"The browser is the new OS" ...(really?)

[zridling]

TechDirt's Timothy Lee tries to expand on this browser = OS metaphor and receives a backlash from his (English grammar-challenged) readers. While I love the cross-platform, proprietary-erasing nature of most webware, I hate the idea that my computer is only as fast as my connection speed, and that for the most part, it's a dead machine without being online.

[attach]

[justice]

The browser is not the OS but once the net connection disappears I become fed up really quick.

[f0dder]

I'm not even going to read the article, the few excerpts zridling posted were enough.

I'm really fed up with morons trying to compare browsers to operating systems - it's like comparing apples and nuclear bombs, it doesn't make sense. At all. On any level.

Sure thing, a browser can act as an application server, but that has nothing to do with an OS, really. And even if simple commodity applications can run in a browser, the browser + web isn't going to replace standalone native code anytime soon. Try doing a graphics editor or a full 3D shooter browser-style? hah.

[mouser]

Well, i agree and disagree with f0dder.

In the past, one ran applications on the local computer, on top of the computer's operating system.  The operating system is the main platform which provides the functions that the local program calls to do things (display a user interface, interact with file system, etc.)

The browser is making real inroads as a platform for running applications, for a variety of reasons.  A major one is that it has solved an issue that the operating systems makers have failed to solve -- platform independence.  The web standards have made it more or less possible to write a browser application that runs the same on mac, linux, windows.

The other appeal of applications running through web browsers instead of being coded directly for local operating system/binary executable, is the ability for such web applications to run as thin clients doing most of their work on the server backend.

For me, i view these things as much as a mark of failure for modern operating systems and programming languages as a success for browsers.

While some people have predicted that we will eventually move to all pcs just being thin-clients which are running web applications that do most of their work on large backend servers, i have a view that might be called the strong-programming-language view.  I hope (and expect) that the continuing evolution of programming languages, and a universal cross-platform user-interface api will unify many of these issues.

That is, I hope and expect that we will eventually rid ourselves of this evil of platform-incompatibility, so that the cross-platform interoperability, and distributed nature of browser-based programs, will eventually be fully supported by full fledged programming languages.  So that all applications will be essentially cross platform, 0-install, and will be designed to run both in a local mode and distributed client-server type mode.

Another way to say what i'm trying to say is that right now we use BROWSER programs that run on our local operating systems, and in the browser we load a page to serve an application.  I believe in the future that the "browser" features that make these programs attractive will evolve into a common LIBRARY used by programmers, instead of being a standalone all-purpose browser program.  So you wont open your "browser" and go to gmail to check your mail.  You will open your gmail program which could be coded in any of the new languages and would run on any pc, and would be as easy to code as it's web based alternative.

To put it another way:
I believe in the future the browser as we know it will stop being a platform for so many different applications.  I believe we will see the features of current browsers incorporated into programming APIs, and then go back to differentiating between surfing the internet using a browser, and using "applications" which will have all of the benefits we currently associate with services that run in browser windows.

[Lashiec]

While I love the cross-platform, proprietary-erasing nature of most webware, I hate the idea that my computer is only as fast as my connection speed, and that for the most part, it's a dead machine without being online.

Cross-platform... if you use Firefox, Opera or Safari users already know how cross-platform Google sites may be. Proprietary-erasing? Really? I thought webware was the ultimate proprietary system. In theory, webware should be as you describe it. In the practice not so much. Just an example, look how keyboard shortcuts or context menus in certain web applications conflict with the ones used by the browser.

An OS is an OS and a browser is a browser. You can run apps inside your browser, yeah, but I think because of things like the ones mentioned above, the browsers are already hitting a limit in their capabilities, mostly because they're duplicating functionality in a way they were not created to do.

And actually, the last paragraph in mouser's post hit the nail right in the head. Things like Mozilla's Prism, Adobe AIR, or some of the capabilities included in Qt 4.4 hint that this is the direction web applications should take in the future, platform-independent apps running as any other native application in Windows, Mac or Linux, but using a web interface and connection capabilities thanks to the underlying engine used in the framework they're running in.

Like f0dder says, certain apps belong in a true OS, and can't be replicated in a browser, or a combination like the ones above without running into problems. But a synergy of both approaches is definitely something that we already have here, and it will be easier once the aforementioned frameworks mature (which by themselves are the true definition of synergy between desktop and web apps).

[f0dder]

mouser: JAVA had the potential to do much of what you want. dotNET has the potential. Neither are going to be embraced for it, though. For a lot of applications, it simply does not make sense - sure thing, all your simple and not-so-dataheavy or precise-interaction stuff can be run "in a browser" or "on the magic platform", but as soon as user-interaction-timing critical things (games/whatever) or data-heavy tasks (graphics, video editing, ...) or "fast user feedback"/"smoothness" (drawing in a paint application) are involved, it's basically FAIL - even if everybody had low-latency high-bandwidth flat-rate fiber-optic internet connections.

And even after all these years of JIT research, we still need native code where speed is critical. OK, so dotNET can produce native code from a bytecode executable (see ngen.exe), but you still don't get native performance. And no, that doesn't really matter for a spreadsheet or word processor, but it sure as hell matters for games and video editing

Run-anywhere is a pipe dream. And even if technology matured and capable programming languages evolved (AND you convinced programmers to use them), monetary interests would block run-anywhere. It would suck anyway, basically having to code for the lowest common denominator - ugh.

I don't really mind simple-apps^TM running in a browser or whatever, my grudge in the previous post was with morons who say "operating system == browser". Application server, perhaps, but no - not OS.

Lashiec has good points as well

[mouser]

i think the evolution from java to .net is the right direction, and i think if you take a few more steps in that direction (and escape the proprietary nature of .net) you get to where i am suggesting: A [language AND hardware] neutral secure virtual platform for coding applications which has all the benefits seen in both the local binary applications and the browser-based services of the current day.

In 50 years people will not understand how it could have ever been the case that some programs ran on some brands of computers and some operating systems but not on others.  And they won't understand the concept of "local binary executables" vs "services running inside of a browser".

[note again that i don't think the browser is going to "take over" anything, rather i think that the benefits of the browser are going to be absorbed by modern programming languages and operating systems; so for me it's the concept of the browser program that will not survive as the key player in the long run (sorry firefox!)].

[justice]

When was the last time the average programmer wrote games or video editing applications? Although flash games proves that it's possible, shockwave has 3d. I think a large percentage of all programming projects would be ok to code in a virtual platform that would be easier to code towards than the desktop situation we have today.

[zridling]

Lashiec, how is webware proprietary when it uses open standards like HTML and ODF, and to a lesser extent, PDF?

By 2006, I was looking to get off the Microsoft wagon, and both open source software and open standards allowed me to. mouser mentioned how "platform independence" gave me the liberty to migrate away from Windows and stay away. Although it's a compromised substitute, at least with webware like Google Docs, I don't have to ever think about downloading, upgrading, or spending a penny on software I use mainly to write letters with.

[mouser]

I guess zaine's post makes me want to underline why i surely hope that i am right and the people who believe the browser and thin client will take over are wrong.

For me, the idea of being dependent on a corporation to host my files and my applications is something i find very unappealing.  I do not want to move to web-based applications run on some corporation's server. I'm not a huge fan of google, and i don't like this trend to hosted and ad-supported webware.

I'm looking forward to the day where we all own our own software again, but where it is as ubiquitous, cross-platform and web-enabled as current web services are, and looking forward to an end to the ad-supported web services.

One of the tricks people always accused Microsoft of employing but which is a more widespread than microsoft is called "embrace and extend (and extinguish)":

"Embrace, extend and extinguish,"[1] also known as "Embrace, extend, and exterminate,"[2] is a phrase that the U.S. Department of Justice alleged[3] was used internally by Microsoft[4] to describe their strategy for entering product categories involving widely used standards, extending those standards with proprietary capabilities, and then using those differences to disadvantage its competitors."

The trick is to lock users into using a specific product and make it hard to move away.  Well the web services have figured out this game as well, and the fact that all your data is already on their servers and not transparently available to you (for the most part) makes it even easier.  I simply do not want to be dependent on corporations to hold all my data online.  It's not that i have paranoid thoughts that they are doing bad stuff with my data, it's that i want control of my stuff.  I don't want to be dependent on some corporation's marketing strategy to decide how i can access it and whether they want to let me move my data to some other competitor's service.

[nontroppo]

I'm looking forward to the day where we all own our own software again, but where it is as ubiquitous, cross-platform and web-enabled as current web services are, and looking forward to an end to the ad-supported web services.

Well, all that is needed is a reliable framework where our home data can be accessed wherever we are via web-based apps right? Then you are in control of your data AND you can roam in the cloud. Nothing about a web app demands its data to be on the same domain it is. Or that is has to be hosted remotely as HTML5 attests to.

Though still largely speculative, here is where Roughlydrafted thinks Apple is going in this domain:

It might also come as a surprise that Apple will soon release a suite of apps that will join QuickTime, iTunes, and Safari on the Windows platform. Like Apple’s existing Windows apps, the new ones will all put the Mac OS X user interface in front of millions of new users. Additionally, they will also advance Cocoa-style development in front of a much larger audience, because Apple is also giving away the frameworks it used to create those new apps.

Another surprise is that all those apps will also run cross platform on Linux. How will Apple do this? Not by shipping a large, cross platform Yellow Box runtime for various other operating systems as it attempted to do back in 1997.

Instead, Apple is refining Cocoa for deployment within the web browser to enable developers to build those so called “Rich Internet Applications” that Adobe wants users to build in Flash/Flex/AIR, Microsoft in Silverlight, Sun in Java, and so on.

http://www.roughlydrafted.com/2008/06/14/cocoa-for-windows-flash-killer-sproutcore/

A standards-based, open source Cocoa-like dev platform, otherwise known as a YAFJF (yet another flaming javascript framework  ).

[Lashiec]

Lashiec, how is webware proprietary when it uses open standards like HTML and ODF, and to a lesser extent, PDF?

You don't have absolutely no idea about what's happening between your computer and the server, less even about what the server doing, so it's still proprietary, contrary to open source apps, where everyone can grab the code and take a peek to see how the whole app works. With webware, this is impossible, and you can only make good guesses at best. Even then, I don't know how many things you can see (the code) of what the web app is showing you via the browser (someone with more expertise than me in web development should be able to clarify this point). Webware at best qualifies like semi-open software.

Add to that what mouser talks about, total lack of control over your files, and how web apps force you to use their supported browsers. In a way, this feels like the dark years, where everything broke if you were not using IE. One could argue that you can always use another browser meanwhile, contrary to using a completely different operating system like it's the case with desktop software, but I thought the use of open standards led by the surge in market share of an open source browser should led to a open web for anyone, not only for those using the two sanctioned browsers.

While it's an advance to be able to choose (while in the past you only have one option), and Firefox is cross platform, I really hope the future brings everything together so people could use what they like to work with its web app of choice. Although if we're going the path I outlined in my previous post, this will not be necessary, and probably Trident (IE) will be out of the game :D

[mouser]

how is webware proprietary when it uses open standards like HTML and ODF, and to a lesser extent, PDF?

Do not confuse the VIEWER being open source with the backend software and data being open source.

The browser, which is the device you use to interact with the web service may be open source, but with web services this is like saying your keyboard is open source.  It tells you nothing about the software and data actually doing the computation behind the scenes.

And even though many backend web services and tools are open source (wordpress, etc.), BUT even so that doesn't tell you anything about how you can access your data when it's stored on the company services.  If you're using a web-based mail program -- your mail is all stored on their company server.  It doesn't help you that your browser may be open source, that's meaningless.  An important question is, how easy is it to get at your data and move it to another service, and are you satisfied not having complete control over your data?  I'm not saying you shouldn't be satisfied with how things are now, i'm just saying that i personally do not like it.  I want to be in control of my data.  Personally i view this as more important than the issue of open source (should we call it "Open Data"?).  I can always write a replacement program if a proprietary piece of software goes rogue.. but if i can't move my data out of a company's web server then that leaves me in a precarious situation.

[Deozaan]

For those here who say Games in a browser suck, check out the Fallen Empire Legions video.

[MrCrispy]

My university professor used to say - "Computing tends to reinvent itself every 20-25 years" and its very true. The evolution of
desktop apps to rich/smart client, SaaS, Web 3.0 etc is all just today's version of terminals connected to mainframes.

I firmly believe in it. Global connectivity IS the future.

"No man is an island" is true in the tech world as well.

The human race has to solve the problems of resource disparity and the gaps in wealth distribution, and make technology accessible to everyone. That's a much harder problem to solve than the technology itself, which is almost here today. The entire world should be living in an always on, always connected metaverse (http://en.wikipedia.org/wiki/Metaverse).

[fenixproductions]

2Deozaan

Hard to say it doesn't suck

InstantAction currently supports:
Mozilla Firefox 2
Internet Explorer 7
Internet Explorer 6
Windows XP
Windows Vista

[f0dder]

For those here who say Games in a browser suck, check out the Fallen Empire Legions video.

And what kind of plugins does it require? I doubt flash is enough