That sounds like "3d games in Windows?" in 1995, or any other "$perf_sucking_app in $new_environment".

The "browser" that's going to run Photoshop 2015 won't be Netscape 3. It'll have a JIT'd JS engine (so it'll be really fast), it'll have local storage (so you don't need to upload anything), it'll have efficient graphics libraries (so it can use your video card), it'll have native-feeling controls, and so on.

It's a "browser" only in that it's a continuous upgrade from our beloved Netscape, but it won't look much like Netscape at all. In terms of user experience, it'll be basically like the Photoshop you know and love, except without the adventure that is the Adobe installer.

Now I am intrigued: an OS that is its own browser. Not eyeOS and the like: I mean the window manager is the browser. You browse a VFS which is mounted, locally or remotely; the browser doesn't see a difference. XUL is the default UI language. The browser handles threading, even. You have no desktop but you have a homepage. Exit the browser, power down.

Am I talking about something that exists? Stripped-down single-app systems like kiosks don't count... or should they count?

To me too. I never understood, ten years ago, why on earth most people would prefer a web-based email client to a native desktop app but they did. And today that percentage is just increasing. Most people just don't have the same concerns you and I do.

>>Their victory is so complete that I'm now surprised when I come across a computer running Windows.

Yeah, this one didn't make sense. If you're surprised when you come across a computer running the desktop OS used by more than 75% ([citation needed]) of the populace, then you clearly move in very narrow circles.

For me it had more to do with the fact that GMail's minimalist UI ended up being more responsive than the likes of Outlook.

"Yeah, this one didn't make sense. If you're surprised when you come across a computer running the desktop OS used by more than 75% ([citation needed]) of the populace, then you clearly move in very narrow circles."

I'm fairly certain that it's still more than 75%. It's also very likely that Dell's monthly sales still exceed Apple's annual sales of computers. Apple's sales have been higher than usual, but still hasn't managed to make much of a dent in the Windows market -- and don't forget that mac users are buying Windows licences so that they can run Windows-only software also... so even though those are probably the minority of mac buyers, it does suggest that Apple's non-victory is even less complete than PG is implying. :)

What you seem to fail to see is that you're an outlier.

Most people want photoshop through a browser. Most people already do most of their stuff through a browser. It's funny to even debate, because it's happening/happened for large amounts of people out there already. Ask your parents what they do on the computer, and it's likely to be 90%+ web, with perhaps a couple of desktop apps. If that.

Ironically, it's the techie usually early adopters that are lagging here, preferring to stick with desktop software. The people leading the webapp charge are people who don't want the fuss of downloading desktop software, with potential spyware etc, and just go to google and find a webapp instead.

The same phenomenon exists when some people discuss advertising. "No one clicks on ads anymore" - it's easy to say, if you install adblock, or just don't click on ads, but this is a total minority. The majority do click on ads. Just like the majority want, and use webapps over desktop apps.

>> "Gmail became popular not because it employed AJAX or was based online, but because it solved a hair-on-fire problem: tiny inbox sizes."

I think actually the reason it succeeded was because it had ridiculously fast searching, and unlike a desktop app, you don't have to worry about your storage. I don't think the majority of people really cared about inbox size. The fact you could easily search for things in a simple and blazingly fast webapp was key. For most people, it is better than a desktop app.

>> "This statement I also find ridiculous. If anything, increased broadband availability has made computers even more of a must-have product."

The fact that broadband access has increased etc means there's less need to have a thick client. Where previously you needed to download a video, click on it, watch it, we all now just watch videos streamed to us by a webapp. When we all have 1Gbps connections, why would we need a full powerful computer each? Why would it not be in most persons better interest to just have a very thin power conserving client that connects up to the big servers doing the grunt work.

The fact is, it's happening, and will continue to happen. You can't stop it. It's inevitable. You might prefer to stick with your computer and desktop apps, but you'll be a small minority, like the people who still buy records on vinyl.

>> "The revolution in web apps was making them behave more like desktop apps."

I disagree with that. The revolution in webapps is making them work BETTER than desktop apps. Gmail can search my mail much faster than my computer can. Mibbit can do things impossible with desktop IRC clients. etc etc

The reason Microsoft are 'dead' is that they don't get it. They're dead in a similar sense to the music industry IMHO - the market changed around them and they completely failed to adapt.

When PG says stuff like "I'm now surprised when I come across a computer running Windows", I think he's more of an outlier than the parent poster.

Gmail can search my mail much faster than my computer can.

There's no technological reason for this to be true. Gmail takes a second or two to answer search queries over my mail; that is plenty of time for a typical desktop machine to process full-text search queries over a few GB of indexed and largely static data.

The reason Microsoft are 'dead' is that they don't get it.

I think most of my objection to the original PG article is the shameless hyperbole it engages in. Microsoft isn't "dead" by any stretch of the imagination -- they are poorly positioned for cloud computing, but the game is certainly not over yet. Claiming that Microsoft is "no longer a factor one has to consider when doing something in technology" is simply wrong -- even if you narrow your focus to consumer-facing web apps, Microsoft still provides the dominant client platform (IE), and a very popular backend infrastructure (SQL + ASP.NET + CLR + ...).

Different kind of outlier though: when you're an outlier because you spend an inordinate amount of time among people creating new technology, then your anomalous experience makes you better at predicting the future, not worse.

Sometimes, but not always. If you spent all your time hanging out with people creating new technology in the mid 1980s, you might have been convinced that Lisp machines would soon become the dominant computing platform.

For most startups I don't think they give ms backend a second thought. It's a no brainer to use free open source.

Also, amongst webapps, IE is no longer the dominant client platform at all in my experience.

I too do not want Photoshop in a browser. I shoot Canon 5D and 5DMkII. My RAW images are HUGE. I'd either have to upload them to the server to "edit", or download a fairly hefty app to run locally. Photoshop (true photoshop) is processor and bandwidth intensive when used as intended. It won't live in the cloud for a very long time.

Now, what my parents and similar folks want is a way to reduce red-eye, crop and convert photos and apply some simple effects. Sure, THAT can be a web-app. But, the cloud is not killing photoshop.

I also personally do a lot of business travel. Until planes have free, good wifi, then "cloud apps" suck for me. I want to be able to read emails and queue responses while in-flight. I want to be able to edit and organize docs and presentations. I want to be able to look up an address or phone number from my inbox QUICKLY, without having to establish a wifi connection, accept some TOS for the free access, go to gmail, find the message, etc. Just pop open my MBP, give it 8 seconds to "wakeup", go to mail.app and pull the message I want.

In a few years time those will likely take less than a second to upload though.

I agree, video editing, large photo editing will take longer to move over.

>> "Until planes have free, good wifi"

It'll certainly happen, it's just a matter of how quickly. I'd bet maybe 5 years :/

Some things are simply not meant to be done in the cloud. I think the right answer is to not fight the cloud OR the desktop appropriate applications. Developers should realize that for many markets, they may have to support both models to gain significant market share. Most companies have gained marketshare by customers what they want instead of forcing a particular model upon them...

Thin clients too- I don't think you can honestly expect that. The difference between using google docs and a native application is night and day. Docs is slow, prone to pauses, and has lots of bugs related to the poor mapping between html and page layout. Either the browser is going to get a lot faster, we're going to need /thicker/ clients, or people will stay with the status quo:

writing their documents in word, copying and pasting them into google docs only when they need an easy wiki.

If anything, Internet has been getting slower and slower due to congestion and server overloads. Upload speeds still haven't reached 100Kb/sec for 90% of the population after 10 years of affordable broadband arrival.

Pushing everything into a centralized location makes absolutely no sense. Not every byte on my hard drive needs to be shared with other people.

Instead we'll be moving towards rich browser runtimes, to allow on-the-fly downloaded&installed code to effectively work with local data.

Anybody who does want Photoshop does not want it in a browser.

The reason Microsoft is in much less of a leadership position is that there are a lot more people who want Picassa than Photoshop. That doesn't mean desktop software is dead (the people who want Photoshop pay handily for it) but it does mean they're no longer as important as they were.

We merely settled on a standardized approach to networked GUIs. Instead of proliferation of 3rd party networking apps like usenet/email clients, ftp clients, various VB/Oracle business forms apps, people simply use HTTP-based UI stack. Frankly, not that big of a deal. People _feel_ like that's a big deal because this opened networking to the masses, presented a lot of business opportunities, but fundamentally, this hasn't changed anything.

The type of applications people are running hasn't changed. If your data needs to be shared with others, you need an application with a network access and it means a web site. If not - why does it need to be in a browser?

Yes, people waste time online a lot, that only means they aren't really using their computers. It's like watching TV: most of them have CPUs and RAM and stuff,yet you aren't seeing "TV evangelists" screaming that new era of computing has arrived: with 50" screens and remotes instead of keyboards. That, however, is happening with the iPhone: look how excited everybody is about "desktop apps" you can run on it.

Paul is right: new website founders have no reason to fear Microsoft because they aren't competing with them. But that's not new: it is exactly like starting a landscaping or dog walking business in the 90s: Microsoft couldn't care less.

If anything, they have won. They have such a strong grip on Windows platform that it takes big balls to start a company in that space. My deepest respect to Xobni for taking on the challenge.

I disagree though, the 'windows platform' is irrelevant for most people for the reasons in the article. Most people don't care what OS it is, as long as it has a web browser.

The 'windows platform' is about as relevant as the computers BIOS is - it's there to allow them to start a web browser.

I work at a web startup myself, we sell SaaS to businesses and, unfortunately, they deeply care about their Windows-dependent data,

Companies move ridiculously slowly - one of the reasons for there still being IE6 usage out there. So obviously they are going to stick with Microsoft/desktop apps etc, but the general population won't.

The advantages of browser apps - standard and ubiquitous deployment target, potential use of cloud - are all improving over time.

The disadvantages of browser apps are all in factors like bandwidth and connectivity - factors which are steadily giving way to the advance of technology.

I'm betting on the future of browser-based apps. PG was right.

http://dev.w3.org/html5/spec/Overview.html

>> you'll be a small minority, like the people who still buy records on vinyl.

You got me there, I totally do buy my records on vinyl.

Cheers!