Hacker News new | ask | show | jobs
by _zamorano_ 803 days ago
These are the little things you don't notice but make a difference.

It's a pity this kind of attention to detail is becoming obsolete in favor of flashy but unconvenient UI

I expect this behaviour on any list I find in a Windows app. I also expect keystroke consistency, like press F2 to edit... but all this UI things seems old fashioned.

I assume I don't get that on web apps, but "modern" Windows Apps are also deprecating these conveniences

I wonder if in the apple sphere they're suffering this kind of degradation.

4 comments

The best example of this decline comes from Microsoft itself.

Here is how Microsoft recommends you do find-and-replace in their simple user-friendly ~20 year old note-taking app, OneNote [0]:

1. On a blank page, type the replacement text to use, or find it on a page.

2. Select the replacement text, and press Ctrl+C (⌘+C on Mac) to copy it to the clipboard.

3. Press Ctrl+F (⌘+F on Mac) to find on page, or Ctrl+E (⌘ + Option + F on Mac) search all open notebooks.

4. In the search box on the top left for Windows, top right for Mac, type the text to find.

5. In Windows, you can select Pin Search Results at the bottom of the results list, or press Alt+O to pin the list. Mac is already pinned.

6. In the Search Results pane on the side of your window, select a search result (a text link next to a page icon) to jump to the page where OneNote has highlighted the text it has found.

7. On the page, double-click or select each highlighted occurrence of the text, and press Ctrl+V (⌘ + V on Mac) to paste your replacement text over it.

Note: When you replace a word or phrase in a sentence, you might need to type a space after the new text is pasted.

8. Repeat steps 6-7 for each additional page in the search results list.

Tip: If you've got a lot of replacements on a single page, copy your text to Word, find and replace the text, and then paste back into OneNote.

[0] https://support.microsoft.com/en-us/office/find-and-replace-...

OneNote seems to be especially bad. I don't think there's been a day in which I've used it and not had it complain about sync errors with my Office 365 account.
I am speechless. Just wow.
> degradation

Definitely, it’s an industry-wide ailment caused by a focus on the web and neophyte users.

The irony is that this power user functionality didn’t impede new users, it’s just been gradually forgotten by folks only raised on the web, where almost everything had to be reimplemented from scratch.

We gained a lot but lost a lot as well.

Every time someone on HN praises Flutter for being the best way to do UI’s, I shudder to think of all the conveniences I’m used to (on both windows and macOS, I’ve used a lot of both) that are likely completely ignored or forgotten by these UI’s as they struggle to reinvent the user interface from scratch.

The UI’s of the 1990s were often designed using actual user studies, with years of hard-fought learning on how to do things in a way that is discoverable, Accessible, and with effective shortcuts for power users. I worry we’re losing all of this as we reinvent the UI without understanding what we’re rewriting.

It's crazy to me that so many software companies dumped millions and perhaps even billions of dollars in R&D collectively to learn how to do UI&UX effectively in the 80s, 90s, 00s, etc., and every new generation just ignores the previous one completely because it looks dated.

Relevant XCKD: https://xkcd.com/1053/

Gen Z, or Gen Alpha, or whatever you call teenagers these days, aren't born knowing how to use a desktop computer. They're born knowing how to use smartphone apps. From what I hear, many of them are terribly ill-equipped to be actually productive with desktop.

They're no different from just an office guy who never used a computer in the 90s or from a kid or got a PC from his parents in the 00s.

Why would the GUI have to adapt to modern users when a new user now is no wiser in practice than a new user back then?

>every new generation just ignores the previous one

One of the easiest first steps when creating your own identity as the new generation is to refuse your predecessors wholesale. This goes triple for something that is about fashion, which GUIs to some extent are a fashion statement.

> I worry we’re losing all of this as we reinvent the UI without understanding what we’re rewriting.

It would get rediscovered in the future same way old ideas are recycled and are new again.

For some reason, this hasn't happened yet, and the time frame that has passed since the "loss" (e.g. from ~2010 on) is comparable to the time frame the UI patterns were originally introduced (e.g. Windows 1.0 to Windows 95).

I'm afraid the prevalence of touch-screen use and the lack of focus on power users by UX designers is making the degradation permanent.

No. Every creation and discovery is influenced by culture and individuals.

Even if you assume the same level of skill and effort will be used (it won’t) there are still many paths and forks to take.

Consider that most “ui designers” come from graphic design or psychology backgrounds, while 90s UI was designed by engineers. Those groups have different values and make different decisions.

Yea, I think a lot of us old timers do notice these "little things" and it's infuriating when they are missing from the Shiny New Web Version of our desktop apps.

And, like others said, the existence of these niceties does not detract from the experience for non-power users. It's simply a bonus for more experienced users, and these bonuses are slowly going away as more and more developers choose the "give up once the thing kind of works" strategy.

I think what makes it worse is that a lot of 'UX' people seem to actively oppose these kinds of power user conveniences because they think they know better than their users.
It’s not that they think they know better, it’s that they know power users are the minority.

Everything is about minimizing expense to an extreme degree to drive “growth” in the current tech economy, so there’s little focus put into things that do not test as an immediate boost to marketability.

I wish this was the reason. It's just mediocrity and pure, raw ignorance. A result of throwing bodies into the technology industry in a desperate attempt to get the software out.
Also power users may be using ad block or opting out of tracking and therefore being left out of the success metrics.

Everything will be optimized towards the clueless.

Considering how much money are poured into AI development would not say that "everything is about minimizing expense". But it is true that user interfaces are not something that will sell your application to the masses.

But to be honest, trying to sell any kind of "advanced software tools" (like compilers or other very specialized tools that provide small gains compared to existing free ones) for "power users" is extremely hard. Was involved in something similar in ~2010 and many power user think they don't need it or they can do it better themselves but anyhow does not want to pay for a complex tool.

Probably the web interfaces is exactly a manifestation of that, many users trying to implement things themselves because they know better. But I prefer writing interfaces with React than with Win32 so I think there is progress.

It's also a lowest-common denominator problem.

Which advanced UX do you implement? MacOS? Windows? Gnome or KDE behaviour? CDE?

One could sniff the user-agent and adapt to a recognized OS behaviour but we all see how superficially shallow the UX on the web is. I've given up on expecting anything advanced.

Any UI is as good as its base layer. We wouldn’t have $subj if MS didn’t implement it. But then when you tell them that the browser UI model is just useless crap with smooth animations, you get hellvoted, yelled and thrown fragile css/$()/useX incantantions at.

Which advanced UX do you implement? MacOS? Windows? Gnome or KDE behaviour? CDE?

Browser advanced UX. Not current browser UX, but a hypothetical one which doesn’t suck.

Accessibility enhancement are not just meant for power users though. Some people depend on those.
There will be no power users when you take away all the powers.
Yes, UX people do recognize it's weird to have something invisibly act a certain way that disobeys the plain inputs.
Desktop programmers got these things mostly for free, even if they gave up as soon as it compiled.

Also, I focused mostly on the web, but things like Apple/Gnome removing menus and titlebars are a problem as well.

I don't know if this is that clear cut that it's worth making sweeping judgement based on it. I don't think it's a great idea to have it so pressing the same letter twice selects the 2nd item in the list with the same starting letter.

From another fellow grumpy old-timer who appreciates details, this looks more like "undiscoverable alternate mode that is invisible to the user, hacked in by a too clever engineer" than power user nicety

This is the way I remember it working originally. It wasn't until later that I discovered you could type the letters of something to jump to that. So for me, typing the first letter multiple times was just the intuitive and obvious way to do it.
It is discoverable through “g windows tips and tricks”. Reading articles like that went out of fashion too though. Thing with “power” and “advanced” is that you have to learn, because otherwise that info for all features had to be on screen and there’s only so much space on it.

This whole discoverability bs is what has driven UI into intuitive uselessness for everyone, when there’s actually two groups: first (/one) time users and just users (“power” or not). For some reason modern UX designers think everyone is in the first group. Even if an app is naturally multi-time heavy-use.

> because otherwise that info for all features had to be on screen and there’s only so much space on it.

there's practically infinite space on screen since that space has a time dimension and can also be tied to context which also varies. Like with this feature: you can show a tip on the first few searches within a list, you don't need to permanently keep the explanation on screen. Or you can show a hint with ll by having a different style of the second l or something similar

Undiscoverability is precisely one big reason why these things get axed since they're not used, so forgotten about in other contexts (web, new UI framework etc)

And yes, needing to reading random articles like this is a major fail in discoverability as you decrease reach and increase cost for the many poor users

I hate the "discoverability" trend. I don't want discoverable UI. I want UI that is consistent with all my other apps, which I have already learned. I should not have to have an app nudge me to "discover" its oh-so-cleverly-hidden features. I'm not Dora the fucking Explorer, I'm a computer user who expects everything on the platform to behave predictably and consistently.

And while we are at it, "lack of use" is not a good reason to axe features. Some things, by their nature, just don't get used as much but when they are needed they are needed. Product designers have become slaves to their telemetry and metrics, and are letting the tail wag the dog.

Probably more historical. My recollection is that lists originally worked that way (same letter many times) for a long time. Then a decade+ later I realized it had become smarter.
> Definitely, it’s an industry-wide ailment caused by a focus on the web and neophyte users.

The constant praise for web UI despite the issues always irritated me deeply.

It's like you're supposed to only say nice things about where we are because other people are saying that. And the more impressionable among us believe that because the Thought Leaders are saying that, then it is truth. And the circle continues.

Because most of the time you didn't have to implement this. The UI toolkit implemented this. When you have to do everything from scratch, most people don't.

Take a simple link navbar with a menu of links on a webpage. This has existed since forever. It should be that you can open the menu with your keyboard and use up/down to navigate it. But that means writing JS code. So devs who did it from scratch didn't do it, and only those who used an UI library with that already programmed in by someone else got it.

But "waaaahhhh native apps are soooo hard" they cry.

IDK man, every project from a single developer who uses Win32 widgets is infinitely better, more predictable, more usable, and more professional looking than any shitty electron app.

I agree. I'm writing a desktop app based on a web stack and I'm agonizing over tiny details in how tabbing should work and how the arrow keys should interact with lists etc. I don't know if my UI is any good, but I know that good UI takes a tremendous amount of attention to detail.