It probably does, and the agent is verifying that he or she is talking to the right person (shocking as it may seem, phone systems have been known to misroute calls). Or the Customer Service application may show a call queue on the screen, and require the agent to select one based on the phone number.
In a more sensitive application such as banking, asking for a phone number, partial SSN, address, etc is a way of ensuring that the agent isn’t talking to someone who didn’t just find a cell phone.
My favorite, though it may be an urban legend: “If you are calling from a rotary phone, press 0” (supposedly the IRS, which wouldn’t surprise me a bit).
The one which nearly caused a phone-shaped dent in my wall was the oil company which allowed me to press 0 to speak with an agent, asked for — and presumably verified — my many-digit account number and PIN, and then told me the office was closed. I used to design and implement phone-menu systems, and a cardinal rule was not to present the caller with options it can’t currently fulfill (preferably advising the caller as to the reason).
Probably UI related: Taco Bell employees have been asking me for my name recently when I order. But when they give me my food they randomly do one of 4 things:
Call out the order number (printed on the receipt); call out the meal number (e.g. “7”); call out the stuff in the meal (e.g. “Nachos supreme with two soft tacos”); call out the customer name (“Joe”).
They randomly choose from the first three options, and have never called me by name.
POS terminals are annoying because they are all different. If I hit “cancel” it’s just about even odds that Credit will be chosen (the desired option) or my transaction will be canceled (not what I want).
In one particularly annoying one I struggled for a couple of minutes at making the selections, somehow choosing “EBT” instead of “Credit” each time. It turned out that the physical buttons had labels printed not on the keys themselves, but equidistant between the keys. This meant it was ambiguous whether a label was for the key above or the key below. Why make users inspect the whole keypad to see if it is a “labels above” or “labels below” version?
In the States we seem to meekly accept that there is a difference between Credit and Debit, and certainly there are differences in the way those are processed, but why should it affect the customer? A customer provides a card, authenticates/authorizes, gets charged, takes purchase.
ETA: Why, oh why, doesn’t Apple Pay just work everywhere? Clearly it’s not implemented in many places, but there are so many times that I see that the workers have to do special things that sometimes work and sometimes don’t. Apple Pay is awesome when it works. I think the banks are intentionally making it difficult.
And I designed software UI’s back before there were National University degrees in such things. I abhor shitty UI design.
You could be talking about any debit/credit terminal here. What you should know is that these are installer/field-programmable system, and absolutely nothing is given less time or shorter shrift than programming and polishing the UI. The minimum number of steps from factory setup to doing what the merchant wants is what’s given. So there are extra steps, useful-but-not-essential steps missing, random wait times, and - absolutely worst of all - completely random order of steps, so that unless you memorize the system of places you visit regularly, you’re always waiting to enter the wrong thing.
This is not an inherent problem. It isn’t even a “crappy system/software/firmware” problem. It is 100% the failing of the vendor who installs these systems and can’t be arsed to set them up according to a consistent protocol, or in a way that makes their use smooth and sensible.
But bitch, and everyone up to the regional manager just shrugs and says, “That’s the way they work.”
And for the record, it should make ABSOLUTELY NO DIFFERENCE WHEN YOU INSERT OR SWIPE YOUR CARD… so the next time you get stopped by the clerk and told not to do it until they say so, keep all this in mind. About the only absolute is that many terminal systems require that one item be rung up before it’s ready to accept payment info, but that’s for shopper safety. (That is, you can’t enter your payment info and walk away for someone else to buy things on your pending card.)
This is something I could have tolerated in the 1980’s, when all programming languages were procedural, and a sequence like “accept request…accept card…process request…etc.” was the norm. But state languages exist now; interrupts have been around for a long time; and the procedure doesn’t have to be in the same sequence every time. As long as the programmer understands how to do it (apparently some don’t), inserting a card should trigger a “card read & store” sequence anytime, not just when the data is needed.
On another subject, our local tax database has been online for about 10 years. Many years ago, I discovered that the valid character set for the property owner’s name included an apostrophe ('), such as might be used for “O’BRIEN”. There are tax records (several) for that name.
But something happens between the online search entry box and the actual search, and the problem is the non-alpha character, ('). Search for O’BRIEN, and you get a not found error. Search for OBRIEN, same thing. The only way you can find the record for O’BRIEN is to search for the partial match, BRIEN, which gives you the record for O’BRIEN.
I notified the various departments about this error in 2007. Still not fixed. Why not? Because they know the workaround, and if anyone complains, they are told the workaround.
Similarly, one of the features of the tax/mapping online database is the ability to draw a circle around any one property by specifying the radius in feet, and the map will outline all other properties that fall into that circle. Handy if you want to send mailings to nearby property owners.
Works for 1000 ft, 2000 ft, 3000 ft, but try it for 5000 ft and you get a program error. Obviously no programmer ever tried it. And my complaint of 5 years ago hasn’t resulted in any fix; the error is still there.
So if I want data for 5000 ft around a property, I call up the Real Property Listing Department and they email me a complete, CSV database for the entire county – it’s just a few megabytes. For $50. Now I can search for all the O’BRIENs I want to myself.
Interestingly enough, just to satisfy my curiosity, I wrote a program to analyze the complete county database to show what characters were used in owner’s names. Besides ABC…, I found / # $ ’ " in use. I suspect some of these were just entered by fat fingers, but none can be searched for online, probably for the same reason as the apostrophe. Apparently no one thought to test for this before.
That reminds me of something: My wife has a hyphen in her middle name. When booking airline reservations she has to enter her full name. Many sites cannot accept a middle name with a hyphen.
I just came across this one yesterday. I’m listing some camera gear on eBay, and in the description for the lens, I put 105mm@f4, and the listing was rejected because email addresses are no longer allowed in listings!
I encountered an online application (something Federal, like Medicare) that didn’t accept a name without a middle initial. It was a very long form, ending with an electronic signature. I don’t have a middle name, and it wouldn’t even let me go back and make one up.
OTOH, it’s pretty well known in programming circles that trying to verify an email address by format rules is practically impossible, if not indeed logically impossible. “Identify a block of non-whitespace text with an embedded @ sign” is about the only rule that doesn’t create false rejections of good addresses.
My personal domain is blahblahblah.us and my email is myname@blahblahblah.us. When I first got the domain and email address (circa 1999) I found a lot of websites that would insist my email address was invalid since it didn’t end in a 3-letter TLD. Fortunately that particular stupidity has long since subsided.
Try having a last name with an apostrophe (which, for the uninitiated, is a string delimiter in most databases). I sometimes have two middle initials and my last name is the part after the apostrophe, sometimes the first letter of the last name disappears, and on at least one occasion my last name was simply “O” — depends on the application.
What’s probably happening there is that the input is being cleaned to strip out any non alphanumeric characters to prevent hacking attempts. (Think Little Bobby Tables). Yes, there are solutions that don’t require stripping every special character out of an input string, but someone decided it was more cost effective to just use the workaround.
In terms of UI design in general, I’m surprised no one has yet mentioned the curse of keyboard shortcuts, which I believe is a plague originally brought to this earth by Microsoft, but now widely infests lots of non-Microsoft software.
Here are some handy statistics from my experience.
With the single exception of copy (CTRL-C) and paste (CTRL-V), the number of times I have used keyboard shortcuts in any application: 0
The number of times keyboard shortcuts have actually been a convenience: 0
The number of times I have accidentally hit some crazy, obscure unintended shortcut due to a slight finger slip or a tiny bit of errant fingernail: 352,983,217,458 (approximately).
The number of times that hitting an accidental shortcut has caused the application to wildly go into some unintended mode, possibly flashing furiously, going into full-screen and windowed mode repeatedly, popping up extra windows, causing Word to reformat everything at random, and/or lose everything I just typed: 352,983,217,458 (approximately).
The number of applications I have found (so far) in which it’s possible to turn off the cursed shortcuts: 0
Every application has Ctrl-Z (or Command-Z on a Mac) for Undo, but they all seem to have their own shortcut for Redo.
Some use Ctrl-Shift-Z; others use Ctrl-Y; and there may be others.
Why is this such a problem? Because if I try one and it’s the wrong key, then it acts like a regular keystroke and the entire redo buffer is erased.
I wouldn’t know where to start! There are so many different kinds of annoying interfaces. I’ll just mention one:
In a Windows File dialog if your hand jitters when trying to open a file, you can rename the file or move it into a different folder! For example, suppose I want to open a Notepad file in my ‘Important’ folder. If my hand jitters when clicking ‘Important’ I may accidentally make that folder become a subfolder of the ‘Imprudent’ folder displayed immediately below it in the File menu. If I happen not to notice, then I’ll never find my ‘Important’ files again! This is not just a theoretical complaint; I once ‘lost’ the entire /usr directory and had to reinstall cygwin from scratch. (Later I figured out what had happened.)
Many interface complaints are about making frequent simple requests complicated. But the flaw I just described is the opposite: An infrequent request which should be difficult to do accidentally is made too easy. It reminds one of the Unix user whose aliases include ‘alias r rm -rf *’. Don’t let your babies and kitty-cats anywhere near your keyboard!
~ ~ ~ ~ ~ ~ ~
Another minor complaint. There are some contexts where the ‘country’ will almost always be U.S.A. (For example, selecting Country of Citizenship when making an appointment for American Citizenship Services at a U.S. Consulate :smack: ) Why should it be necessary to scroll through Afghanistan, Albania, etc. to select U.S.A.? Note that it would be harmless to have multiple U.S.A.'s in the list, e.g. one at the top, one in alphabetic sequence. What gives? Political correctness? Lack of imagination?
~ ~ ~ ~ ~ ~ ~
While, I can think of good reasons to ask for state in addition to zipcode, do you have a cite for your claim? Do the exceptions arise at the peculiar Tennessee island inside Kentucky, or whatever it is? I see no examples in the popular free online zipcode database.
But speaking of zipcodes, I’m reminded of an annoyance in Thailand. There is one zipcode for each post office, and exactly one post office in most (but not all) districts. I’m in one of the districts that has a 2nd post office and much of my mail used to be addressed (via computer algorithm?) to the wrong zipcode, despite that I’d provided correct zipcode to start with. For a while, I think there was some guy down at the main district post office reading my “undeliverable” newspapers and magazines! :mad:
With a lot of POS devices, you have options options to clear or accept. Some have “accept” on the right; others have it on the left. Would it be so hard to standardize?
They’ve fixed it recently, but for a long time, web meeting software required admin rights to run, long after most organizations stopped allowing their users to run as admins. So whenever anyone had a web meeting, someone from IT had to install it.
@wolfpup: If you watch any professional Windows user their hands almost never leave the keyboard. Mouses and menus are the clunky add-on; keystrokes are the way to do damn near everything. This is less true in a drawing app than in, say, a word processor or programmer’s toolbox. Other than that class of apps though, if you’ve got a hand on a mouse more than about 5% of the time you’re doing it wrong.
@minor7flat5: At least in Windows, Redo is standardized as Shift-Y. Microsoft doesn’t have a method to enforce that apps follow their compatibility guidelines. And if they did people would really howl about monopolistic overreach. But they do publish standards on how it ought to be done to play nice with others. The solution for us non-developers is to not buy apps written by idiots that A) are ignorant of the basics of app development or B) refuse to play nice with others.
Here is the official word from Microsoft to the entire PC development community. Guidelines for Keyboard User Interface Design | Microsoft Learn. Note it was first written in 2002 and is still applicable today despite the somewhat dated graphics. 15 years later IMO there’s not much excuse for non-compliant commercial software.
I think you are an unusual case. I use keyboard shortcuts in almost every application, and so frequently I couldn’t begin to estimate daily usage, it would be in the hundreds-per-day minimum, and I am only a medium user in the grand scheme of things.
In looking closer, I realize that there are really only two competing keys: Ctrl-Shift-Z and Ctrl-Y.
I just looked at a few of the most important apps I use: Toad uses Ctrl-Shift-Z, Photoshop uses Ctrl-Shift-Z, Eclipse uses Ctrl-Y, and UltraEdit uses Ctrl-Y.
I looked up Windows Redo in your doc and it is Ctrl-Y, so it looks like Eclipse and UltraEdit are playing nicely.
But the problem is that big players in the field (Photoshop and Toad are leaders in their respective fields) are not collaborating. And somehow I never seem to remember which ones followed the guidelines and which didn’t.