I won't argue that the Internet is not important as a communication medium. I think that's another point all together. It is very effective at communicating information.

My point was that the interface to that transmission medium could be better. What we have instead of a phone receiver is something more akin to a rube goldberg mechanical turk. The message you were intended to receive might not be what you actually receive and depending on what kind of mechanical turk you have on your end; it may crash or behave unexpectedly.

Of course, the core functionality is all still there: data can be stored and shared on publicly accessible servers. Stateful and stateless protocols can be built to interact with that data. Just because you don't load up a web browser with a tonne of HTML/CSS/JS just to view which friends are online doesn't mean it's not a social communications platform anymore.

that integration is an effect of data being stored on publicly accessible servers and is not a virtue of "web applications" alone. I agree that the Internet should be more ubiquitous if possible, I just believe that the technology stack we've chosen to implement it is feable and silly.

But.. but... I don't know how to program anything else! The web is the future! FUTURE!

In all seriousness, I basically feel exactly the same way. I've been building 'web applications' for companies for years because that's all they're hiring people for. It sometimes surprises me that it ever works at all. The sheer number of brittle components all hobbled together... there are so many weak points where something can go wrong. It just makes for one big headache after another. X11 is a server and has been delivering stateful GUIs across the network since the early nineties at least! It amazes me, the amount of technology we have today, and what we've chosen to do with it. It could have been so much more, but instead the worst possible solution won out the day... and now a whole generation of developers have no exposure to anything else.

Is everyone seriously impressed that we're creating stateless GUIs to remote applications by scripting marked-up text inside increasingly bloated and resource-hogging third-party applications? Is this the future? Really?

I'm with you on this one.

It's easy to make that claim. Especially if what they deliver really is tailored to netbooks. No need to ship it with a gargantuan kernel full of drivers. It might be smaller, faster, more secure because it literally is smaller and thus supports a smaller range of hardware.

Amen.

It does have it's own challenges (being somewhat difficult to configure on its own for non-technical users), but the flexibility it affords is awesome.

... and it's a server too! Maybe if we'd spent the last 15 years working on a standard X11-like network protocol instead of hacking stateless application GUIs out of scripted marked-up text, we'd have a more useful Internet than we do now. But I digress.

Loves me X, I do! :)

amen.

To this day I have a hard time convincing people to write unit tests, to adopt continuous integration, or to even automate their build and deployment processes.

If you want to build things the right way, it sometimes feels like an uphill battle! There have been many times in my career where I've had to defend my decisions to let a new feature slide in favor of refactoring some important bugs out of a critical component of the system. It never ceased to amaze me when it happened. Were they happy to release a new feature and then fume at the meetings when it was buggy and full of errors?

I think a large part of the problem is that management that doesn't understand what it is we do are ineffective in the long run. They are essentially making decisions based on faith. If they trust their developers, they have no way of validating their trust in them and must base their decisions on whatever they're told. How is that an effective process?

Adding more process onto the programmers isn't going to fix the problem.

The only effective managers at a technology company are those who started off as technical staff. Even if they weren't the best programmers on the team, if they were decent they'd know who those people are and be able to recognize good advice.

It's really not as primitive as you make it out to be.

I've worked on projects big and small, but no matter the size emacs fits the bill for me. I actually find that it's a rather elegant tool.

Anyway, I'm long past my days of flaming people over IDEs and languages. I think you had the right idea in the beginning, but you lost me with your closing line.

Assholes, exactly -- people who are not genius' can act this way too. I've met countless sales people and executives who've had sexual allegations against them and who've siphoned off company money to fund affluent lifestyles; act like complete pricks to everyone they meet; and generally be very "Josh" like. Yet they seem to lack one thing: intelligence.

I can understand a lot of the "devil's advocate" positions; but the reality is that this editorial is supporting a straw-man argument.

A genius developer isn't universally predisposed to defecating in public and verbally abusing people.

They may be eccentric and the scale of their eccentricities vs. their practical value to society will certainly determine how far they go. More often than not, such people will either disappear from corporate life or else end up running it. The worst thing possible for an eccentric genius developer is probably being stuck as a "cog in the machine." They'd probably be more likely to flourish in research, leadership, or academic positions.