth of a second you’d see them blinking (they’d be off for .2 or .3 seconds in some cases).
They possibly didn’t design it with tenths - whatever library they are pulling the result from (or more likely the one used to format the value) may have that as the default precision.
This is a fun one. Many programs do the thing where when you put the mouse on some element of the program a little box (tooltip) will appear with some words or something in it. Firefox has a bug where the box will not always disappear when Firefox is no longer the foreground window. This is annoying, as then there is a little box floating around on the screen covering up whatever else you happen to be doing until you go back to Firefox, and get the box to go away.
Through the power of open source software, the bug was finally fixed, after 22 years. It was fixed by a new contributor to Firefox, who had just made an account 2 weeks before submitting the fix. It is possible the bug is older than the person who fixed it.
One person on Mastadon gives a likely explanation:
Maybe they experienced the bug as a child, it annoyed them to their core, they went to university to learn coding, then graduated, made an account, and fixed it. Life goals complete.
Yowsa. They were using a mouseOut event to dismiss the tip box, when it needed a blur (or onBlur) event. I wish I was still teaching programming at my college, first to demonstrate the difference between those two events (I think it was even a question on one of my quizzes), but more importantly that anyone can make improvements to open-source software.
What I cannot abide are web sites where you want to do something that requires filling out forms and when you are finished and try to go on, they tell you there is an error and won’t tell you where it is.
I just tried to buy tickets for the Adirondack train Montreal to NYC. I went to the Amtrak web site, found the train (there is only one) and tried to choose it. There was a continue button at the bottom of the screen, but it was dead. Finally I realized that you had to click on your “choice” even though there was only one. So I chose that train and now the continue button was live. I then spent 10 minutes going through the regamerole of enterting my name, my nationality, passport number. Then the same for my wife. Then I entered the CC number, the name on the card, the address (even there, the next-to-last line asked state/province, but only offered states. The last line asked country and when I clicked on Canada, provinces came up on the previous line) and postal/zip code. I entered all that and at the bottom of the screen was a purchase button. But it was dead and I still have no idea what was missing. Finally, I found the Amtrak telephone number. They had to transfer the call to a Canadian agent but she was extemely pleasant and took all that same information and sold me the tickets.
Whoever designed that web site should be taken out and shot. Or at least fired. Do they never try it from the POV of an ordinary user?
I really hate how every single streaming service’s video player has a COMPLETELY different video interface, especially in regards to how rewinding works.
Here’s what I want to happen
-
You press back once, you go back 15 seconds
-
You press forward once, you go forward 30 seconds.
-
Holding rewind or forward will steadily rewind that way if not smoothly in 30 second increments.
Instead what I see all the time is hitting rewind or forward once goes into light speed and goes back minutes at a time. If you ever tried using the PS4/PS5’s native media player it might be the worst thing I’ve ever seen since if you rewind you’re going back minimum 2-3 minutes, it’s so fast. On the opposite problem you got video players where you hit rewind and it will rewind or go forward 5 seconds at a time which is far too short if you actually want to catch a line you missed, and it’s also far too slow if you actually want to rewind back or go forward to a specific scene. I think YouTube currently has the best video player I’ve seen in TV mode.
this is game related but in Sid Meiers Civilization 6 if you build a military unit you can click a button to have it auto-explore the world which helps advance gameplay in several ways but there is an idiotic bug that they never fixed
if you run into the “barbarians” who fight and pillage everything and one of your units gets attacked Now you can manually move the unit away from the line of fire to safety but if you’re still close by and you turn on the auto-explore button it will go right back to where you movedit from getting attacked and killed in the process…
Now I don’t know if this one is because of the way it was made it a product of making it compatible with modern hardware but in diablo 1 the character cant use any ranged weapon properly because of the point-and-click movement …you aim and start firing arrows or use your sling etc your person will run to where your aiming at and get stomped by the horde that’s trying to kill you
I was a technical writer for over twenty years, serving as a bridge between engineers/designers/programmers and ordinary users. One of the problems I faced was trying to explain to the engineers (etc.) that what they thought of as ordinary users, were not what were actually ordinary users.
I remember documenting a restaurant accounting program for a POS system years ago. The engineers (etc.) thought they were creating the system for experienced accountants. Nope, they were creating a system for a 20-year-old part-time bookkeeper who doubled as a bartender before entering the day’s receipts into the system. The engineers’ “user manual draft” assumed that all users were either experienced accountants or computer software engineers, and our marketing department’s user surveys demonstrated that our users were neither. I hate to say it, but I “dumbed it down,” and while the engineers hated that, our users loved having something that gave them clear instructions on how to complete tasks with the software.
On another job, the engineers’ user manual draft for the project was about as thick as the Manhattan phone book, but it really only needed to be about one-quarter (if that) the size. Why? Because it was so full of “Caution: Not to be used except by authorized personnel,” and “Warning: Attempting to impinge the gizmo on the framistan may result in injury or death.” Well, given that the gizmo and the framistan were both inside a box marked “No user serviceable parts inside, access by authorized repair personnel only,” no ordinary user had any reason to pay attention to the warning. Yet the ordinary user was presented with a Manhattan phone book full of such things, of which maybe one-quarter was useful to the ordinary user.
Here’s the thing about ordinary users: if they don’t know that something is possible, they won’t do it because they don’t know about it. Tell them that it is possible, but they shouldn’t (“authorized personnel only,” and all that) and they’ll undoubtedly try it at some point.
Returning to Hari’s point: I beg engineers (etc.), to not make things really, really cool, because they show off your skills, while confusing the hell out of users. Rather, make things really, really easy for the ordinary user, who you can assume is as dumb as a bag of rocks when it comes to your subject matter. Remember, the ordinary user buys your products and thus pays your salary, which means that they are more important than the applause from the engineers in the lunchroom when you’ve confused the hell out of some ordinary user.
In my life as a computer programmer/analyst, a co-worker and I developed ‘the rule of cool’. That rule stated that if you looked at my design/code and said “that’s really cool”, it had to be redesigned/coded to make it maintainable by someone else in the future.
1000x this. I have to maintain some software that was written 20 years ago by a guy who did all kinds of “cool” stuff in his code, just because he could. With no comments either. It’s a nightmare for me to try to make any changes to it, because I have to spend half a day first just trying to figure out what the hell he’s doing.
On the basis of this, at least half the developers in the world need to be taken out and shot (or dug up, resuscitated, and taken out and shot). The lack of internal documentation influenced me to the point that by the midpoint of my career at least 20% of my code — even things I wrote for my own consumption, since I never assumed I’d recall exactly what I was doing — consisted of comments.
That was actually a plot point in Silicon Valley. When Richard has the beta version of Pied Piper ready, everyone he shows it to tell him how cool it is, except for one woman at the venture capital firm who’s funding them. But she just says if everyone else says it’s good, I guess I must be missing something. Then it’s a flop when it actually goes live, because all the people telling him how great it was were other engineers, who were praising things ordinary users don’t care about.
I made this point recently. The audience for most comments is future you.
And sometimes the “ordinary user” is very different from a “new user”. If the ordinary user is somebody who uses the software frequently, making things easier for an experienced user, but sacrificing the experience of a new user, is probably the better balance. For example, a method to automate repeated tasks may be too complex for a new user, but extremely useful for everyone else.
Our standard joke was ‘When a new Design/Coding graduate joins the company it takes about 6 months to re-train them to produce something useful’.
This is also why Functional Programming didn’t get very far.
It seems really cool, until you need to make any changes to your programs (let alone anyone else’s code)