"based on 71 users. "
"Billions in lost revenue"
Core blimey. Thank goodness for the experts to right our wrong paths, eh, peeps?
Personally, I blame Micro...hangon. It works here. So it isn't funny.
The mania for "flat" user interfaces is costing publishers and ecommerce sites billions in lost revenue. A "flat" design removes the distinction between navigation controls and content. Historically, navigation controls such as buttons were shaded, or given 3D relief, to distinguish them from the application or web page's …
UX testing can be performed on 71 users and give very good correlation.
It depends on the tests, the question/task set etc and the diversity of the pool of users (e.g. you couldn't ask 100 people in a pensioner home to carry out a task and think it applies to the whole population). This is very different from a statistical survey based upon opinion.
UX testing with a pool of just 5 users can provide sufficient result for your own apps to be clear about consensus if you have a clear problem you are trying to address. It may not suffice for a general subjective or opinion based question about your website or app.
Heatmap testing general requires around 40~50 participants.
Yes. But Billions in lost revenue?
Heatmap testing in a pool for a specific UI, fine. But then to extrapolate that across the, well the entire world using public with a claim of billions in lost revenue?
Sorry, I don't buy that. And being attributed to purely the aethetics and not the layout, information flow, and the segment of population etc. Not for something claimed this high.
If someone had said that 'KDE users more productive than Gnome' there would be a massive outcry at, well lots of things.
But hey. This is the new world APP rapid design, 'fail forward' and all those wonderful terms that mean half arsed conclusions are used to make grandiose decisions.
Gah - aethetics = aesthetics.
Anyway, on an 'ire roll' now so lets see where this goes...
If companies, who directly don't lose the billions by inserting a lovely 3d glow around their pop-up 'Hey - subscribe to our newsletter because we ROCK' as soon as you go to their web-app, really were concerned about lost productivity - I wonder how many billions, nay, Trillions, are lost daily because of 'Hey - your call is important to us, please stay on the line so we can bounce you around the globe a few times, in a phone version of a text adventure game!'
'Hey - your call is important to us, please stay on the line so we can bounce you around the globe a few times, in a phone version of a text adventure game!'
Personally I've found those sorts of shenanigans more common on complaints and other 'information lines'
- If there was an actual number for 'I want to give company X some money' you can be damn sure the only bouncing will be minimal to filter out the 'Buzby callers'
>Yes. But Billions in lost revenue?
I am really not sure what your point is. For example, Google/Alphabet reports ~$25B USD revenue per quarter. I'll use that rather than some breathless "size of Web economy" stat. Halving that, to remove non-web Google stuff, then taking the remainder as a rough proxy for the order of magnitude of the web economy*, if there is a 22% loss in navigational efficiency on web pages due to flat design, then yes, revenue on things like ad-clicking or purchasing may very well drop in the billions.
Assuming of course the 22% drop is real and not just something to sell more Nielsen services.
There is a difference between calling BS on some made up marketing numbers and just being cantankerous for no particular reason. The UI-expertise commentards already called you out on your dismissal due to methodology and sample size. I am calling you out for not realizing that "1-2% of very big numbers in the triple digits range of $ billions" may very well equate to "billions in lost revenue".
>This is the new world APP rapid design, 'fail forward' and all those wonderful terms that mean half arsed conclusions are used to make grandiose decisions.
Glibness <> insight.
* Assume Google has 5-10% market share of web revenues: $24B * .5 (a hypothetical share of their web vs non-web revenue) * 4 (quarters) * 10 (the inverse of their market share if at 10%) => $480B/year.
1%, not 22%, of 480B$ is already $4.8B.
Ok - so lets see how these figures transfer:
"if there is a 22% loss in navigational efficiency" - which comes down to time taken to perform various tasks. This doesn't say this results in a task not being completed. Just taking longer. So even if we then pick another figure out of the air -
"then yes, revenue on things like ad-clicking or purchasing may very well drop in the billions." How? It may, but equally you can say it may not. So we are no clearer. Or rather, I don't think we are.
"There is a difference between calling BS on some made up marketing numbers and just being cantankerous for no particular reason. The UI-expertise commentards already called you out on your dismissal due to methodology and sample size."
You think they did, but my comment was based on the leap between this study of 71 users then being extrapolated to a comment referring to billions. Not that you can't get any meaning stats on a UI from 71 users. UI expert I am not, but I *do* know you can design a good UI without and with 3d indicators. So throwing a figure of 22% navigational effiency change based on using 3d or not, I am sorry, but I find it very hard to believe you can then say Flat UI is results in lost billions. This is not a scientific methodology. A flat UI vs a 3d UI - these are VERY large areas. To start talking about Metro and OS design based on this study is a little disingenious. I feel.
"Glibness <> insight." True. Forgive my glibness. However, there is a trend to make sweeping statements based on 'marketing style evidence'. I think this is one of them. Admittedly, this is my thought process. Others will disagree.
Overall what we are talking about, potentially, is just being clear on what is and isn't a link on a web page. What is and isn't clickable on an app. This *can* be done with 3d, but equally can be done in a flat way.
Edit: Something bugging me. You put: "1-2% of very big numbers in the triple digits range of $ billions" may very well equate to "billions in lost revenue", which looks like a quote but I am struggling to see where that quote came from? Trying to work out how 1% came into this.
Edit - missed the edit window but the actual study goes on to say this:
"These findings also confirm that flat or flat-ish designs can work better in certain conditions than others. "
"Notice that those characteristics are also just good, basic UX design best practice: visual simplicity, external consistency, clear visual hierarchy, and contrast. In general, if you have an experienced UX team that cares about user research, you’ll do better with a flat design than other product teams that don’t. If your designs are already strong, any potential weakness introduced by flat design will be mitigated. If you’re conducting regular user research, any mistakes you make in implementing a flat UI will be identified and corrected."
So what we are now showing is that actually, this isn't a Flat vs 3D simple study. Which means that the Reg article we are discussing, is slanted differently to the conclusions actually put forward by the NN Group.
Well, you're not wrong on doubting marketing numbers in general. But given the magnitude of the web economy, even small effects can add up dramatically. So, in this instance, I'd give it somewhat the benefit of the doubt.
Keep in mind: a huge web metric is a user abandoning at a landing page for whatever reason - this is something a pretty, but inefficient, UI could easily worsen even while it made all the deciders salivate. (think no further than Windows' 8 UI for example).
Honestly, I'd be more doubtful of the 22% efficiency loss claim. That's a big number, pretty easy to game and likely to motivate much spending on remedies and white paper purchases. On the other hand Nielsen lives and dies by its reputation for truthfulness.
If you're aware you're being glib, just as I am aware I am BS-ing some back-of-envelope numbers out of thin air, then, yes, we are on the same wavelength.
In terms of the time lost, 22% isn't high. I've just completed some UX testing like this, and found issues where people take 2-4 times longer to complete tasks in areas of the site with bad labeling. That is 200%-400% longer.
Every participant under 60 who I have tested with over the years is short of time, so taking longer to complete a task means they may not have time to do another task, and the risk of an interruption e.g. phone call, increases in proportion to the time taken. An interruption has a high chance of abandonment the task altogether.
Within an organisation, a 22% reduction in employee efficiency while using a website such as the company intranet is very significant.
Ad revenue comes from clicks and views, not how long each ad is visible to the user, so this will reduce advertising revenue.
For any eCommerce website 22% more time consuming will translate in about the same loss of revenue, plus competitors are just a click away, so under performing sites can lose customers to competitors.
This is a sound methodology. 71 users is hugely more than needed to come to a solid conclusion. There is not enough space to go into it all here, but UX Matters tackle the sample/validity question in "Studies for problem discovery" section of http://bit.ly/2wDVkrc
User testing is not Marketing hocum. In my experience Marketing managers can get angry when they receive the results of user testing as it cramps their style, debunks assumptions, and creates a high and measurable bar for design quality where traditionally there was no accountability. If it was easy to manipulate, they would do that rather than get angry.
On the other hand in my experience, devs more often claim that the methodology is invalid due to statistical sample size requirements. But this type of study is comparable to testing software for bugs. You don't need thousands of testers to find the major bugs. More than 3 is a waste of money because the extra bugs they find will be minimal.
If you work in the web industry and think UX techniques are unreliable, you are missing out on an important tool to built quality UI.
God I hate KDE, it is an abomination to gui's everywhere. No icons on the desktop at all, who was the braniac that thought that was a great idea. I tried KDE for an hour and realized it was one of the most non intuitive interfaces I have ever used. Nothing was where I expected it if I could find it at all. Sure they got the bugs out of a lot of the programs that come with it but then the interface design itself became a bug of enormous complexity when gui's are supposed to simplify the OS for the users this monstrosity went the exact opposite direction. I think hiring a focus group of normal users would be the best money they could ever spend.
Which version of KDE are you talking about? Past disappointments with an early version of very unstable KDE 4? Or disappointments with _any_ version of KDE 5 (or whatever they call it)? KDE 5 is one example of a flat-inspired design, with hard to distinguish monochrome icons (extremely stupid), and countless idiotic decisions. For some "traditional" features, you must now use the thing called "Activity", which I never figured out (it's easier to compile your own kernel, really, that says it all). I managed to stay on KDE 4, for the time being.
Confused about KDE? Try GNOME 3! I gave it a try, and binned it relatively quickly. Eventually, the developers of GNOME decided that what makes a desktop UX as we commonly know it must be completely stripped from the gtk library, enforcing people to have a program looking like a phone app. Unsurprisingly, I heard that gtk does not have the following it used to have. There seems to be a convergence toward qt.
Insanity just about everywhere.
A small number (71 ) of users can but often do not give good correlation.
When using a small number, those tend to have similar demographics/attributes.
Who were the testers?
Were they Apple Users, Windows Users, Linux...?
What were there ages?
The article specifies that the finding are not real-life.
"To get comparable, interpretable results from this experiment, we had to ask users to do very focused, short tasks on a single page. In real life, users don’t do tasks that way. They arrive to your site, and don’t know who you are or what you do. They navigate to pages, and don’t know for sure that they’ll find what they’re looking for there. They explore offerings and options."
To call it official, is less than scientific or honest. It is more likely, the writer of the article is not a fan (more likely a hater) of the flat design.
I can tell you that 8 people are enough to test on. NNg were using more than one site. You don't get any more information by doubling to 16. I have been in the eye-tracking and testing labs and I have dealt with the data. Even my marketing colleagues freaked and spent ££££ on focus groups, to get exactly -- EXACTLY -- the same results.
Sorry - this story is presented as Flat UI means lost productivity vs 3d.
I don't buy this. A well presented UI is a well presented UI with or without 3d effects. I am not descrying the fact you can test UI elements with a few users. I am arguing against the flat '3D means more £' rhetoric.
That was exactly my thought - based on the heat map illustration it looks like the same layout was tested with flat UI elements vs 3D elements. However, changing to a good modern design requires more than just changing a theme - it should use the placement and coloring to give the user cues which elements are active and which are passive.
Wrong! There needs to be a clear distinction between content and controls. As more and more apps have been skinned, now the Windows GUI makes content and controls indistinguishable and very web page has its own look'n'feel, the result is that usability conventions have been eroded away completely, and users have to *learn* the vagaries of every app/tool/application/web page. What a mess.
I wish developers would understand I just want to use your f*****g tool, not gaze upon the magnificence of your lovingly crafted individual creative masterpiece.
Windows 3.1 -> 2000 had very clear GUI guidelines and applications were MUCH easier to learn than pre-Windows DOS based apps. Of course MAC/GEM GUIs had their conventions, but the real difference could be perceived when PC users went from pure DOS Text-based applications to Windows with its standards and usability conventions.
When I am using a computer to get something done and get fouled up on a too clever UI the only extra anybody in the world gets from me is a few swear words. Can you put a value on that?
I might add that any sort of an advert comes into the same category. Adverts are part of most user interfaces. Fine if you are shopping but if you are working they are n b u at all.
Slower is NOT better. FASTER is _ALWAYS_ better!
Besides, the article clearly points out that eyeball focus went to things that weren't "productive", like looking longer at titles [maybe to recognize it as a title?] or buttons or other things that are NOT content, apparently to recognize them for what they are. You know, those thermal plots near the center of the page...
So this "extra time on the page" is just inefficiency, and is NOT an indicator that your page is better, or more important, or more interesting. It's just HARDER TO READ when it's 2D FLATSO FLUGLY, vs Elegant 3D Skeuomorphic. (if you ask me, I want EFFICIENT design so I can get more done).
As for the article itself:
THANK YOU THANK YOU THANK YOU THANK YOU THANK YOU!!!
It's ABOUT DAMN TIME that the TRUTH about this 2D vs 3D came out, and I have instinctively HATED the 2D FLATSO since it was excreted from the evil bowels of Redmond's "force the world to change" department. I particularly blame THIS person, though ti's possible that not ALL of it was her fault, directly. Sinofsky was merely a convenient scapegoat, and being male, easier to fire.
Did they really go with the '3D' look for buttons because of usability studies in the 80s? Or was it because they wanted to show off the computing/graphics power they had and make flat buttons look dated and backwards? I'll bet it was the latter. Apple seemed to be the only company that ever gave much credence to usability studies back then - if Microsoft paid attention to them back in the 80s and 90s they need their money back!
Once everyone and everything had 3D buttons by 2000 or so, I guess Microsoft decided to be different and unfortunately for the rest of us Steve Jobs liked what he saw. He always liked simplicity, so I guess its not surprising, but too bad everyone else went that way too.
Maybe it is like hem length, and everything that is old will become new again. The old flat buttons rule today, but maybe soon we'll see the 3D buttons return in a UI here and there as a way to stand apart from the crowd, and it will come back into fashion.
There's a general belief that people see what they eyes point to, but that's far from being true. Work on "inattentional blindness" shows that even seemingly obvious material is overlooked when it is not related to what a person is searching for. Advertisements that trick a user will just annoy them, and make it even more likely they'll simply abandon a site.
Once upon a time UIs were all flat, then UX experts did some research and discovered that introducing 3D elements increased navigability. This must have been sometime around 1988.
Not surprising that modern research agrees, as meatspace hardware changes terribly slowly.
Agreed. What happened here is certain companies (Apple, Microsoft, etc), pushed the "new, modern, 50% less fat!" flat GUIs so they could either (a) run it on cheaper hardware or (b) make it run faster on underpowered hardware. iOS changed to make it run faster. The abomination that was Metro in Windows 8 was so they could run it on phones barely more powerful than digital watches.
The companies made the flat GUIs, then hired some brain-addled UX "experts" to convince us all that it was actually better not being able to figure out GUIs
"UX experts did some research and discovered that introducing 3D elements increased navigability. This must have been sometime around 1988.
Not surprising that modern research agrees, as meatspace hardware changes terribly slowly."
And that's the entire thing neatly wrapped up with a bow.
Proponents of flat design tell us that skeuomorphic design was needed back in the early days of PCs because people didn't know how to use them, and the physical resemblance to actual tangible objects helped them to grasp such a heady topic.
Now, we're told, people have been using graphical computer interfaces for years, and they no longer need the "training wheels" of skeuomorphs.
Human brains have been dealing with 3d objects in the real world for hundreds of thousands of years (assuming 'human' means 'H. sapiens'). Our ancestors also dealt with 3d objects for millions of years before that. Beings that can quickly recognize threats, useful objects, edible objects, etc., in this 3d visual world are at an advantage compared to those who take longer to do the same. Everything about our evolutionary legacy is about creating brains that are adept at identifying objects in that 3d context.
Skeuomorphs are not training wheels for people who don't know that a box containing the word "Ok" means "Ok". It sounds dumb to put it like that, but that's essentially what the flat UIers are trying to tell us. Skeuomporphs are representations of UI elements using the native "design language" our minds have evolved to work well with over thousands or millions of years of living in a 3d world full of 3d objects. Of course we process things faster that way; that's not shocking or puzzling in any way. The shocking thing would be if, despite this evolutionary legacy, we were just as good at dealing with flat UI stuff that we've been evolving to use for four or five years (while continuing to spend all of our non-screen time in the 3d real world).
Can we finally put this whole flat trend to bed and worry about what works rather than what is trendy and new?
@mort - Bad UI design is a known killer. What the study is saying is much of the UI 'research' is not based on trying to understand how users actually interact with the site but on what looks cool to graphic artist. Good UI design should try accommodate both young eyes and geriatric eyes that do not work so well these days. This means fonts, font colors, background colors, visual clues, etc. need to carefully considered. Also throw in color blindness and other lifetime visual issues and many 'cool' UI designs are flatly idiotic because many will have trouble using the page.
Too often in IT there is a tendency to ignore human biology like visual acuity, hands, etc. which limits the range of valid options for an effective interaction.
Too often in IT there is a tendency to ignore human biology like visual acuity, hands, etc. which limits the range of valid options for an effective interaction.
Ahh, like websites that have a perfectly good scrolling system at their disposal, so they can provide a number of posts that users can scroll through with just the cursor keys, or a mouse wheel.
And then they add a useless bit of bling that collapses longer posts, forcing your flow to be interrupted as you navigate to the "expand post" link.
Or, believe it or not, some sites replace the perfectly good post time/date with an XXX mins/days ago mashup that is useless for any time longer than an hour. And even more crazy, if you disable this iavascript bling, instead of falling back to the old date format, it just shows the date. No time.
So if you are reading a new topic, all posts have the same datestamp.
had although many point this out, nothing changes.
El Reg should do an expose on these sites....
Scientists usually need to replicate their studies to be sure of their findings.
But what strikes me about this whole thing is that nobody did any such study, before flat design became popular.
Some people and companies just said it was good to do. And the majority went along, as if this was God's truth. Nobody even bothered to check how this change affected software usability for people.
I think this should be a lesson for the IT industry, not just about flat design, but about computer programming in general. Evidence-based programming is what the IT industry really needs.
Because there are other examples in software development where some popular guy or a group of popular guys say that this or that is good to do. And a lot of people in the industry follow along and treat it as if it's some kind of religion.
It's almost like people trained in Computer Science don't know much about science. Because they aren't using the scientific method and they aren't conducting scientific experiments the way true scientists usually do.
The studies were done in the 80s.
Faux-3D where clickable things look like they stick out of the screen and repond to a click by appearing to become depressed was the consensus opinion for "best UX" throughiut the 90s and 00s.
Several different studies have shown "flat" to be less discoverable and slower to navigate.
It's just that the design people have forgotten history. No doubt this will be "rediscovered" soon.
Forgetting past studies about flat design is another sign that people in Computer Science need to have a better training in science.
Because true scientists never forget to look for evidence, before they make any kind of claims or conclusions. Scientists always need to back up any claims they make either with published studies from the past or with new studies and experiments they themselves have done and published.
Evidence is everything in science. Without evidence, you only have speculation and beliefs, rather than some kind of knowledge. And it's not just any kind of evidence. It has to be evidence based on the scientific method, with proper experiments and controls.
"But what strikes me about this whole thing is that nobody did any such study, before flat design became popular."Actually they did. When the web started it was all flat. When 3D elements became available studies were done. Not just on 3D, but all aspects of web page design. Mostly ignored by so-called web page "designers", but there are examples of excellence. The Google home page and Amazon for instance.
There's an aspect to this that Neilson probably took into account. While a person browsing the web when it was new would invest some time, today with so much choice a page that frustrates through design will send users looking for something easier to use. One thing that pisses me off royally are websites that I must use where I end up going around in circles looking for what I need.
"Actually they did. When the web started it was all flat."
When the web started, links were blue and underlined, and they turned purple once they'd been followed, so the user could easily see where he'd already been. They were not flat or unflat; they were text, but there were consistent clues about what was a link and what was not.
Now, I guess it's not cool enough to have underlined or color-coded links. At least on a traditional PC, we can mouse over the elements and see if something that looks like it might be a link really is. It's really the same thing as the flat design trend; it's "designers" who care only about aesthetics (their own, subjective aesthetics) winning out over everything else.
A designer presented two alternative design concepts to me for a major project 3 years ago. One design was flat. I had my team test both designs with users and rejected the flat design, and built the other.
The reason I didn't shout this from the rooftops was that I had no reason to help competitors make the right choice.
Can't attest to the veracity of the number, but I can tell you that I have never personally met anyone who likes "flat" designs, and as a matter of simple observable fact most people tend to avoid things they don't like. So the question of how long you spend looking at "relevant" (i.e. money making) content versus trying to figure out navigation is actually moot, when you're disinclined to view that page or any other part of that website at all.
Considerably less anecdotal proof of this can be found in the pitiful market share of Windows phones, and the fact that the only way Microsoft could get anyone to use the flat-UI desktop equivalent was by sneaking it onto their machines by stealth. Although having a global monopoly on the desktop pretty much guarantees a captive audience that you can foist any old garbage onto with wild abandon anyway, so its debatable how necessary this subterfuge really was, other than as a means of continuing to spread its tentacles into a PC market that is essentially dead.
And the Sky EPGs are the worst. I suspect they take at least 20 times longer than they need to to find anything. And why on earth will they not provide an alphabetical list so you can find a particular channel quickly.
Money. Channels pay for their positions in the Sky EPG, if you could just search for them by name, then they wouldn't need to.
If the menus on my Freeview box are anything to go by, the time isn't spent trying to find the information on the screen, it;s spent waiting several seconds for the button press on the remote to register. I can only assume the processing power in these things is akin to that of a small legume. Either that, or the programmers are so bad that they are writing massive nested loops into their code just for shits and giggles.
A user friendly EPG would establish the user's channel preference by frequency of use and order by that, or at least put the most frequently used few first and the rest in alpha order.
Oh, wonderful! An EPG whose ordering changes over time because one's watching habits change according to what's being shown on each channel.
That's not going to be confusing or annoying, not at all!
And the Sky EPGs are the worst.
I was fine with the Sky HD EPG but Sky Q is just badly organised and implemented. Several multi-page views (like, for instance 'My Recordings') don't have a scrollbar. The TV guide now requires several button presses to get to it. They've also wasted nearly a third of the screen width with a stupid corporate logo and the mini-TV.
Then there's the amazing context menu that stays open and defaulted to 'delete' after you delete something so that if you press 'Ok' twice another programme gets deleted (there is an undelete option).
It does what I pay it to (yeah..I know...) but the UI pisses me off pretty much all the time I see it.
You don't really want an alphabetical list.
The king of shit UIs, Samsung TVs, tells you it can order the channels properly if you give it your postcode for your area (more likely, they just want your postcode). So you input your postcode but it doesn't work because you forgot to input the country on another screen (or more likely you couldn't find the option or didn't even know it existed in the first place).
So now all the channels are in alphabetical order, and all the adult channels at the start because they begin with "ADULT".
The apps on Smart TVs like Samsung where to enter the PIN code you have you use the cursor keys and press OK.
1) There is a keypad on the remote for typing numbers
2) You have to send the kids out of the room and make sure they aren't peeking through the gaps as although the code has **** when you enter it the fact that you hover over each number and select it makes it trivial to know the code.
Or not have say Dave and Dave + 1 a hundred channels apart.
I wonder what Channel 4 have done to piss Sky off - C4 HD is way down in the list. Even further down than ITV HD. But I don't pay much attention to channels these days. Just browse the EPG looking for something interesting then mark it to record. Eventually I'll bother to watch it.
Well there is more than just the number of participants that is relevant. The study could have some serious flaws...
...however the results of the study are exactly what you'd expect from a GUI which removes important visual clues to how it works. Just imagine having a room with an invisible touch sensitive sensor instead of a clearly visible light switch. You probably still touch the right space on the wall many times, but if you don't, it's hard to see where you should have touched it.
A good example is the car park entry system where I work. It has a large 20cm square black panel that you need to hold a tag up to in a particular place. If you hold the tag in the wrong place the barriers won't open. If they were to draw a square on the pad so you knew where to place your tag you would get in a lot quicker (not that that is a good thing).
Black text on black buttons on a black panel. This stupidity started long ago for DNA to notice!
I have TV with dark grey over simple touch icons on black.
I have LOADS of LCD screens with a row of unlabelled buttons almost invisible under the front bezel.A nightmare to select inputs or change settings.
And in Arizona, USA, it's Cold until the hot water gets there and HOLY MONKEY BALLS HOT HOT HOT HOT AHHH MY HANDS until it finally cools down to cold after ten minutes. (at least in most of the houses I've been in- apparently they never heard of putting insulation on the pipes to keep things hot or cold.)
At least in the summer. In the winter, it's actually normal.
Even "simple" can be confusing. Last week I encountered a Petrol pump with A,B, C options (let's call them Unleaded, Plus and Premium). They only had A (Unleaded), so they put a paper sign on the panel, above the two other buttons saying "Unleaded Only".
What happens? People only try to press B or C (Plus and Premium), thinking there's no Unleaded,
I pointed out to the attendant that they should put their paper sign OVER the two buttons to physically hide them. I get some response about how they did that, but the rain messed up the ink. BS, it hasn't rained for five days at that point. [Yes, this is the great 2017 Texas Not-A-Gasoline-Shortage]
"imagine having a room with an invisible touch sensitive sensor instead of a clearly visible light switch. "
Come over to my Kickstarter page for my ermm Invisible Cloud IoT digital IO switch for stuff.
£500, 000 required and I'll send you a photoshop mock up of the invisible device and update you with empty promises until say 6 months time? Then errm, Chinese, quality control and errm someone else was to blame, money spent on nice holiday and pissing it up the wall, sorry, I mean R&D.
"Just imagine having a room with an invisible touch sensitive sensor instead of a clearly visible light switch."
I know EXACTLY what you mean.
I'm currently working on a project that has 'an interface' that requires users to do something specific with it, in order to control the device (while simultaneously rejecting 'accidental' activation of the user interface). However, the amount of feedback that could previously be given to the users [to help them 'get it right'] was inadequate. I recently fixed this, after making some architectural changes to support it (priorities, priorities), so that the feedback was a LOT more intuitive (and gives more information to the user to help him use the interface). NOW it's much easier to use, and hopefully will be acceptable to the end-users. That's the goal (usability).
So yeah, it's extremely important for the UI to be as intuitive as possible. Otherwise people will have a hard time using it, and are likely to complain or "just not use it".
And, naturally, any influence I have over web design, icons, etc. WILL contain more 3D skeuomorphic and veer away from the 2D FLATSO and 'hamburger' menu. And I'm passing a link to the article along to "the powers that be" to help justify it.
Just imagine having a room with an invisible touch sensitive sensor instead of a clearly visible light switch.
Sounds like modern monitors. If you're lucky there will be some symbols at the bottom that indicate that if you fondle something, somewhere in the area then something might happen. If you're not lucky then you get to thoroughly molest the poor thing. I must say though, as an IT person, I do get a perverted thrill out of asking "how do you turn this on?"
"I must say though, as an IT person, I do get a perverted thrill out of asking "how do you turn this on?"
Same here, although the users are more likely to point out that that's why they called me in the first place. They never switch the screen off once it's on unless by accident, and that's when they realise they don't know how to switch it on/off because the buttons/touch sensors are too well hidden :-)
"invisible touch sensitive sensor instead of a clearly visible light switch. You probably still touch the right space on the wall many times, but if you don't, it's hard to see where you should have touched it."
umm... except for the dirty marks on the wallpaper where everybody thought the switch was ??
Sample size requirements are hard to "intuit"
For instance, if 14 people out of a randomly selected sample of 70 are X, and the remaining 80% non-X, you've already got a 95% confidence that the true population frequency of X people is between 12% and 28%, however big the population.
This is why you can get reliable polls even if you ask fewer than 1 in 100,000 people. The randomness of the sample is vastly more important than its size.
"This is why you can get reliable polls even if you ask fewer than 1 in 100,000 people."
Ahem:- polls *can't* account for people like me who refuse to take part in polls, they can only assume a pattern of behaviour and/or beliefs for people who refuse to participate.
I *am* the error bar.
*EDIT*:- There is no way to correlate the projection of the poll against what I believe or do. How are you going to check, take another poll?
"There is no way to correlate the projection of the poll against what I believe or do. How are you going to check, take another poll?" We can do an autopsy on your brain to see what you believed. It probably won't work, but then it sure will save us some hassle with the next poll.
I like to answer polls. I give answers that will skew the results & make anyone reading my answers scratch their head & wonder WTF.
Question: What's your favorite sandwich? Answer: Spotted owl, Dodo liver, & Narwhal pate'!
Let's see you corrolate THAT one with your desired outcome! HA!
Why yes I *am* a cranky old curmudgeon, thank you! =-D
"I like to answer polls. I give answers that will skew the results & make anyone reading my answers scratch their head & wonder WTF."The Gitling purchased a game for me that I'm having a problem with. In order to get support, I was asked "What's your favourite animal? We really want to know." So I responded: "Alan Price".
"Ahem:- polls *can't* account for people like me who refuse to take part in polls, they can only assume a pattern of behaviour and/or beliefs for people who refuse to participate."Oh dear, you don't really think that refusing to participate in a poll doesn't generate information for the poll do you? Yes, you are counted as a non-participant. Yes, you are the error bar. The percentage of refusals is very useful information indeed. Well-designed polls have a very low refusal rate.
It wasn't a survey and the 71 people isn't a sample.
When I run surveys I select a sufficient sample size to get the desired confidence interval and margin of error. When I run user testing I select the number of participants based on a logarithmic curve. There is good literature on the web about this.
An analogy is if a car company builds a new model car and they have some customers test drive it. The first 5 customers all claim that it has bad handling. They test it with another 10 customers who all say the same thing. How many more customers should they test it with to validate this? Do they need to test it with a statistically significant sample of the population of the country, the world?
In reality the company would start to investigate the issue at that point. They would probably retest with some more customers after making improvements, rather than continue to test after the first 15.
It is a rubbish idea that just looks like a laser printed page. Fine for content and totally rubbish for menus, buttons, navigation or interaction. NNG have done other tests and been saying this for ages.
I've worked on GUI design since 1980s and it's definitely purely a presentation aesthetic, very poor for interaction. I still find Windows 10 "control panels" / "Forms" totally baffling in terms of what is informational and what is interactive/clickable.
Even Linux Mint with Mate desktop has some "flat" applications creeping in from Gnome3/Unity/Ubuntu.
People copying things because they LOOK clean is a curse, as is excessive glass/smoke/blur/shadow and Skeuomorphic. Simple clear obvious icons, text on a button (using language resource file if the function isn't super obvious. Very simple 3D effect of two bright lines and two dark lines.
People imitating the worst "koolaid" from MS, Apple or Wordpress themes is mental.
but I do know I hate flat interfaces.
I've often given up on websites when I can't easily spot the scroll bar to go down to the buy option, or because I can't even find the button I need.
It isn't just the billions lost in terms of wasted time, it's the billions of lost due to customers sodding off to someone else's website which is easier to use.
Remember that the billions lost aren't a "total" as such.
Eg I go to site X to buy a £10 widget, give up, try site Y, give up and finally buy it from site Z which has a UI I understood/could find said widget.
Site X lost a £10 sale and Site Y lost a £10 sale, so a £20 "loss" but I still spent my tenner and got my widget.
Of course from time to time when using Amazon I think "why don't Amazon make their website look a bit more modern", forgetting that a) I'd hate it and b) at Amazon everything is tested to within an inch of its life so it must be done on purpose because it means more sales.
I'm not surprised that flat user interfaces are slower to navigate than interfaces with more obvious visual cues. That's a no-brainer. What does surprise me is that "creatives" just follow the trend like zombies when its obvious its a step backwards. Microsoft makes a UI style change for the worse and the industry mindlessly follows like sheep.
"Back in my day, we had to walk to school barefoot in a blizzard uphill both ways..."
But dad, surely there wasn't a blizzard every day?
"Well, son, if there was no blizzard we didn't get the day off to go to school and had to stay and and work the farm."
What does surprise me is that "creatives" just follow the trend like zombies when its obvious its a step backwards.
> You don't spend much time with "creatives", do you? Mindless trend-following - it's what they do. Well, 99% of them, anyway, to be fair.
yeah, like "academic arrogance" at its WORST.
I bet that MOST truly creative people spend time CREATING, and it's their "creative" BOSSES that tell them to "do it that way". At least, that's how _I_ see it. (In effect, they're rewarded for 'bad behavior' and therefore are encouraged to propagate it).
More like too many companies aren't willing to allow any kind of feedback from users of their products. I still shudder recalling efforts to give feedback to a couple of companies, both of whom assumed that absolutely everything any customer might complain about could be adequately answered by a FAQ, but theyd make you jump through multiple hoops and several web pages before it became apparent that their 'contact us' section didn't really hold any contact details at all, unless you fancied making an international phonecall to spend upwards of half an hour discussing with someone what you found problematic about the product in question and why - rather than being allowed teh ease of sending an email or completing a text box in a web form.
The surprise is no one admitted finding this earlier.
Microsoft MUST have done something that showed the same effect in the new Windows UI, they test a lot and would inevitably have caught this amount of difference. And went ahead with it anyway.
Still, nice to know what I and others were pointing out 2+ years ago has finally been confirmed. Who needs efficient if it can be pretty... oh, it's not even pretty.
You've not used Visual Studio have you?
Don't get me started on that. For a premiere development tool it is disturbingly badly thought out and not particularly well implemented. After several iterations the controls in the properties pages still don't resize/flow properly. The Quick watch still doesn't choose sensible column widths or remember what I last dragged them to. Context menus need vertical scrollbars even before you've added a few essentials.
And it takes so long just to open. And opening a solution takes even longer. During both of those operations it will sort of, sometimes, appear to respond to commands. Background processing? The VS development team have heard of it.
I'm grateful every day that the VS team added so much chrome back into the VS UI after the first Win8/10 style builds presented acres of empty white space devoid of clues. It's always going to be a pig but just occasionally a little lippy does work wonders ;)
Nothing will ever make it handle Unreal4 source in realtime though :(
Who needs efficient if it can be pretty.
There are a strong following that believes if they find it aestheitically pleasing it must be more efficient.
Maybe we're all missing something, and we're supposed to be staring past the visual cues at all the white space in order to achieve some sort of zen enlightenment...
...but then again, I've yet to see any signs of higher thought in any flat ui fans whatsoever...that only leaves some levelling bacchanal.
The unstated idea behind the flat design is to make it easier for users to click on the wrong thing ... and the wrong thing is usually an advert thus generating cash ... flat designs in any application that is web driven generates more "click through" than the older designs where it was easier to see just what you were doing.
And AC above, I think you've both completely missed the point. The unintentional clicking is a happy accident (from site owners' PoV) Never attribute to malice what can easily be explained by stupidity...
A better experiment to back up this somewhat weak study would certainly help, but it's only confirming what intelligent power users have been saying for years isn't it.
..the conversation went something like:
"All the kool kidz are doing flat, we must do it too"
"Well, if it's good enough for he kool kidz..."
..some time passes....
"Oh look our bounce rate seems smaller and look how much longer people stay on page!"
"Yeah, we are totally engaging traffic with this new style - we must be geniuses."
"Maybe now we can get some of those Google-style bean-bags for the office"
I've been talking about that ever since Windows 8 was released.
Design created for the sake of extreme simplicity without also being functional leads to disastrous results.
More on that here: https://tech.slashdot.org/story/17/01/27/1425205/ask-slashdot-a-point-of-contention---modern-user-interfaces
I know what you mean about Windows 8 - but I think it was more of a convergence thing than a real design decision about an effective UI for computer users. IMHO the rot set in when someone decided that what works best on a low-res 4-inch phone screen would also be the right UI styling for desktops, laptops, TVs, and every other human-computer interface under the sun. Because, you know, smartphones are really popular, and at the time Microsoft desperately wanted to be Apple (as opposed to now, when they desperately want to be Google). So it made perfect sense to remove all the visual cues which enabled people to use a point-and-click interface intuitively.
On the same basis, I look forward to bicycles with steering wheels, trains with off-road capabilities, and tinned horse-food.
If the new would-be trend setters had even bothered to read old facts they'd have ignored them because, you know: OLD!
(mega-thumbs up for that)
You are SO right! The millennial generation decided "it's OUR turn, now" and, like a copy of Arthur C. Clarke's "Superiority", they went on to re-invent *EVERYTHING* and "do it THEIR way".
with predictable results.
I went totally blind long before Win8.x & have never seen the modern UI, nor whatever Win10 has done. Keeping that in mind, I'm seriously wondering what is meant by a "flat" UI.
Are you talking a nested tree of +'s & -'s to expand or collapse a set of navigation sub entries? Just plain text on plain buttons with no shading, iconography, nor embellishments? No 3D, background images, wallpaper, animated GIF's, or stuff like that?
When I hear of a flat UI it makes me think of a simple list of links to same page content, other page content on the same site, or links to off site content. Put them in a Navigation Menu block at the top of the page, a Back to top link at the bottom, & you've just made an intuitive UI. Don't hide any of it, don't make it visually pretty by hiding links/content behind blobs of scripting, just give the potential customer what they're looking for so they can find it quickly, easily, & most importantly, in an Accessible fashion that doesn't cause a screen reader to shit itself.
I know you sighted folks like the pretty stuff, I used to myself, but having turned blind I'm now all too aware of how badly some of that visual only pretty stuff turns into headaches for anyone not equipped with perfect vision.
So what's meant by a "flat" UI? Anyone care to try & describe it?
*Leaves a pint as a thank you gift for anyone whom does so*
The simplest example is rather than having a button which is 3d shaded to make it clear and distinctive that it is something special you may click on, a 'flat' design may just have a box surrounding the text which may (if you are lucky) change colour if you hover the mouse over it to show you can interact with it (particularly good when you are using touch screen style of interface). Of course other objects may also have boxes round them, or they may decide the box can be removed just leaving the text that changes colour when you realise you might be able to click over it. Similarly a flat webpage may choose to style a hyperlink so that it isn't underlined, isn't in another colour, if you are lucky it may be bold/italic but possibly only if you hover over it. Or a collapsed tree may not show any sign that it is collapsed, other than maybe the text being bold until you click on it and realise you can expand it.
Basically flat is removing any visual indicators to make it look 'clean'
Each test image shows pictures of three anniversary wedding gifts: one for the first year's anniversary, one for the second, and one for the third. They're laid out like a glossy magazine advert. Each gift has a title, under which is a paragraph of descriptive text, and beneath that is a click-through link. In the low signifier version, the click-through link is not underlined and is the same colour and font size as the paragraph of text above it. In the high-signifier version, the links are a underlined and in a brighter blue. That's the scale of difference we're talking about. We're not talking about putting 3D buttons around the hyperlinks or the kind of design you're outlining.
You could pull apart the test in multiple ways. For example, the heat map shows the user looking at the title the most. So if the title had been hyperlinked, all would have been well. It was insisting users use a piddly hyperlink under the page was caused the problems.
But the design is more suspect than that. Because a shopper is not going to go, "You know what, I won't buy the second year's anniversary present this year, I'll buy the third." So it didn't need complicated exposition. It just needed to get the shopper to answer the question, "How many years have you been married?"
As for Windows, you'll have to ask the haters. I barely noticed the difference between Windows 7 and Windows 10. But I don't user Microsoft's dev tools or Office, so I'm not deeply exposed.
"It just needed to get the shopper to answer the question, "How many years have you been married?""
What I did notice is the website saying 1 year anniversary = Gold, 2 year = Garnet, 3 year = Pearl. My Oh, My, they really want to condition shoppers into buying expensive and unnecessary bling, aren't they?
I thought it was pretty common knowledge that Golden anniversary is the 50th and Diamond is 60th. Pearl is 30th, and I don't think Garnet anniversary even exists:
Of course it's not so sexy (not to mention lucrative) to be promoting presents made of paper, cotton, leather, wood etc, not to mention with modern divorce rates the amount of people making it even to silver anniversary must be fast diminishing.
I guess they are trying to recreate the success de Beers had with associating diamonds with engagement rings in the 1940s:
Icons, and widgets were actually made to look like real things. The drawing style used shading , shadows and different colours to make things look like the genuine object.
The "save" button, for example looked like an actual 3.5 inch floppy disc. A "new email" button looked like an actual envelope. It actually bulged.
Now, they tend to be simple line drawings. If we are lucky, they may be coloured in in the style of what a 5 year old can do. A trend has been to remove all colour from the oversized toolbar.
If you are able, please think on the toolbar from Microsoft Word or Excel. The current one has much fewer items in it because they are much larger. Beside some of them are tiny triangles that lead to more.
You may remember that buttons used to have dark lines on two adjacent sides. This was showing the shadows that would have occurred if there had been light coming from the opposite corner over a genuinely raised object. Now they are either gray/coloured rectangles or a line drawing without the coloured.
I said a long time ago that these things are not only harder to use but they look ugly and amateurish but, apparently because I was born in 1960, I have old fashioned visual preferences and habits so my opinions were not needed!
Hi Shadow Systems, your second hypothesis is correct, the flat UI style is similar to a piece of paper with all the elements drawn on the same plane with no 3D, shading or embellishments.
The design typically has all the text, graphics and the UI elements within flat rectangles on the flat 'paper', and the icons as just more text within another rectangle. Kind of like a piece of paper with various rectangular post-it notes of different sizes arranged on the paper containing the text, images, and navigation elements, but the post-it notes have been ironed perfectly flat onto the paper so there is no visual cue to any depth or priority. Opening another page typically lays one flat element on top of another flat element, and the scroll and navigation bars are also flat so it becomes more of a challenge to navigate around the user interface. In addition to this, the icon design has also become flat in the same way, and the colour palette has been reduced to mostly primary colours to un-necessarily simplify the user interface.
I hope I have described it adequately, someone else may have another description.
I am in the process of going blind, so I feel qualified to answer at least part of the question.
There are two trends in web design that is making my life very hard.
1. The pastel and low contrast movement. People are moving away from high contrast in web and other UI design and using muted colours with very little contrast. Flat UIs are particularly guilty here. Often Buttons, or other actionable parts like links etc, are not obviously high contrast like they used to be, but looks like all other text on the screen. (Sometimes the lines surrounding button is so thin and low contrast compared to the rest of the colour scheme that I just cannot see it)
2. Flat controls - buttons dont have hard shaded areas, and are sometimes just thin lines surrounding some text. So, very subtle, and the screen looks more like something printed on paper, than an obvious computer UI. Again, for someone with low vision like me, it is sometimes impossible, without moving my mouse over every part of the screen until the pointer changes to know if I have hit a UI control.
Not sure this helps answering your question, but I least I got to rant a little. :)
Wow, sorry for the delayed response! Havent been on here for a while.
So I run Windows 7. Windows High Contrast is pretty useless because web pages dont respect it. So what I have found though that you can defeat pastels by running your screen in inverse colour mode!
So I run Windows Magnifier, set the zoom to 100% (no zoom), set it to full screen, and set colour inverse to true. I pin it to my taskbar and then click on it when I need to switch colours to see if any given UI works better in its inverse colour mode. So far it has been reasonably useful.
Except for my colour palette. May just be typical male.
I cannot tell the difference between pastel shades they all look the same to me, light blue, light green ect.
So when I see light grey and lighter grey I cannot actually see what it there (Google maps is a prime example).
So as far as I am concerned there are about 15 or so usable colours, and that a lot of the pastel shades would be the same.
But I have no problems with photos and the like.
One gem of the flat style is the so-called hamburger menu. On a reasonable design if you have a link which expands to give a choice it will at least be labelled "Menu" and might even be in the form of a button. Or it might even be a series of first level items on a menu bar - much like our beloved "DATA CENTRE SOFTWARE" etc menu across the top of elReg and each then dropping down to give more choices. With a hamburger menu you get 3 short lines, supposedly a schematised drop-down list but looking equally like a schematised hamburger. This will be lurking in an odd corner somewhere and quite possibly looking much like some of the other irrelevant bits of graphic design on the page.
<cite>Oh, right! You mean that "Strictly equivalent" ≣ icon. Yeah. Hate those. And "Cogs".</cite>
The gear icon I can at least grok as meaning 'settings' or some such. As a menu? no. My main hatred for the hamburger menu is that it's not a universal thing- on some UIs, it opens a context menu, on others it opens what used to be a sidebar menu (or left hand column of items), and yet on others it's something completely different.
Chrome, for example, doesn't even have a 'normal' hamburger button, but three dots stacked in a vertical line- that's the menu for you.
And don't get me started with Office 2013 and 2016, or I'll be here all night.
Office 365. Where the page for downloading a standalone version of the software is accessed by clicking on the words office 365 in the top horizontal bar just next to the cluster of geometric spots that looks like something you should be grabbing a screen wipe to tackle but which is actually a drop down menu to switch applications... and, of course, a cog.
I'm seriously wondering what is meant by a "flat" UI.
simple explanation: all 3D effects are eliminated. The buttons in the upper right corner have degraded into a plain underscore, box and X with no effects around them to indicate they're buttons. And they're farther apart so that fat fingers can mash them more easily on a touch screen.
Additionally, the borders around various 'boxes' are gone, including THE WINDOW ITSELF. If there is a border, it's a 1-pixel-wide line, barely distinguishable from the content surrounding it. White-on-white is NOT uncommon. Usually the UI item has a different color, but doesn't always. Sometimes a button has no border or color, and is merely text placed "wherever".
So in short, the old visual cues of a 3D-looking border, combined with a standard color scheme, are gone. They were typically replaced with BRIGHT WHITE [which is hard on the eyes; they need to use #FFFFE8 instead], with sometimes grey or other "hard to see" color text on the UI elements, and NO BORDERS [or in a limited case, a light grey border on white, which is also difficult to see].
I'm not bilnd, but I'm old. my eyes don't easily distinguish certain colors with similar luminocity, at least not without causing them to hurt a lot. They look like "blur" to me. And a lot of the "bright bluish on white" combinations that Micro-shaft seems to like so much are EQUALLY HARD TO SEE. And this is what you get a LOT of in Win-10-nic. It hurts 'old eyes'. [ultimately it will cause eye strain on YOUNG eyes, so those young whipper-snappers better get a clue, or get glasses at an earlier age].
Shadow Systems, imagine some trendy designer thought it would be "cleaner" or "cheaper" to just draw Braille dots rather than make them 3D. A "flat" UI isn't even in the same ballpark, but it's annoying and stupid regression. The ongoing increase in processing power makes it unnecessary.
Of course, on-screen 3D isn't really 3D but it does make buttons stand out better visually. I guess a temptation then is to make the button smaller and hence harder for people with shaky hands and weak vision to hit :-(
What "The Mole" said...
Or, in other words, lacking clues as to what is content (text), what is interactive interface (buttons, toolbars, etc), and what is informative interface (window or group-box titles, etc).
In old MS Windows, menus were distinguished by being in a separate colour and place in a row along the top; toolbars and buttons looked 3-D by lighting and shading; grouped elements could have a line around them with a group name as an interruption of the line itself, and windows had a clearly distinguishable title; the rest was actual content you worked on, and was usually "flat-looking" -- like, say, text in MS Word or Notepad.
From about Windows 8, the 3-D effects were apparently declared old and un-hip, so now everything looks the same: It has become much harder to find what just tells you something, what does something, and what you can edit.
As an example, my wife uses Windows 10 with the new Web browser, Edge. When she wants to show me something or, worse, wants my help with something, I usually can't see where the browser ends and the Web page begins. I have to find my way around by trial and error, clicking on stuff to see if anything happens.
I see that Jakob Nielsen is still fighting his personal war against good design. Wonder what happened to make him hate with such a passion that he's devoted his entire career to making the world ugly. Did he get rejected from art school? Whenever I see his name I know there's no point reading any further; the conclusion is always the same: an undifferentiated mass of text is always better than any attemp at layout, and the best UI looks like Windows 3.1. Don't believe a word of it. Good design is invisible - not conspicuously absent.
If you consider design to be "functional design", then in this case good aesthetics != good design.
Something my be aesthetically very pleasing, but functionally useless.
The ideal is to have something that is both aesthetically pleasing and functionally effective.
However, when it comes to something that is meant to deliver concrete information (to buy something click here, to get details on this legislation click here, to get the specs of the widget I am looking at incorporating into my industrial process, click here, to update my payment details click here, to fill in and submit my tax return click here, to fill and and submit my visa application click here) then aesthetics must give way to functionality. That doesn't mean it has to be ugly, un-artistic, but those are secondary to functionality.
Ideally a UI would both look good AND be useable, but if one of those two has to suffer then let it be looks, because whilst users can learn to ignore something that looks a bit iffy if it's supremely easy to use, they'll almost certainly never learn to love something which looks gorgeous but constantly hinders them from doing whatever it is they're trying to do.
You're right, good design is invisible, the user should never need to think about it. Flat UI design however seems to fall more onto the conspicuously absent side of the fence - taking away pretty much every bit of visual guidance to show users where to click, drag, type etc. and leaving them to guess at which parts of the UI do what really isn't good *UI* design, even though a static snapshot of the UI might look like really good graphic design.
Only 11 downvotes? You can do better than that!
Meanwhile, here's what Mr Nielsen thinks a well designed website should look like: https://web.archive.org/web/20120123103227/http://www.useit.com/ Oooh, lovely text only interface. Nice one Jakob, did you make that all by yourself :D
"Oooh, lovely text only interface."
It might not be pretty but you don't have to hunt for the functions and that's a win.
I can imagine the flat interface carried over into a self-driving car. Suddenly the car announces you have to take control. You look around and can't find the steering wheel or brake and the accident is fast approaching...
"Oooh, lovely text only interface."
Yes, it is quite lovely actually. Clean, simple, no nonsense, allowing the visitor to find the information they want with the minimum of effort.
And why go delving into the web archives for this particular iteration of his site anyway? If you want to criticise his current views on UI design, it feels a bit like clutching at straws to use a 5 year old version of his site instead of the current verson, the UI for which looks nothing like the old one...
"I see that Jakob Nielsen is still fighting his personal war against good design."
And three cheers to him for that. Look at all of the comments here... on a site frequented by people who spend a lot of time looking at screens, yours is the first one in several pages of comments to take issue with the article, or to mention "design" and "ugly" and "art" without a word about usability, intuitiveness, workflow, or anything else that actually matters more than aesthetics to people who actually want to get things done and not just admire whatever it is that is supposed to be a tool to accomplish a task. It speaks volumes that all you can say about any of the objections to flat design is that it's "ugly."
If I may be so bold as to speak for the other commenters: We're all part of this "war against good design." It seems like we're starting to get some traction here... maybe we can win this war and vanquish "good design" once and for all.
Here we are forgetting that also Google was a big promoter of flat monochromatic interfaces for its web applications, and laid out the direction for web sites far more than MS did with its Metro interface.
On small screens, a flatter UI can have some advantages because it uses less screen space (less relevant today because of high res high dpi screens). As long as controls and non-controls are clearly identifiable.
On large PC screens and more complex applications, it becomes quickly less usable especially when mingled with non-controls and becoming ambiguous.
You don't really need a GPU to run an UI, Windows had non-flat interfaces far before hardware acceleration became common for UIs.
"On small screens, a flatter UI can have some advantages because it uses less screen space"
Could have. But my bank with its flat design uses so much white space on its web site that I can't use it without maximising the browser window and they've even taken to adding text hints about where to click as a substitute for a control that might be out of sight. Lunatics and asylums.
" On small screens, a flatter UI can have some advantages because it uses less screen space (less relevant today because of high res high dpi screens). "
Actually, on small screens, 3D-effect controls are even more useful, because most small screens are now touchscreens.
One of the less-discussed benefits of the 3D button is the matter of target areas. To the user, the click target is the "top" of the button, or the inside of the check-box or radio-button. To the computer, the whole area is a valid target area.
This means that the whole area containing 3d effects becomes part of the margin for error in user clicks, and allows you to put buttons closer together without compromising the usability of the interface. This density of buttons is often visually more appealling than using blank space as padding between buttons.
When you're interacting with your finger rather than a mouse, you need a much wider margin of error, and in flat interfaces this often leads to sparseness of elements, and only the centre of the active area is marked -- the boundaries are absent. This makes it significantly easier to accidentally go over the boundary and hit the next interactive element.
For example -- the YouTube interface. Even on a medium-small device like the iPad Mini, it's not uncommon for me to accidentally press the timeline and rewind a video instead of playing/pausing it. This is pretty frustrating.
The 3D interface comes from HP Windows from the early 80s, they were doing this on machines with a 68000 CPU. It was then incorporated in Windows 3 which was supposed to run on i386 boxes. The 3D stuff wasn't the problem as I recall, the problem for the 386 boxes was the font rendering speeds. Surely the WindowsPhone CPUs could scrape up the performance of a 386?
However they could EASILY even in 2002 drive a decent Windows 3.1 / NT / Win98 /Windows 2000 interface. Aero is a mad trip to the other extreme. Best disabled.
I had 9100 and i9200 "smart phones".
Even a 320 x 240 four shade mono LCD with a Z80 can do a reasonable "3D" hinted effect for buttons, icons, scroll bars etc.
Not to mention the absurd contrast theft which plagues "modern" websites. The Register is in the minority with its crisp black-and-white text. Many sites are now only readable with constant eye strain. Or they were until a few months ago. Those who brought us semi-visible text are now in full reverse-ferret, but my accountants still server up limp whiteish on whiter text, and Google search is still pretty gutless.
Actually, I find black on white headache inducing. My text editor is light grey text on a blue field. If I have to read printed black text, I prefer the matt beige of a musty paperback to the glossy white of a text book. I seem to recall, back in the day, there was a lot of research to support this. (Although both are high contrast by modern standards.)
There's also the practical issue that white backgrounds use more power than black, which matters for mobile.
minimal eyestrain comes from black text on #FFFFE8 background.
This is based on its similarity to PAPER, and the fact that blue light depletes the orange pigment in the macula, resulting in macular degeneration. It's why you should never use 'cool white' or 'sunlight' lamps in your workspace, but only "soft white" [which is like an incandescent light].
now, if El Reg could fix the font size in the edit box for comments [it's TOO FRICKING SMALL] I'd make fewer mistakes and wouldn't have to re-edit everything I type all of the time... but if I hit the alt+ to zoom, then the document text is way to freaking big. cannot win...
... the absurd contrast theft which plagues "modern" websites.
The designers of those web pages seem to have learnt at the school of theatre programme designers, who have been delighting us for ages by printing the synopsis, cast list, etc., in (say) pale blue on top of a sepia half-tone photograph. Just what you need for good legibility in the low lighting in the auditorium!
> Historically, navigation controls such as buttons were shaded,
Now I feel old...
Historically, the first UI:s as in the original Machintosh and Windows versions were flat. Although this was mainly because of technical limitations - cannot make very nice shading with low-resolution 2-colour or 8-colour (all of them garish) displays.
You don't need shading or much extra space just a 1 pixel border with top & left pale and right and bottom dark. The windows 3.1 3D enhancement was fine on 16 colour EGA. Though some CGA displays might only have done 8 shades of the 16 colour CGA text mode. CGA graphics was only 4 colours or black and white (2 bit or 1 bit). Windows 1 to 3.1 only used the 1bit mono mode of CGA.
The next lowest depth wasn't 8, but black and white Hercules and CGA monochrome modes. Then the paler lines were dotted. However CGA and Hercules Windows was a curiosity, The minimum practical windows was 640 x 350 16 colour EGA.
Seems I misremembered the number of colours on CGA. Maybe because I never personally owned one. I had a Hercules clone on my first PC (a Bondwell PC/XT clone), and did sometimes use Windows 2 on it. Hercules graphics was 720x350 pixels, way better than CGA and even bit more than EGA horizontally, but of course allowed only black and the slow, green phosphorus of the MDA monitor.
While the flatness is a problem there are several things to like in Microsoft's signpost inspired design, which is one of the reasons why it became so influential and was quickly able to displace Apple's increasingly wearisome skewomorphic designs.
Users love clear navigational clues. This has been best summed up as "obvious always wins" by Luke Wroblewski (discussing the infamous "Hamburger" menu icon). But other respected researches such as Steve Krug have been saying this for years.
Personally, I think that Google has done a great job with Material Design in paring the chrome back but still keeping visible cues. But a UI framework is only ever a start. For any particular website there's still plenty for a designer to do.
I think the main thing here is the conflation of flat vs 3d with abstract vs skeuomorphic.
The 3D interface was never completely skeumorphic to start off with. Sure, the buttons were skeuomorphic, because buttons in the real world stick up and then move down when you press them; but I have never had a desk where the paper was below the desk surface, which is where the 3D in Windows 95 et al placed it. Worse, look at the scroll controls in Windows -- you click "up" to move the paper down.
So the 3D interface and its success can't have been down to mimicking the real world, because it didn't.
Thus the 3D interface is not truly skeuomorphic.
The problem with 3D came in Windows XP when Microsoft made icons more 3D. Prior to that, the icons were front-on representative drawings of computers, monitors etc, with 3D implied by lighter lines on the top and left and darker lines on the bottom and right; in Windows XP, all the icons were 3D representations of the ideas with continuous graded shading that made them less iconic, lower contrast and overall harder to process.
Debating "flat vs skeuomorphic" is an oversimplification of the debate, because it conflates icons (which should generally be mostly flat, as they are symbols) with UI elements (which should generally be 3D, as it marks things out as interactive).
The problem, of course, is that this confusion is now universal, because icons in almost all settings act as buttons.
If we look at iOS, the lines are very blurry, because the fact that icons must (Apple rules) fill the rounded-corner-square, means they're visually both buttons and icons.
On Windows, icons have transparency, so they're not visually buttons. So Microsoft invented "tiles" -- i.e. flat buttons to stick your partially-transparent icons on. The name alone tells us that "flat" is a stupid interface: you interact with buttons, but tiles just sit on your wall.
I know it takes three times longer to do what I need to do one a month at an ATM newly refitted with one of their bug-ugly flatscreen guis (get a balance, use it to decide which account to pay my credit card bill, transfer remaining balance to deposit if that was a checking account and grab some folding money). Mostly because each of those operations now spits the card out, requiring me to re-insert it and log back in to do the next bit. It's like the ritalin-addled foetus that wrote the algorithm didn't even use ATMs ...
UIs are just like clothes. You have to climb on board the latest fashion or get forgotten, or at best sneered at.
Remember when the craze for 3D-animated virtual spaces first arrived, though only a $5k workstation could cut it? Then we flung across back to 2D but this time shiny, so shiny that we next flung back to minimalist. Oh look, it's time for another fling. Where to next? Chunky retro? Text shell (because we all have voice assistants now)? Greying-out everything not in focus (so we simply can's shift attention away)? Surely not something functional, that will never beat fashion in the market place. (he types on his beautifully functional but deeply unfashionable Debian/MATE desktop, grin)
"every GUI using it looks like coming from 1982. No surprise it's also inefficient."
Actually the example you linked is efficient. There are relatively few objects systematically laid out, you can navigate them with the keyboard, menus are clearly labelled, there are no extraneous graphics and there's no cause to wonder where you are on the screen. It's everything a modern flat design is not.
I think that there are 'flat' interfaces and undifferentiated interfaces. TWM, Windows 1, Windows 2, System 7 (in fact, all pre-MacOS 8 MacOS) are all Flat - but perfectly differentiated (and, to my eyes at least, very clear and elegant)
Windows 3 - Windows 7, macOS 8 - macOS 10.9, iOS 1 - iOS 6 are all differentiated and 3D (to my eyes they're very clear, but also a bit gaudy)
Anything recent, as far as I can tell, other than some special Android / Windows skins or Linux variants, are flat (which is elegant) and undifferentiated (which is sadly unclear).
So don't have a downer on flat UIs. Have a downer on undifferentiated UIs instead!
Looking at the example in the article, I'd say that the one on the LEFT is flat. The one on the right is just completely borked.
Have a look. On the left, the "shop" buttons are blue, caps, underlined. No 3d shading or even 2d box around the border of the button. That, to me, is "flat". On the right, the "shop" buttons are in the same style as the body text - grey, mixed case. You could only find them by guessing from the words that it is clickable, or by trial and error (unless there is a mouseover).
This does not appear to be an argument in favour of the pre-iOS 7, and equivalent Windows, button styles. It's just an argument for not completely borking your UI.
It's worth looking at all their tests. Some are better than others:
Approximately the bottom third of the page is tests where they didn't find any difference between the two versions.
Years in the tech industry have demonstrated conclusively to me that nobody knows how to design, test or refine a user interface.
We peaked at about Windows 3.1 and it's been downhill most of the way from there, with only the occasional upward slope.
And nobody seems to realise how to test this.
You get a line of old grannies, pay them £10 an hour to wiggle the mouse. You give them a bunch of things to do and then leave them to it. Everything from "turn on your new PC" to "print out this document" to "can you backup those files for me".
Then you sit the developer and the granny alone in the room. Every time the granny asks something, he has to help the most minimal amount possible. Every time he is forced to change the UI out of sheer frustration with being asked so much, he has to start again with a fresh granny.
Only when people can walk in, do the tasks, find the way to do them themselves, and not get lost looking for where the damn Portrait/Landscape button is now, and the developer doesn't tear his hair out (or worse) answering the same questions over and over do you let him go back to developing new features.
And then after the new feature, you repeat the testing all over again.
Why this is complex, difficult or in any way worse than "Let's just change everything and hope the users recognise our One Supreme Vision", I can't fathom.
You know, when you wanted to print a file, you used to go to File.. Print or, at worst, the icon that looked like a printer. Those menus / icons are now GONE, hidden under the gumpfh of ribbons and contextual interfaces.
Gone are the days of actually making things simple, when people can make them "designer". This kind of thing sums up EVERY problem I have with the word designer. Expensive. Worse than non-designer. And nothing at all to do with "design" (i.e. fit for purpose, cleverly engineered solutions to the problem at hand) but to do with "designer" (i.e. looks fancy until you actually try to use it, when you realise it's the most overpriced junk ever).
Personally, I still install Classic Shell. I don't get why I'd want a recently used list over an alphabetical one, why I'd want a massive multi-column menu over a little diddy list of programs, why I'd want to hunt a tiny arrow that's liable to mis-clicking over nice big "Shutdown" or "Restart" buttons. Don't even get me started on Metro, which is basically Active Desktop 2.0. We rejected that back in 1995 for a reason, guys.
I honestly just want an OS / application suite that's customisable. Let people theme it and customise it, and then if they WANT it to look like Office 2000, well someone can just distribute that theme and all the buttons go back to 3D buttons with File menus and suchlike. Why does the UI determine how you use the program and not the other way around. And then, when that happens, let's track download of the Office 2000 theme versus, say, the Office 2016 theme.
And why the hell the UI ever changes for servers, I've really never understood. I can only imagine it's literally to sell the new MCSE/MCSA qualification where you've rejigged the question "What menu do you use to..."
I don't know whether it's the Windows programming model, or whether it's just crappy coding in Microsoft applications (or both), but Microsoft UIs are getting worse and worse. Just look at Skype on Windows - the amount of odd redraws, flicker and so forth is awful. Visual Studio when loading a solution - like a drunk trump staggering across a slippery road at night.
My favourite bit of UI nastiness is when some cranky old bit of Windows draws the default styled Window chrome, followed by a bit of flickering (presumably an application vs Windows message fight), then the whole chrome is redrawn in whatever it this year's in-house Office style. So frickin' amateur. I thought even Windows had a modern enough window manager so UI flicker and multiple redraws could be banished.
Although to be fair, iTunes on the Mac and a few built-in iPhone apps are heading down the same woeful path. So maybe UI shitness is universal, made acceptable by the 'take your time mate, settle down when you're ready' rendering style of Web browsers.
Visual Studio when loading a solution - like a drunk trump staggering across a slippery road at night.
Yeah gotta love that 'Oops I can't render the skin for this refresh so will have to just leave the underlying framework visible'. Always puts me in mind of the 'Mr X' image Homer Simpson used on his web page.
It might not have any 3D styling, but it does critically provide a very high level of contrast between it and its surrounding page area, making it absurdly simply to find and identify as something that can probably be clicked on, hovered over or otherwise interacted with in some way.
Let's not get too hung up over "flat" here - whilst the move to the more minimalistic flat styling was a bit of a shake-up after decades of 3D-styled UIs, at least those early flat UIs still provided clear delineation between their elements. The problem is *now* that UI design has progressed even further down the road of minimalisation, stripping away pretty much anything that lets you know which part of the UI does what. In some cases the designers didn't even stop there, and continued on to remove *every* visual hint as to where the active parts of the UI were in the expanse of seemingly anonymous whitespace they so graciously decided to dump in front of our eyeballs.
So I think when people are, quite rightly IMO, complaining about "flat" UI design today, many of us are really complaining about minimalistic UI design which is still promoted by some as "flat" design, if only because those few bits of UI styling which do still exist *are* just as flat as they were in the early days of "flat" UI design...
The rot has even gotten to LibreOffice which installs by default its "human" interface with fugly flat monochrome icons. Doesn't look to me like it was designed for humans.
First thing I do when installing it is download the "galaxy" interface and switch to it - color 3D icons that actually have some resemblance to what they do. But it IS a matter of taste, so YMMV.
The rot has even gotten to
LibreOffice our local <redacted> shop
Some bright whizz bought new tills for our local shop. The customer display (say 6 x 5 inches) displays individual items & price in black on near white, but in something like Font size 6 or smaller. The total however - while in a bigger font - is in pale yellow letters on a pale grey background.
Utterly effing useless; needless to say complaining achieves nothing. What tosspot thought that was a good customer - friendly design?
And sometimes you can't fix it...
And I thought it was me. I thought it was my fault I struggled to see the wood for the trees, or could not readily decipher light grey 6-point text on an expansive white background, or even know where to look to find the wretched button. Actually, I'm lying. I thought, "which tosser ever thought it was clever to camouflage the UX."
Its not just Brits - there number of sites everywhere where people think its sensible to use a mix of greyey blue contrasting not with bluey grey (or similar practically monotonic themes) which are deemed by 'designers' to look cool or stylish and yet breach every simple accessibility rule.
One of the joys of the internet is you can normally force your own css onto a page so you can see what the fuck is going on. You need an old colour monitor with access to all the colour contrils to make some UIs U.
My wife still cannot navigate or use a UI effectively, and flat designs have just aggravated the issue. I can no longer point her towards useful clues as to whether a single-click or a double-click is required - flat UI seems to depend on familiarity with the technology, rather than imparting it intuitively.
Once upon a time, a long time ago, Bill pontificated that windowed designs were "intuitively obvious".
Twentysome years later, billions of training hours later, we've discovered what many people believed BB (before Bill)................namely that this is complete crap.
....because every time Apple or Microsoft reorganise menus, every time we get an "improvement" like the Microsoft Word "ribbon", every time we get an improvement like "flat UI"...........we get a learning cost of billions of hours of user retraining.
Another example of this crap can be seen in the early 1990's with Windows Server. Who ever thought that managing a server needed anything more complex than -- gasp -- a terminal window?
But the "intuitively obvious windowed technology" got Microsoft billions of dollars of licence revenue.
The "improvements" (MS Word "ribbon", flat UI) got Microsoft billions of dollars of licence revenue.
Am I missing something here..............or is this all about dollars, and NOTHING to do with usability?
I've despised flat ui design for years.. the absence of any navigational cues. sure, it makes things sleek and clean, but also hard to use.
sometimes I've had to help others find how to do something simple and obvious on a web page, and had to ask for help, simply because it wasn't clear where you had to click.
yet another style over substance stupidity of the modern era.
I think the idea of putting fault to UX-designers (or computer people in general) is wrong and the actual problem is that user interface is designed by graphic designers in the marketing who have absolutely no idea of how user interface should be made. And they don't care at all.
People who see user interface as a piece of paper and usability is a word that doesn't even exist: They want just "clean" looks and that's that. That's what the graphic design is about.
So marketing opinion ("looks") trumps knowledge, basically. Is anyone suprised?
"Microsoft MUST have done something that showed the same effect in the new Windows UI, they test a lot and would inevitably have caught this amount of difference. And went ahead with it anyway."
Exactly that as marketing decided, not those people who actually know (UX designers). Yet another stupid decision in a long chain of them within Microsoft.
But not only in MS, it's a larger trend that marketing is overpowering the IT and that causes huge amount of totally stupid designs everywhere as marketing has no idea about functionality, they see only looks and nothing beyond that.
The flat UI worked well in places where the navigation paradigm was consistent, like a Windows Phone or a Zune (or to a lesser extent on Windows proper). You didn't need to know what a button should look like in Windows Phone because the navigation buttons were in pre-defined spots. It worked well even in unfamiliar applications because placement was consistent enough that you *knew* to click on the title, or look in a submenu.
This is not "Microsofts Fault". It's the fault of unimaginative copy cats getting it wrong. As one of 12 people world wide that bought a Windows Phone, I can tell you that the UI paradigm was leaps and bounds over everything else and nobody has caught up. Unfortunately, Microsoft turned around and watered it down in an attempt to find acceptance and dismantled some of the best features.
Really it is Microsoft's fault for the ridiculous assertion that what works on a tiny touch screen for a smart phone should be extended in both form and principle to the design of a workspace that is intended for expansive, professional use rather than limited, social consumer use.
Good UX design does not include the assumption that the user just knows where to touch the screen without adequate visual cues. This conflates and confuses the importance of an appropriate, clear and consistent UI in a constrained usage environment, with the fundamentally poor principles that result from a denuded interface lacking necessary visual or behavioural constructs.
Really it is Microsoft's fault for the ridiculous assertion that what works on a tiny touch screen for a smart phone should be extended in both form and principle to the design of a workspace that is intended for expansive, professional use rather than limited, social consumer use.
Microsoft have never managed to understand that desktop devices and mobile devices are used in different ways, at different times, and for different things. There is nothing to suggest that a single UI can be appropriate for both kinds of device, and everything to suggest that different UIs are required.
If you cast your mind's eye back to about the year 2000, and look at the UI of Windows for Pocket PC (for PDA devices -- smartphones without phones in them) you will see that it looks very like a cut-down version of Windows 2000. Those "Windows" PDAs failed because they tried to look as though they were running desktop Windows (rather than Windows CE) which (a) was unsuitable for the device format, and (b) led to the expectation that they could run desktop windows applications.
It wasn't until Windows Mobile and its Metro UI that Microsoft started to make any traction in the PDA/smartphone sector because they finally had a UI that was appropriate for handheld devices.
One might have hoped, at the time, that Microsoft had finally learnt the lesson that desktop and mobile are not the same ... but no, they saw the success of the new mobile UI and failed to understand it. They failed to understand why it was successful and that its success depended on the fit between hardware and software -- and promptly tried to move the new UI to the desktop.
The 'flat' and related interface principles are utter crap when it comes to usability - there are sadly very few neutral formal assessments of the usability of the more recent elements so this is a refreshing confirmation.
The sad part is that too many people actually buy in to this rubbish (corporate and consumer) because they fail to acknowledge the impact of the design and don't assess the difference between new versus better/improved.
This isn't an expression of personal preference and aesthetics - this is about good UX design. Anyone who spends time working in any sophisticated way with computers knows already that if you remove the visual indicators from screens, or you make them transient, or you jazz up the view with unnecessary prettiness, or you create a requirement for motion where it isn't needed, then your ability to work in an efficient manner is reduced.
"The 'flat' and related interface principles are utter crap when it comes to usability - there are sadly very few neutral formal assessments of the usability of the more recent elements so this is a refreshing confirmation."
I disagree. The need to "show" a user what constitutes an interactive element can be in itself a failure of user interface design. Metro worked very well because the the user intrinsically knew where to find what they were looking for. Not so in any of the copy cats that took one *minor* element of Metro (flatness) and decided that was enough to make a good clone.
I agree with your point (good UX should not include unnecessary elements - even Microsoft state this in their UX design guidance) but the 'flat' UI problem is that it omits necessary elements. A completely blank screen with touchable areas doesn't need to show what is where if the user already knows, but even then the visual cues would help use it more efficiently.
FWIW no user intrinsically knows where to look for something without some initial visual cue. Although users can learn or be trained to make those assumptions, and a consistent UI will generally help increase usability, an assumption of intrinsic knowledge isn't a useful principle for designing generalised user interfaces (although understanding acquired knowledge is useful when designing modifications to UIs).
Also this is why we test usability design with users who lack familiarity with the prototype specifically to identify how effectively they will achieve the working knowledge to complete tasks.
"This [free] book answers two questions. The first question is "Why should I make my web site more accessible?" If you do not have a web site, this book is not for you. The second question is "How can I make my web site more accessible?" If you are not convinced by the first answer, you will not be interested in the second."
Isn't anyone looking at the examples? The "good" one has the link in a contrasting color, all caps, and is underlined. The "bad" one has the link the same color as the text, upper/lower case, and no underline. Neither link example has 3D or shading effects, which means this isn't a test of flat design. It seems more like an example of bad flat design vs. good flat design.
Flat design typically puts links in a contrasting color, or perhaps in white text in a shaded box to stand out (no 3D button effect, just strong contrast). It can often use all caps. Standing out is just good design. You don't need 3D effects to do that.
My conclusion is that the testers have no idea what flat design means, and created a faux test to try to create news.
If you read the linked article, the researchers talk about signifiers, and comment that "weak or absent signifiers" cause the problem. They also assert that flat designs tend to have a lot of weak or absent signifiers, and while they didn't attempt to prove that in the paper, I agree with them. Often the "hamburger menu" button at the top of webpages has no button-border (so the only signifier is the icon itself, hence weak) and even the search function is an unmarked clickable textarea (absent signifier) made worse by the fact that in some UIs you have to click in the text area, in other you can click in the text area or on the magnifying glass, and others still, you have to click the magnifying glass. Stronger signifiers makes this inconsistency less problematic.
Yes, flat design doesn't mean no signifiers, and 3D design doesn't always mean strong signifiers; but flat design biases away from strong signification, and pseudo-3D design biases towards strong signifiers. (Whereas true 3D design often eschews "signifiers" altogether and relies on stupid "object manipulation".
" The "good" one has the link in a contrasting color, all caps, and is underlined. "
Irrelevant when all those elements exists in the page body too: There's no way to distinguish links from text or next level, control buttons of the software itself: It's all one big soup, 'flat'.
Totally stupid already at concept level as there's never any advantage of having an UI like that.
That's why you have to have something the body can't have for control. Like fixed header with fixed buttons.
Was introduced by engineers so that peole could make split second decisions unhampered by any especial need to interpret the interface.
Then engineers stopped designing stuff, and the task was left to art students who could drive ';software productivity tools'
Leading to such wonders as the new Barclays Smart Investors site, which features an airbrushed pastelized couple being 'smart investors' but won't let me withdraw money.
Smart Investors are waiting until the site works well enough so that they can remove their portfolios.
It's been 20+ years since I've used CAD software so my skills have atrophied. I initially assumed that it wouldn't be that difficult to relearn. That has not been the case. Now that I know it isn't just me, I would like a solution. The current plan is to bring at least one more person. This creates two problems. First, capital is extremely limited and paying a salary blows my budget. Second, I want things done right so I prefer to do as much as possible myself. I prefer to finalize the design, then have a licensed engineer sign off on the government required paperwork.
Aside: I did a test using my GF as the guinea pig. In her case, 22% is far too low. She gave up on the flat UI so I have no clue how much worse flat is for her. The test was done using two different subjects as using the same basic content would have skewed the results in favor of the second page. Both subjects were outside of her knowledge base.
I absolutely loathe flat controls that are so often hard to distinguish from the content and/or decorations.
Skeuomorphic designs, so long as they are not carried to the absurd lengths that led to the "flat" revolution, are easier to navigate, easier to see and easier to manipulate.
Why the fark should I have to change my entire mindset when I move away from my (real) sound-mixers with analog controls to a screen-based version which is totally non-intuitive and (for me far slower)?
Wanna see some nice icons? BeOS had a big set of perfect icons that immediately told you what they were for - now I have to look at a black square on a white background and see if I can remember what it does and why it is different from the same square rotated 90 degrees.
Flat shit - first thing I do with a machine that I am going to use seriously is to whip up a theme that makes some kind of sense intuitively rather than a bunch of meaningless boring Mondrians.
El Reg publishes this article about flat UIs being shite.
The commentards seem to overwhelmingly agree that flat UIs are shite.
The next day, El Reg alters the symbol at the top of each article that leads to the comments section. It used to be oval (ish) and blue and stood out. It is now rectangular (ish) and the same colour as the body text, and doesn't stand out at all.
I notice that "flat" icons and buttons lack robustness to unexpected changes in visibility or composition. Take a pale single color shape without an outline and put it on a background of uniform color and it will likely disappear if both colors are similar enough, or have same luminosity. The outline or the shadow of a 3D icon with multiple coherent layers would still remain visible, so the icon can be read from the fill or the line, whichever works. A color clash commonly occurs if you look at an LCD from an angle, or with icons that can be freely arranged.
Another vote for beautifully designed BeOS icons, or the similar ones from Windows 98/2000.
In 2014, Jakob Nielsen confessed: “I really do hate it [flat design]. It’s this hot, strong word, but it’s almost true. I do think that it embodies some of the worst aspects of design, which is fads and fashion and disregarding the user’s needs”. I really don’t understand why NN/g tries to “conciliate” flat idiocy instead of declaring exterminatory war against it.
Useful collection of scientific articles about flat design: FlatIsBad.com
UI design doesn't get much better than what Windows 95 introduced, if you're talking about navigating efficiently. Later on some UI elements were introduced that helped speed up controlling applications. Starting from Windows XP it started to go downhill. Fortunately, you were able to select 'Classic' UI on these modern versions, which made a perfect combination, until it was removed from Windows 8 (or was it 10?).
This is all due to the hordes of under-educated people who call themselves UI designers, combined with clueless self-proclaimed software developers from ever lower wages countries. Blame it on corporate greed.
As for Science; several renowned institutions have done usability research in the 80's to scientifically support what some people already intuitively knew, but what "UI designers" seem to not get.
The people who intuitively got it were working at Xerox PARC, developing the Xerox Alto in 1973.
Famous Jakob Nielsen quote: " I can just barely understand companies that ruin their user experience because they don't want to pay $298 to find out what the usability research says."
Biting the hand that feeds IT © 1998–2019