Is this thread about working or talking? Surely it's hard to collaborate with people using desktop applications without an internet connection.

The point I was trying to make is that "death of desktop" is a myth. A new buzzword, invented to get more PR for whoever is NOT developing "desktop software".

While harsh reality is that there are applications and there are web applications. And we all make our money using "desktop" software every day. Moreover, lately it become accepted (and even hot) not to pay for it (apache, mongrel, linux, eclipse, ruby interpreter, PHP, photoshop, firefox, internet explorer, even freakin BIOS). And proclaiming it "dead", while being 110% dependant on it, is at least superficial.

And if you ask people why they build web apps, the answer will not be "because desktop is dead", but probably (if they are being honest) because regular software takes a lot more skill, more learning time, more development time and frankly it's just not as "sexy" these days. Sexy or not, it surely is not dead.

We depend on "desktop" so much that people often pay for it, while "web software" is almost never considered valuable enough to get people to open their wallets for. To me that's pretty pathetic: not to be able to charge for what you've built, and becoming yet another semi-useful parasite sitting on AdSense wellfare.

I actually started my career in desktop apps and moved to web "because desktop is dead". My first couple tasks projects at my current employer were Swing apps, IDE plugins, etc.

I was pulled onto web development because we basically can't sell desktop interfaces anymore. I'd prefer to do desktop software, IMHO it's easier (particularly in Java, where the web frameworks suck), but there's just no demand for it. Almost all new development at my employer uses a web front-end. Almost all of our competitors are still using thick-client desktop apps. When we go out to prospective clients and give demos, our web interface is a significant selling point.

'"web software" is almost never considered valuable enough to get people to open their wallets for'

Our customers give us close to a million dollars a pop for our webapps.

Sure, it'll be years before everything we do happens on the network...but not many years. It's just too damned convenient to have it floating in the cloud, accessible wherever you happen to be and with whatever devices you have handy.

Take 37signals app. I can't use it for work related because that would resulted in me storing the confidential work related items under their server. Work also takes much of your time.

Another example would be to store Word docs in Google. One slip-up and your word docs is searchable by Google.

So no, the desktop isn't dead yet.

In fact, I only use websites to read news and forums.

Privacy awareness and assurance will come to these services. Most small businesses already outsource email...that's as "confidential" as most businesses get.

I already know dozens of people that use gmail for everything (business and personal). It's not a large gap between that and presentations, spreadsheets, etc.

And, of course, "web" doesn't just mean "running at Google". There are numerous installable products that aren't desktop apps, but keep data within an organization.

"In fact, I only use websites to read news and forums."

This merely exhibits your own bias, not the coming reality. Sure, it'll be years, as I mentioned. But it's definitely coming. The convenience, higher efficiency (both for the provider and the consumer), and faster pace of innovation will win in the end. Someone else mentioned that desktop software is harder and that web apps can be developed by hacks...while that isn't entirely true (the apps we're generally willing to rely on were developed by seriously good developers regardless of whether it's desktop or web-based), it's another reason for web apps to win. More developers equals more apps for more specialized needs. Just like DSLs are the best way to solve many programming problems, custom apps are the best way to solve many software problems...with more developers more custom apps can flourish. And if you have a choice between a poorly fitting desktop app, or an ideally fitting web app designed exactly for your business and industry, you'll choose the web app. Again, I don't know that it's really true--all software is hard, when it comes to making it ready for the masses.

Anyway, denial is a part of the seven stages of grief. For those grieving the death of the desktop, you've probably got a couple of stages to go, and that's alright. ;-)

Reality is that there are always 3 big components to any software: code, data and runtime. You can download/cache any of the 3, in different proportions.

Now define "online app". Is it something that downloads all 3 components? Nope, since runtime is always local (browser/flash/silverlight).

My original point is that it makes alot of sense to push as much code down to customer's computer as possible, distributing the load (because it simply makes sense). The runtime will always sit at consumer's computer, because runtimes are fat, ugly and a paint to install securely.

So... basicaly it all comes to where you want to store your data.

But slogans like "desktop apps are dead" (just like popular "everything will be XML" a few years ago) are plain stupid. The application you spend most of your time in front of is a "desktop app", it's called a browser.