I think the author is wrong to dismiss platform HIGs (or equivalents) so quickly. I'd argue that even the fundamentals are platform specific.
I also think that the situation is not as dire as the author posits wrt tooling. The world is kinda moving away from classic widget based UIs, Qt is moving to QML, on Windows there is XAML, and generally Electron-style solutions seem to increase in popularity and capability (personally I have hopes that Servo could fill a hole there).
After 40 years since PARC, perhaps the issue is a weakness in the abstractions of WIMP. The base assumption for WIMP is that the user cannot type...the reference hardware at PARC was a tablet like device for children, the first Apple GUI's were on the business executive targeted Lisa platform.
The ubiquity of computers has rendered that assumption to the midden alongside buggy whips and typing pools. Using networks requires typing: even appstore apps require usernames and passwords even when typing web addresses isn't required. WIMP was designed for a time when to a first approximation all new users were new computer users. That's just not the case anymore.
4 comments
[ 4.4 ms ] story [ 14.5 ms ] threadI also think that the situation is not as dire as the author posits wrt tooling. The world is kinda moving away from classic widget based UIs, Qt is moving to QML, on Windows there is XAML, and generally Electron-style solutions seem to increase in popularity and capability (personally I have hopes that Servo could fill a hole there).
The ubiquity of computers has rendered that assumption to the midden alongside buggy whips and typing pools. Using networks requires typing: even appstore apps require usernames and passwords even when typing web addresses isn't required. WIMP was designed for a time when to a first approximation all new users were new computer users. That's just not the case anymore.
most comments on this one: https://news.ycombinator.com/item?id=10823735