Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Microsoft hasn't had a coherent GUI strategy since Petzold

jsnover.com

Post reply on HN
Print view View original

591–600 of 604 posts

Re: Microsoft hasn't had a coherent GUI strategy since Petzold

#591

Post by throwaway2037 »

irishcoffee wrote: post #496
vovavili wrote: post #470

Earlier quoted context omitted.

In my experience: - Qt Widgets worked fine, but looked like a piece of software made in 2013; - QML looks stylish and is a very nice language, but had a lot of weird bugs. Neither of these are issues I'd run with if I were to make a web app.

Widgets looks like whatever you want them to look like, if the feel like they're from the 2010s its because the implementer made that choice, not because of a limitation in qtwidgets.

As I understand the philosophy of Qt, it does not (for the most part) use native widgets. They seek to draw widgets that match native widgets on a pixel-by-pixel basis. You can also change the look-and-feel (similar to Java Swing widgets) by using what Qt calls "styles". Ref: https://doc.qt.io/qt-6/style-reference.html

Re: Microsoft hasn't had a coherent GUI strategy since Petzold

#592

Post by account42 »

iamcalledrob wrote: post #471
throwaway2037 wrote: post #425

Earlier quoted context omitted.

What about Qt? It is the gold standard for cross-platform desktop UI frameworks. It is (1) performant (C++-based), (2) does not look outdated, and (3) not bug-ridden.

Qt apps don't feel great on macOS, though it's by far the best for mac-ish UI. Dropbox was Qt for a long time and I'd argue it worked well for them. Its easy to fall into "uncanny valley". On Linux, Qt apps feel a bit off in GNOME, though you can never satisfy everyone as its the wild west. I think Qt also suffers from not really being anyone's favourite. On the one hand, you have web developers who tend to not reall…

> On Linux, Qt apps feel a bit off in GNOME, though you can never satisfy everyone as its the wild west.

It's much better than how GTK apps feel under Qt-based DEs - at least Qt tries to adapt to the DE style.

Re: Microsoft hasn't had a coherent GUI strategy since Petzold

#593

Post by account42 »

irishcoffee wrote: post #495
gr4vityWall wrote: post #492

Earlier quoted context omitted.

I think Qt is only missing well-written, feature-complete bindings for a major JS runtime, including support for hot reload. Developing UIs without hot reloading is too painful.

I think what you're asking for has existed for a long, long time. QML.

QML is an entirely separate widget toolkit that really isn't a great fit for desktop programs - it's not much different from using a web-based toolkit.

Re: Microsoft hasn't had a coherent GUI strategy since Petzold

#594

Post by account42 »

jasonlotito wrote: post #423
shmeeed wrote: post #379

Earlier quoted context omitted.

>their telemetry showed that few users used them I wonder if they ever stopped to think that power users are the ones that disable telemetry immediately upon install.

That's not remotely universal, but they did consider that. It's immaterial.

It's really not immaterial. Power users are also the ones that get anyone to install Firefox in the first place.

Considering Firefox market share trends all decisions at Mozilla should be by default considered textbook material for what not to do.

Re: Microsoft hasn't had a coherent GUI strategy since Petzold

#595

Post by gr4vityWall »

irishcoffee wrote: post #585
gr4vityWall wrote: post #542

Earlier quoted context omitted.

QML doesn't have a way to define interfaces with JSX and doesn't integrate with the wider JS tooling. From my very limited experience, it still feels too close to the C++ world.

This isn't true, but I understand why you feel that way. Qt had bindings to many languages, C++ is a choice, not a demand, if you'd like to use Qt.

> This isn't true, but I understand why you feel that way.

By all means, I'd be happy to be proven wrong. Last effort I saw on this direction was node-gui + React. I believe that project isn't being maintained. Being able to use something like Vite to build desktop programs would be a blessing.

Re: Microsoft hasn't had a coherent GUI strategy since Petzold

#597

Post by Doxin »

bvanheu wrote: post #581
Doxin wrote: post #578

Earlier quoted context omitted.

IMO the ubiquitous Yes/No/Cancel is even worse. No and Cancel are too conceptually close. Doesn't help that these usually show up when you're about to lose all your unsaved changes. We've got big screens now! use more words! Save changes/Discard changes/Don't quit.

"cancel" means cancel the last operation (e.g. "quit the program", "yes/no" is an action taken on the prompt.

I understand those prompts perfectly fine, but they are panic inducing for e.g. my mom who has about a 50% chance of clicking the wrong button and losing work.

Re: Microsoft hasn't had a coherent GUI strategy since Petzold

#598

Post by raxxorraxor »

jasonlotito wrote: post #423
shmeeed wrote: post #379

Earlier quoted context omitted.

>their telemetry showed that few users used them I wonder if they ever stopped to think that power users are the ones that disable telemetry immediately upon install.

That's not remotely universal, but they did consider that. It's immaterial.

The power users I know do that immediately and then it becomes uncertain. Also, most of these metrics suck. We didn't always collect such metrics and I wouldn't say it did improve software in general. On the contrary, it probably brought in far more dark patterns than better UI.

To hell with it and then some. Also to ignore uncertainty means you will be wrong. There is no "perhaps" or "likely" here.

Re: Microsoft hasn't had a coherent GUI strategy since Petzold

#599

Post by int_19h »

hyperrail wrote: post #510
hyperrail wrote: post #499

Earlier quoted context omitted.

(disclaimer: I was an individual engineer in the Windows division during the Windows 8 project, i.e. reporting through Steven Sinofksy) I think you're being a bit unfair to the Windows division during the Win8 lifecycle. Maybe that's just my rose-tinted glasses though. I know there are some HN/proggit commenters who like to harp on the supposed toxic rivalry between the Windows orgs and Microsoft developer tools orgs…

> What actually happened [re Windows RT, but I think the point applies to UWP in general] was that the users went WTF because none of their native apps - which, contrary to his take, were very much alive and kicking! - worked there, and devs went WTF because they were told that they'd need to rewrite everything yet again in some new thing that was kinda sorta but not quite like WPF, because Windows just hated .NET th…

Thing is, Silverlight and Windows Phone 7 were already two steps that made developers go WTF. There was WPF, hot off the press and obviously very promising (but also obviously requiring a lot of polish; I don't think it was truly ready until .NET 4). And then instead of actually, you know, polishing that, like we did with WinForms, there was suddenly that new Silverlight thing, which was obviously very similar but not quite. That was when third party devs first started balking, but it really went into high gear when Windows Phone 7 guys said that their XAML will be yet another different thing. By the time we got to Win8, the developers for the platform were already allergic to all this nonsense. So I don't think it would have helped much to support WP7 apps (although it certainly wouldn't have hurt!).

Re: Microsoft hasn't had a coherent GUI strategy since Petzold

#600

Post by int_19h »

hyperrail wrote: post #504
hyperrail wrote: post #499

Earlier quoted context omitted.

(disclaimer: I was an individual engineer in the Windows division during the Windows 8 project, i.e. reporting through Steven Sinofksy) I think you're being a bit unfair to the Windows division during the Win8 lifecycle. Maybe that's just my rose-tinted glasses though. I know there are some HN/proggit commenters who like to harp on the supposed toxic rivalry between the Windows orgs and Microsoft developer tools orgs…

> I was in DevDiv during his great WinRT push and the overall feeling I remember was that the guys in Windows had zero clue as to what the devs actually wanted, but were hell bent on scorching all the ground that wasn't theirs. My team actually did some prototyping for Python/WinRT support, and we had it working to the point of the visual WPF designer in Visual Studio even. Unlike JS, it was full fledged - you could…

> In the name of winning over new or inexperienced Windows developers with "simpler, safer" projections, we in the Windows division almost completely failed for about 5 years to document or even explicitly say that WinRT was essentially just "COM: The Good Parts, Version 2012".

I'm the person who wrote this StackOverflow answer: https://stackoverflow.com/questions/7416826/how-does-windows...

You might note that I had to forcibly reiterate that, no, Win8 apps don't have to be written in HTML/JS, because people genuinely got that impression from what they saw at BUILD back then. It's not that the "COM 2.0" and the "this also works on .NET" parts were completely missing, but they (.NET especially!) were de-emphasized to the point where it genuinely created confusion and alarm among the developers. And, ss far as I can tell, this was entirely Sinofsky's idea, and one that he still refuses to admit was an epic failure.

Post reply on HN