Gadgets, desk accessories, widgets – whatever you they were called, they were a must-have feature for various operating systems for a while. Windows in particular has tried making them happen six times, and every time, they failed to really catch on and ended up being killed, only for the company to try again a few years later.
Microsoft has been trying to solve the same UX problem since 1997: how to surface live information without making you launch an app. They’ve shipped six different implementations across nearly 30 years. Each one died from a different fundamental flaw – performance, security, screen space, privacy, engagement. And each death triggered the same reflex: containment.
↫ Pavel Osadchuk
There’s quite a few memories in this article. I never actually used Active Desktop back when it came out, because I seem to remember the channels feature was either not available in The Netherlands or the available channels were American stuff we didn’t care about. The sidebar in Vista had a lot of potential, and I did like the feature, but there weren’t a lot of great widgets and we hadn’t entered the era of omnipresent notifications begging for out attention just yet, so use cases remained elusive.
Now Metro, that’s where things came together, at least for me. I was en enthusiastic Windows Phone user – I imported two Windows Phone devices from the US to be an early adopter – and I still consider its live tiles with notifications and other useful information to be the most pleasant user interface for a mobile device, bar none. It may have taken Microsoft six tries, but they nailed it with that one, and I’m still sad the Windows Phone user interface lost out to whatever iOS and Android offered.
On desktops and laptops, though, it’s a different story, and I don’t think the Metro tiles concept ever made any sense there. Widgets as they exist in Windows now mostly seem like an annoying distraction, and I’ve never seen anyone actually use them. Does anyone even keep them enabled at all?
I tried to use Active Desktop, but it always needed a lot more RAM than I had, and was very unstable. I liked gadgets in Windows Vista, but again, they used too much memory to be turned on all the time, although it was far better than Active Desktop’s implementation. Since then, I haven’t used Windows enough on a system where I have appropriate permissions to really use the later implementations. I liked the live weather tile on the Start menu in Windows 10, it was disappointing that they killed off the feature.
I like the idea of gadgets, and I use them in KDE, but I understand why they aren’t for everyone. One of the issues I have with the concept is that they are free-form designs, with no consistency enforced. Secondly, despite the convenience, gadgets tend to be too limited in functionality. KDE’s implementation, at least in KDE Plasma 5 and 6, tend to be much less resource demanding and more flexible, which is in line with KDE generally, after all.
Angel Blue01,
I also like the ability to customize widgets. Some, like system monitoring widgets are extremely useful. I don’t use it all that much though. The main problem for me is that I multiple computers with different accounts running different desktops that don’t even support the same widgets. So I end up throwing my hands in the air and giving up on too much customization. If there was a better way to make customization persist across all of my desktops & logins, then I probably would make better use of it.
This seems like a big ask considering all the desktops I run (cinnamon, xfce, kde, and ideally gnome too).
It was always more niche than mainstream, but at the time I was a full fledged windows user and built components for active desktop. Visual Basic could easily build components to be used as desktop widgets. One I built was a multiuser painting problem that let you draw notes and sketches on the desktop and share them across other desktops on other computers. And it had the ability to play back simple animations. It was for no other reason that because I could.
ActiveX was pretty neat as a technology to embed components on the desktop, inside of documents, on web pages, etc. But it was notoriously insecure and it never recovered from the reputation for being exploitable.
The idea is still neat, but it’s obvious any implementation of it for consumer devices would require robust isolation. To have it be multiplatform would be cool. Alas, noone would agree on it.
Ohh, the lessons learned…
I remember learning the first two by trying those new features for an hour. It’s curious how they went beyond the development stage at Redmond.
I don’t remember learning the other four lessons, because… Why should I bother? I began to lose trust in Microsoft’s approach to OS development and I was also no more in my 20s and so took a more conservative approach to using computers.
Hey Satya, here’s another lesson: Everything you did since (and including) 8 is hated by power users, and they’re leaving in droves. Make 7 great again.