So...Android? I mean, I get there's a lot of differences, and Android, the platform, would have to be much more open to begin to compete with the web. But, it ticks off an awful lot of the boxes he's named. If one could run any app in the world on Android without installing it and without significant delay, it'd be a reasonable substitute for the web, I guess (which it can, because it has a web browser, but we're back to the previously listed failings of the web).
Of course, maybe one can argue that openness is the killer feature of the web and unless Google really embraces open, it'll never surpass the web.
But, I still don't buy it. I was critical of the first article, and I'm maybe even more convinced now that the web won't be beaten by a "better 90s", which seems to be what's being proposed, in many regards. The author seems to want to reset the clock just before the web exploded, and build something new from where we were in about 1994, but with a bunch of lessons learned from the web.
Some specific areas of contention I have:
1. IDE-oriented is a terrible idea, IMHO. IDEs are fine, I guess, but they always strike me as being indicative of insufficient/incorrect abstraction, rather than being a great productivity booster.
2. I think designing UIs for the web is today better than for desktop apps. Maybe this is indicative of my lack of experience with desktop apps, but even when I've tinkered with Android development, I found UI builders to be tedious and frustrating, especially when it comes to making them responsive. With Flexbox and Grid, and a component library, I can whip up a nice, responsive UI in minutes for the web...and I'm not even good at it!
3. On data representation and binary formats, I tend to think everybody settling on JSON is Good Enough. It's slightly sub-optimal, but it's standardized on the front and back ends, it is universal, and it can be compressed using standard universally available tools. To me, it looks like a classic "Worse is Better" scenario. Anything else will take years to work its way into common usage.
Which brings me to the biggest issue: By the time a new platform reaches even a tiny fraction of the reach of the web, the web will have likely caught up. It is very difficult to overstate how fast the web as a platform is moving today. No single developer can fathom how advanced a lot of the tools have become just in the past couple of years. I've been sort of giving myself a crash course in Node/React/etc. lately, and it's, frankly, astonishing. You can go down any of dozens of rabbit holes and find incredibly powerful tools in all sorts of domains. Building apps for the web is stupidly easy once you get over the (admittedly large) learning curve for putting all the pieces together and overcome analysis paralysis and stop reading about every new library and framework.
I guess I remain unconvinced. I come away feeling the same optimism I had for the web before reading (maybe even more, because for nearly every point he makes, I can see a clear path where someone is already working on solving the problem for the web, or it already exists in a rudimentary form), and the same strong doubt that anything can beat the web short of being the result of the continuing evolution of the web.
Of course, maybe one can argue that openness is the killer feature of the web and unless Google really embraces open, it'll never surpass the web.
But, I still don't buy it. I was critical of the first article, and I'm maybe even more convinced now that the web won't be beaten by a "better 90s", which seems to be what's being proposed, in many regards. The author seems to want to reset the clock just before the web exploded, and build something new from where we were in about 1994, but with a bunch of lessons learned from the web.
Some specific areas of contention I have:
1. IDE-oriented is a terrible idea, IMHO. IDEs are fine, I guess, but they always strike me as being indicative of insufficient/incorrect abstraction, rather than being a great productivity booster.
2. I think designing UIs for the web is today better than for desktop apps. Maybe this is indicative of my lack of experience with desktop apps, but even when I've tinkered with Android development, I found UI builders to be tedious and frustrating, especially when it comes to making them responsive. With Flexbox and Grid, and a component library, I can whip up a nice, responsive UI in minutes for the web...and I'm not even good at it!
3. On data representation and binary formats, I tend to think everybody settling on JSON is Good Enough. It's slightly sub-optimal, but it's standardized on the front and back ends, it is universal, and it can be compressed using standard universally available tools. To me, it looks like a classic "Worse is Better" scenario. Anything else will take years to work its way into common usage.
Which brings me to the biggest issue: By the time a new platform reaches even a tiny fraction of the reach of the web, the web will have likely caught up. It is very difficult to overstate how fast the web as a platform is moving today. No single developer can fathom how advanced a lot of the tools have become just in the past couple of years. I've been sort of giving myself a crash course in Node/React/etc. lately, and it's, frankly, astonishing. You can go down any of dozens of rabbit holes and find incredibly powerful tools in all sorts of domains. Building apps for the web is stupidly easy once you get over the (admittedly large) learning curve for putting all the pieces together and overcome analysis paralysis and stop reading about every new library and framework.
I guess I remain unconvinced. I come away feeling the same optimism I had for the web before reading (maybe even more, because for nearly every point he makes, I can see a clear path where someone is already working on solving the problem for the web, or it already exists in a rudimentary form), and the same strong doubt that anything can beat the web short of being the result of the continuing evolution of the web.