Loud

I made this here comment on Mike Cane’s post on the cloud, chromebooks and that whole malarkey. Preserved here for whatever reason.

Unfortunately, you’re making a classic extrapolation mistake here: You’re looking forward on one branch of tech while ignoring the possibilities of the other.

Namely, you’re assuming that client-side software will stand still and that cloud-tech won’t turn into a clusterfuck of mistakes and greed-driven idiocy.

Which it will.

Web apps are and will be great for certain apps, but is worse than ‘native’ apps for some things. This will continue because as web tech evolves and improves, native apps will as well. There are a lot of low-hanging fruits for improvement of the native APIs of iOS, Android, and even the Mac, and, unlike most of the tech the cloud/web industry is based on, they aren’t designed by a committee of infighting autocrats who are completely disconnected from actual, on the ground, use cases (HTML5, I’m looking at you).

So, in short: ‘Native’ apps are better than web/cloud apps at some things today, and they will continue to be better at them tomorrow, because this is a race where they’ve got a 40 year head start and they’re not giving up their lead.

Fun fact: the web is over twenty years old, we’re in the middle of the birth of a new generation of web tech (HTML5 and related standards) and we still don’t have any way of doing usable rich text editing on the web. The only thing done by HTML5 on the text editing front is standardise the crap that’s already implemented. Assuming that this will be solved is daring, to say the least.

Cloud tech also disregards some of the basic concepts of what made the internet powerful: The end-to-end principle ( http://en.wikipedia.org/wiki/End-to-end_principle ) and the Rise of the Stupid Network ( http://isen.com/stupid.html )

A dumb network of smart clients will always outperform a smart network of dumb clients.

The Cloud is also much more vulnerable to disasters, mistakes, failures, and any other sort of Black Swan event that can’t be accounted for.

So, the cloud: Good for some things, better than native for others, worse than native for most. It’ll be a big part of some corners of the industry, but don’t expect it to take the board.

I added this here comment as well:

Another point: Any piece of cloud tech is a bitch to scale to many users, costs a fortune and consumes an ungodly amount of resources. A native app that performs the same task can be made by a small team of people and only needs a dumb server to sell and distribute, scaling up to many users is relatively cheap, and you’re not tied to any piece of infrastructure or capital expenses.