Backwards compatibility has always been a driving force in software. Without it, our loose-knit software ecosystem would quickly become hopelessly divided. But is it always good?
We often see the term "usability" thrown around. How usable is product X? This term might as well be a synonym for familiarity: Does product X work like something its users are already familiar with? Point-and-click was revolutionary in 1984. Most people then weren't familiar with computers, but they were (of course) familiar with moving physical objects around! Point-and-click was a perfect physical analog to jump-start the mental computing model.
But desktop computing has always had a not-so-hidden expert mode. Want to input text? Use the keyboard. Want to speed up a repetitive task? Learn keyboard shortcuts. The venerable vi text editor (1976) and its brethren even contain an ingenious system of modal keyboard input (we are fans of spacemacs!). i3 (and a long history of tiling window managers) apply similar keyboard-centric concepts to the desktop. But this stuff is for "experts"! What about the everyman?
Fast-forward to today, and touchscreen devices are ubiquitous. Rare is the mobile device with a physical keyboard (for good reason), so those of us that want to do real work in such a constrained setting are stuck in fat-finger purgatory. In short, point-and-click was great in 1984, but we've still got the training wheels on over 30 years later.
We at memwris are sick of dragging boxes around and ticking drop-down menus. Swiping is just a special case of the same old. Our interfaces are hopelessly cluttered with different-but-similar widgets that are often difficult to distinguish from static content. We think all this is ridiculously inefficient, but people put up with it because it is familiar. In our opinion, the modern user experience has evolved more out of fear than pragmatism. Fear that interfaces might confuse new users; fear of the unfamiliar; fear for the bottom line.
Then there are the band-aids. Rather than improve on expressing our intents accurately and efficiently, we use fancy techniques like predictive text, or voice commands, or smart error-correction. These work great, except when they don't. They certainly don't work better than getting it right in the first place, because if they did, the keyboard would be a relic of the past. For you programmers out there, would you ever consider seriously programming on your mobile device? We think it's possible, but not with point-and-click.
And sure, there are the trailblazers dreaming up new VR, wearables, etc., and we think that stuff is exciting! But as of yet this futurism doesn't really translate to everyday computing. We want to fill that gap. In short, we think there is a time and a place for making breaking changes. For UI, that time is long past due.