It's a rant of 1h 17mins. The video is funny but it's very centered on the author's usage of Windows UI, which is obviously very specific.
When you ship a product like Windows you have to be very mainstream, you can offer options to customise the UI, fair enough, but honestly the best option is to be dead boring. Maybe once in while you can bring in a new UX concept. Thinking out of the box and trying to reinvent the wheel like they did in 8 doesn't work.
I'm not trying to defend Windows, I think it's been going downhill in term of UI since 7 but not for the reasons he mentioned.
These were my first thoughts too, he's ignoring the kilotons (megatons even) of constraints of a company like Microsoft and its flagship have to navigate around. He's also so far above the average user's understanding of things. Most of my neighbors are not at ease when I tell them to `right-click`
MacOS is if anything more constrained than EWindows in terms of GUI (largely: desktop) customisation.
As the video actually explores at some length, Linux, or more accurately, the world of X11 window managers and develoment libraries, shows some real exploration, most notable with wheel/pie menues.
The video is long and centered on the Windows shell, but lots of it is applicable to the other platforms. It’s exactly what I’m interested in and I’m thoroughly enjoying it:
“Hi! One of the things I like about the Internet is the long tail, which means everybody has specific interests that won’t appeal to most people, but for the people who are interested, it really delivers!”
[...]
”Plus what a wonderful choice I’m being given! Choosing between a more secure system, or being able to use my GUI the way I want to!”
“Yeah, I get it! I’m a freak and I’m the only person who’s doing this!”
”But riddle me this: why do security updates need to screw with the graphics like this‽”
After 20 minutes of being shouted at, I got his opinion on icons on the desktop, that he likes "nice looking icons" (icons aren't about nice but recognisability). He wants a one-click wallpaper changer. He likes pretty art. MS UIs are going to crap[0]. I wasn't going to watch more.
He says upfront he's not an expert and he's not. I thought it was close to worthless. YMMV.
I found a lot of very valid points presented in the second half from gesture navigation to inefficiencies in default navigation I hadn't thought of.
Watching at like 2x speed is a must tho. I wouldn't have been able to make it thru otherwise.
I don't think its specifically about having a one click wallpaper changer so much as having the ability to define and configure what happens when he clicks different UI elements. For HIM that means the wallpaper changes another user may want differing functionality.
For the slowly delivered part, you can turn up the playback speed. I found 2x speed fairly comfortable (I occasionally turn up playback speed but normally not above 1.5x).
One thing that surprises me (and may be the source for some of his frustrations) is how much for a power user he uses the mouse. This isn't a criticism, just an observation.
He, fairly, criticises how long it takes for the filename to show up when he hovers over an icon on the task bar. However, I think the vast majority of people at his level would use Alt+Tab which immediately brings up not only larger icons for active applications to swap between, but the filename above.
If you have a lot of apps open, Alt-Tab no longer shows them all, you have to repeatedly keep bashing Tab to get the window to scroll down to the hidden ones - there's no scroll bar and mousewheel doesn't work. For this reason, I've completely stopped using Alt-tab on Windows 10.
Win+Tab is the 'replacement' but is slower and riddled with bugs - one annoying one is that some times it doesn't show the preview of half the apps. Also, the order of the windows is always changing, so you waste time scanning for the app you just used.
You can also type Win+<num> where <num> is 1-9 and that number icon activates from the taskbar. Which is great if you have 10 sticky icons down there, but if like me you don't 'stick' icons to the taskbar, it not as fast (muscle memory etc) or as useful as it could be.
I totally agree with the first part of his video (the main Windows 10 rant). Each iteration of Windows makes it slower for the power user. When the Year of the Linux Desktop is finally upon us, I guess I'll eventually switch. ;)
Yeah. I've always wondered why MS puts the window previews there when the Timeline view and multi-desktop view shows the window contents.
macOS imo does a much better job: Mission Control is for all graphical work. For power users, Alt+tab is apps only, no previews, and then once you're in an app, you can Cmd+` to cycle through each window. Or, set a gesture to 3-finger-swipe down and see all the windows that belong to a specific app.
What struck me is how, given a trackpoint (or trackball), all the mouse arguments fall away. Fingers are on the home keys, keyboard homeless or otheerwise.
I'm not fully opposed to mice, and having each of trackpoint, trackpad, mouse, trackball, and tablet could be a plus.
Long. Much better at high speed (I went 2x). "Gee-Yoo-Eye" rather than "gooey". But quite good in its criticisms, observations, and possible moves foreward.
One of the strengths is the list of recommended features for a GUI. That really should be called outin the video description or several linked discussions, but doesn't seem to be.
1. The GUI should be asthetically acceptable with heavy and long use.
2. There is no perfect GUI for everyone.
3. There are many features that are ideal for the majority of users.
4. The GUI should get out of the way when you don't need it.
5. The GUI should be as efficient as possible if you know what you're doing.
6. The GUI should activate when you want it and not when you don't.
7. It shouldn't be easy to do things you don't want by accident.
8. Switching between mouse and keyboard should be minimised.
9. Use multiple mouse buttons for maximum efficiency.
10. Both low-contrast textual and stark contrast graphical elements should be avoided.
I'm not in full agreement (#3 may be debatable) and would add a few items (especially: preserve user state if at all possible). But it's a good starting point for discussion.
In the video the narration spells out "GIF", which offends my space alien cat sensibilities tremendously. However I forgive him this, the analysis is sound and informative.
I mean just because I invented the television it wouldn't mean it would be my right to claim that the correct pronunciation of TV was actually TP. But let's not get bogged down with who's right and wrong on the internet and focus on the important stuff.
The changing UIs are essentially a marketing sceme, not having to retrain how to use a UI is almost always significantly better. Mobile UIs on the other hand have really improved in the last decade.
When I first moved to Japan I was appalled by all the old GUIs used everywhere and the never renewing systems (still using fax for example), now I think that changing styles every few years without good reason should be a crime.
To have a great GUI, beyond looking nice and getting out of the way, it must be intuitive, and for that you need a metaphor that works, but the moment you use one you are also constrained, it's a very hard problem IMO.
I like in the video is touching on Science fition interfaces, I found this video[1] you might ejoy, the author Chris Noessel has a google talk[2] about intefacing with AI in different ways and touches on interaction design, I'm trying to learn more about this.
There was a link posted on HN that I can't find anymore, it was about better UI metaphors dropping breadcrumbs to traceback your path when you navigate across different sites and some 3D desktops with piles of documents ... I know is vague description, if annyone knows that link, please share :)
> To have a great GUI, beyond looking nice and getting out of the way, it must be intuitive, and for that you need a metaphor that works.
I don't buy that there is "good" GUI design in any absolute sense. "intuitive" is thrown around with too little thought in GUI discussion even though it is ambiguous.
When disassembling the meaning of intuition is turns into familiarity - but there is a whole spectrum of familiarity. At the most fundamental there is _innate_ familiarity e.g direction, which is why arrow keys and cursor input are innateley intuitive. At the other end of the spectrum we have high level design mimicry or historical use, skeumoprhism, e.g the floppy disk save icon, checkboxes, 3D relief in UI elements as a reference to physical input devices.
This is why "good" and "efficient" GUI design is also highly subjective, not because we don't all share a history that can be tapped for familiarity, surfacing conventions which are hard to escape - but because it's possible to be more efficient, more effective if we abandon those short term shortcuts and only hold onto the innate intuitions.
For example I use a tiling window manager, I get to define my own shortcuts - all of them. I find it intuitive to alt+arrow-key to change window focus, and shift+alt+arrow-key to move a window - this comprises innate intuition + artificial almost contextless fabrication - i.e no normal user would be able to figure it out without reading a manual or defining it for themselves.
This would also not generally be considered good GUI design, because it's not discoverable by people who are already familiar with historical GUIs, you have to put in effort to learn it - as would an alien using windows UI. and yet it is more efficient...
In summary, "good" GUI design for the masses is highly constrained and dictated by history, not ability to conjure up good metaphors.
> I don't buy that there is "good" GUI design in any absolute sense. "intuitive" is thrown around with too little thought in GUI discussion even though it is ambiguous.
Long ago, in the primitive time of the 1980s, people actually did studies to show what worked and what didn't in GUI design. Almost all the lessons learned from those studies have been thrown away and replaced by... I don't even know. Gut feeling? Ease of implementation? Prettiness? Certainly not usability studies.
The idea of hiding all the icons when not on the screen surprised me... I've been using computers since 1979, and never thought of doing that. I do like his point about the increasingly small and easy to miss targets that GUIs have been forcing on us... especially now that my laptop has a 1920x1080 display and a crappy NO BUTTON touch pad. (more than 1/2 of the time I get a right-click instead of left-click!)
If I could somehow activate a menu with maximum 8 choices in each keyboard direction (that are NOT subject to automatic moving around, so that a given motion is repeatable)... and cascade that as many levels deep as required.... I could get a lot more done, a lot faster.
Watched about 20 minutes.
Maybe there's a point hidden somewhere in between his rantings, buts it's quite hard to grasp it.
He moves back and forth between actual criticism of the Windows UI and just stuff that he personally likes as an advanced user.
I agree that Windows is far from perfect (I'm mostly a Mac user) but IMO the biggest issue is complexity. For tech savvy users all these are non issues, but for my Mom it would be impossible.
I have always spelled it out in my head, and the first time I heard someone say it I was utterly confused as to what they meant. To me, if I were to try to pronounce it, it would sound like "guy", as in the first syllable of "guide". Unlike Nasa or scuba, there's no universally obvious pronunciation for GUI—and thus I think it should be spelled out.
I have not changed my pronunciation or any of my own opinions based on this video; I only listed lessons to be learned from watching it in its entirety.
While there are loads of valid criticisms here, a lot of time is wasted on complaining that Windows broke the weird, buggy shell customisation tools and modules that the author uses.
Some of the problems ("why are these resources in DLLs and not files") are there because if Microsoft changes them, loads of programs break for many more people. As opposed to the people behind many shell customisation tools, other programs do use the APIs that were intended to be used by developers and Microsoft actually tries to keep compatibility with those above some kind of quick launch menu some guy customised back in the XP days.
I think this guy would be excellently at home on Linux, although he picked some of the more fancy distros as examples in his video. His Windows setup reminds me of LXDE and XFCE. The dock, the floating panels, everything can be replicated and practically guaranteed to work despite updates. Now, program compatibility for stuff like video editors is an entirely different subject so I don't expect this person to switch, but as far as GUI problems go, Linux is the best place to solve them to prove your point.
There's one problem though, and that's that nobody cares about you and your long tail. Linux people used to be all about this stuff and that's why Linux was always considered arcane and unusable. Linux shell developers used to develop a lot of their software for themselves and their usage patterns instead of for others, making their only usable for people with the same expectations. Because of the lack of a unified GUI, there will never be centralised settings for the entire system, there's always long text files. If you are in the long tail, pick up Arch and go ham. But please, for the love of God, leave the GUI for normal people alone.
I used to do tech support over the phone, which most commonly consisted of getting people to launch Teamviewer and fixing whatever problem they had. Every fancy, smart, helpful feature that some whizzkid installed to help their parents made it completely impossible to help those people. They don't know what a start button is, they can't describe their screen to you, there's no standard buttons, labels or anything to help people, manuals are suddenly all wrong ("I've got the start menu open but I don't see control panel" because the shell renamed it to "settings") and auto hiding task bars made people think their computer was broken.
There's improvements to be made, but most people require simplicity over customisability. Custom icons break software, break expectations and often break metaphors. Custom task bars break old, buggy business software that hundreds of thousands of people rely on. Hell, I've seen some tools break because I decreased the width of window borders by a few pixels. Every customisability option needs to be tested for functionality and compatibility and developers' time is much better spent at making features normal people care about.
You want a perfect shell no matter the cost? Build one. The shell launch API is there in Windows, changing most system icons is supported through the registry. Window theming for most of the OS can still be used through the Windows 2000 theming system. Even more is possible by passing certain system messages to windows of other programs like explorer. You'll break stuff, but you'll break them just the way you like it. Don't expect anyone to build a GUI for you if you know your requirements don't matter for the vast majority of people.
I watched half of it. Will most likely finish soon. I like the guys passion. I had some experience trying to run litestep in the past and I was really into things like windows blinds for a while. Looking at his setup, it looks infuriating,but if it works for him, that is what is important.
No, I was a heavy LiteStep user during its prime Windows XP days, but even then it was a brittle, badly developed and hard to set up thing. Development has stalled once Vista and 7 came along.
I think Web-style navigation was introduced as an option in Windows 98, along with Active Desktop. Back then, Microsoft envisioned merging local file system and web browsing in Explorer, a direction they later reversed when the IE/Windows lawsuit hit.
95 was focused on a spatial metaphor like old school MacOS and 98 started the switch to a browser metaphor.
"Why do we have lists of tiny icons to click on instead of hiding big ones until we need them"
It's called putting the parts on the floor, it's why Lego is intuitive.