Linked by Kroc Camen on Wed 6th Oct 2010 16:37 UTC
Editorial In response to Jean-Louis Glassee's article "The OS Doesn't Matter..." I wrote quite simply: the future of the browser wars is he who integrates with the OS best. This phrase came from my article lambasting Microsoft's use of HTML for their IE9 jump lists, which caused quite a stir. In the wave of ever increasing web browser capability, the operating system is going to matter to web users more than it ever has before.
Order by: Score:
Os course the OS matters!
by Sodki on Wed 6th Oct 2010 17:21 UTC
Sodki
Member since:
2005-11-10

If the OS didn't matter, then there would be no OS wars. Windows, GNU/Linux, MacOS, none of it would matter and everyone could switch to GNU/Linux in a blink of an eye. But that doesn't happen. And people still buy Apple hardware because of MacOS. The OS will always matter.

Reply Score: 3

RE: Os course the OS matters!
by umccullough on Wed 6th Oct 2010 17:49 UTC in reply to "Os course the OS matters!"
umccullough Member since:
2006-01-26

Edit: Oops, didn't meant to reply to the first comment

I think what JLG was really trying to say in his article is that the underlying kernel doesn't matter any more... and that everything hinges on the UI toolkit and above.

As he pointed out in his article, OSX and iOS are pretty much the same underlying OS - and essentially UNIX-compliant, similar to Linux, Android, WebOS, etc.

He then goes on to point out that Windows is not UNIX, and thus the "other option".

In short, the only real differentiator any more is the UI/User Experience.

Edited 2010-10-06 17:59 UTC

Reply Score: 5

RE[2]: Os course the OS matters!
by Kroc on Wed 6th Oct 2010 19:02 UTC in reply to "RE: Os course the OS matters!"
Kroc Member since:
2005-11-10

I don’t believe that’s the case--that’s just giving up on innovation. There is no way we have run out of ideas on how to get a circuit to do stuff. The more everybody accepts that they can’t change what is simply a given, the more room there is for the upstart to try something new.

Reply Score: 1

RE[3]: Os course the OS matters!
by Zifre on Wed 6th Oct 2010 21:39 UTC in reply to "RE[2]: Os course the OS matters!"
Zifre Member since:
2009-10-04

I don’t believe that’s the case--that’s just giving up on innovation. There is no way we have run out of ideas on how to get a circuit to do stuff. The more everybody accepts that they can’t change what is simply a given, the more room there is for the upstart to try something new.

The fact that this is 2010 and we still don't have widespread transactional file systems or metadata indexing is perfect proof that there is more to be done on the low level parts.

Though ultimately, the what the user sees is all that matters. The point is, a bad base OS will make it impossible to have a perfect user experience.

Reply Score: 5

Eddyspeeder Member since:
2006-05-10

Well I think Urias has a point; the general tenure of your article feels as though you have only read JLG's title, disagreed with that statement, and started to write an article focused solely on that statement, without stopping to discover that the article really is about all OSes turning into variations on UNIX or having UNIX at their cores, except for one brave little village of Gauls resisting UNIX from invading on their territory.

Even in your comment just now, it feels to me like you haven't even considered JLG's arguments. He speaks of various expressions of the UNIX core, be it Chrome OS or the Mac OS, Android or iOS. How are they "giving up on innovation"?!?! To disprove your point, let me also go back in time; the rise of DOS can be explained by its low pricing, but using the UNIX philosophy is a conscious choice by all companies who decided to go for their own Linux distribution, for creating an OS from scratch with UNIX underpinnings (BeOS), or for replacing their existing OS with a UNIX-based one (NeXT, and various smartphones). All the way up to the highly praised QNX, innovation is achieved through a single foundation all companies (except one) have come to agree upon, sometimes after years of deliberation: UNIX. That is not giving up on innovation; it is acknowledging the wheel has already been invented and you can create any kind of vehicle or machinery around said wheel.

Will we witness a sea change in the way we interact with the Web through our OSes? To an extent, yes I believe we will. Solely for that reason, Apple cannot afford to lag behind, which is also why I did partially disagree with your "Apple won't embrace the Web" article ( http://www.osnews.com/permalink?427209 ) as I myself am more optimistic about this age of Web competition preventing a return of the days of IE6. In fact, that is why I hope the OS integration with the Web, which you speak of, will happen on a large scale. And I do thank you for pointing that out in this article.

Reply Score: 3

RE: Os course the OS matters!
by Lennie on Wed 6th Oct 2010 20:35 UTC in reply to "Os course the OS matters!"
Lennie Member since:
2007-09-22

The OS itself just matters to a small number of people.

Most people just want the applications (or websites in this case) to (just) work.

Only fan boys (I'm on the side of Linux myself) and the creators of the OS itself really care.

Reply Score: 2

RE[2]: Os course the OS matters!
by Kroc on Wed 6th Oct 2010 20:44 UTC in reply to "RE: Os course the OS matters!"
Kroc Member since:
2005-11-10

People care that when they switch their computer on it starts quickly, doesn’t randomly decide to stop working and that when they click things they happen in a responsive way. The OS matters in all the ways the consumer is not aware are happening under the hood. They don’t want to be confused by meaningless dialogues, they don’t want their machine to bombard them with interruptions that get in the way of what they want to do. It is the OSes job to make the operation of the computer easy and enjoyable and matters a whole lot.

If the OS didn’t matter, you’d be typing in binary code by hand.

Reply Score: 1

UltraZelda64 Member since:
2006-12-05

If the OS didn’t matter, you’d be typing in binary code by hand.

Hey, that'd make keyboards less complex and cheaper. The only keys you would need would be 0, 1, Backspace, Delete, Enter, a few directional (arrow) keys, and maybe Escape. Maybe throw in a Ctrl and Alt key as added value so we can use our Delete key to its fullest potential. ;)

It might not be very pleasant to use, though. ;)

Reply Score: 3

Windows is not the only one
by shmerl on Wed 6th Oct 2010 17:46 UTC
shmerl
Member since:
2010-06-08

OS matters because of the OS, not because of the browser only. Windows will not gain much advantage with hardware acceleration - it is available on all modern OSes, including Linux and MacOSX.

BTW, Firefox team is also focused on hardware acceleration, see for example http://hacks.mozilla.org/2010/09/hardware-acceleration/

Reply Score: 2

RE: Windows is not the only one
by Kroc on Wed 6th Oct 2010 17:52 UTC in reply to "Windows is not the only one"
Kroc Member since:
2005-11-10

The mention of hardware acceleration is simply that Microsoft are making best use of Windows API, without that necessitating IE-only websites—a good thing.

Mozilla I think are going to face some real challenges in the near future. They haven’t understood native behaviour as well as the other browsers. Safari and Camino behave *far* more natively on the Mac, and this does matter in the small details. Just simple things like using the system-wide dictionary, auto-correct and system services.

The more a browser can behave native to the operating system, the more this is going to matter as HTML starts to touch the hardware.

Reply Score: 5

ssokolow Member since:
2010-01-21

Agreed... though I'll be curious to see what ends up happening with notifications on Linux. Firefox 3.6 still hasn't merged the patch Ubuntu uses to send its notifications via libnotify and I'm not sure about Firefox 4.

Perhaps more interesting, when they implemented the notifications for web apps, Google specifically justified implementing their own notification system with "OS-native mechanisms like libnotify aren't powerful enough." ...which I can understand as a developer (especially given Ubuntu's efforts to design a more crippled notification daemon), but I'm not sure I can excuse as an end user.

Reply Score: 1

theosib Member since:
2006-03-02

Well, even Camino isn't completely native. It integrates fairly well, but it too suffers from the long-standing firefox bug that prevents a Mac from going to sleep if you leave it idle. This is why I stick to Safari, because it doesn't have this problem. With Safari, I can set my MacBook Pro down and forget about it, and it'll go to sleep automatically. With Firefox and Camino, it'll run the battery down. (And I have one of the 2007-2008 MPBs that ruins batteries, so mine just suddenly powers off at some random point when it gets below 20%, and I lose work.)

Reply Score: 2

RE[2]: Windows is not the only one
by lemur2 on Thu 7th Oct 2010 05:41 UTC in reply to "RE: Windows is not the only one"
lemur2 Member since:
2007-02-17

The mention of hardware acceleration is simply that Microsoft are making best use of Windows API, without that necessitating IE-only websites—a good thing. Mozilla I think are going to face some real challenges in the near future. They haven’t understood native behaviour as well as the other browsers. Safari and Camino behave *far* more natively on the Mac, and this does matter in the small details. Just simple things like using the system-wide dictionary, auto-correct and system services. The more a browser can behave native to the operating system, the more this is going to matter as HTML starts to touch the hardware.


I didn't get your point at first. Firefox 4.0 has just as good hardware acceleration as IE9, (when it include Jaegermonkey) at least as good Javascript if not faster,

http://www.mozilla.com/en-US/firefox/beta/technology/#feature-perfo...

slightly better standards compliance and better support for emerging standards such as CSS3, and equivalent features such as Panorama and Apps Tab.

http://www.mozilla.com/en-US/firefox/beta/features/#feature-apps-ta...

But Firefox has all kinds of extra features as well

It has WebGL and Animated PNGs as graphics features:
http://www.mozilla.com/en-US/firefox/beta/technology/#feature-graph...

It has geolocation, Orientation and Multitouch support
http://www.mozilla.com/en-US/firefox/beta/technology/#feature-devic...

In addition it has behind-the-scenes features such as web worker threads. As always, Firefox will still have XUL, giving it the largest extensions library of any browser.

Personally, I am of the view that other browsers on any OS will have trouble matching the features and performance of Firefox 4.0.

Reply Score: 2

Kroc Member since:
2005-11-10

But at the same time Mozilla are really struggling to make a good web browser on mobile devices because it doesn’t ‘fit’. They practically don’t exist in that market. N900, and that’s it. The Android port thus far has been horribly slow and clunky. OS fit and finish matters here, where before XUL could get away with lacking polish.

Reply Score: 1

o.O
by rafaelluik on Wed 6th Oct 2010 17:50 UTC
rafaelluik
Member since:
2010-10-06

I can agree OS integration is a key for success, but I can't agree it's the primary key. I'm pretty sure some web browsers offers more features than others, and that's what matters for me, these features are present in any OS and that's what I like.

We can conclude unique features are the way of getting users, but the fact is actually web browsers are used because of their company ads or because some are bundled to the OS from factory and many people don't realize others exists.

Reply Score: 2

time to rethink the OS then
by aeischeid on Wed 6th Oct 2010 17:52 UTC
aeischeid
Member since:
2009-05-25

Chromium OS, or Peppermint Linux both have interesting approaches in that they the see web browser as central. I don't think they diminish the importance of the OS, they instead seem to aim for a more integrated system. The OS still matters, but it can be less visible. Instead of building a browser to integrate with the OS they build the OS around the browser.

Reply Score: 2

But it is declining in relevance
by kaelodest on Wed 6th Oct 2010 18:40 UTC
kaelodest
Member since:
2006-02-12

I mean to people like us who go to the more technically inclined sites we sort of care about the OS but not that much. If you will succeed in one then you will likely succeed in any environment. But for the consumer I think that they are reacting indifferently to an OS that has failed them and I can qualify that with the question of how many of us have experienced or heard of a windows horror story.
What I see is that people in general want the OS to be helpful and stable and transparent. So that they can get on the Web and watch a video and send some emails. The OS is not something that they want to think about, that is what engineers and technicians do, there is a plain assertion that there is only the two principal OS environments Windows and Unix\Unix-Like. And with the lack of innovation in games, where the market has chosen games that are fun (the Wii) as opposed to expensive but pretty games with low replay value(ridiculous shooters). In the same way that I do not want to think about what embedded SW is in my fuel injectors, I just want to drive. I am not trying to squeeze 15 HP here or there with polished headers or a cold-air intake, but to a car nut that matters. Windows or Mac or Linux, Firefox or Chrome do not matter to most of my family most of the time or even most of my clients and I am ashamed to say this but even some of my tech friends. So I do not think that the OS is dead per se. It will always be there but what I do see is an increasing reliance on the browser. At work many tasks are run on a citrix session in a browser, or as a web app front end to a data store somewhere else. The reasons and benefits are clear you patch an web app in one place or upgrade and add features to all of the users at once.
Look at the plummeting market share of MSIE and that is what only runs on windows. That is where the relevance of windows has gone. That OS and work stay in the office. Because of the emergence of the MacOS in the VIP space, weather they need it or not - if it is just a prestige/vanity item the web app still has to work. Or to reduce two paragraphs into a simple way of looking at it is, if you are not chasing specs and numbers - fastest CPU burliest GPU biggest HDD then all I want is an OS that works, keeps working and stays out of my way. Windows has failed in that space.
Facebook and twitter and who knows what comes next these are what people think is important. And in that context any decent browser should be able to take on the world. That this browser is not Microsoft Internet Explorer, well that is just icing on the cake.

Reply Score: 2

deathshadow Member since:
2005-07-12

How can it (IE) be declining in relevance when IE has MORE users today than it had when it was at 99% of the market.

DO NOT let market share be used to lie to you!!!

Let's rub a couple brain cells together and go over what share really means. It comes down to a question people don't ask when they hear a percentage and should; "A percentage of WHAT?"

Let's make this easy and compare 2005 to 2008. We have pretty concrete numbers to play with in that range. Besides, it's the most up to date numbers available on google public data, and the wikipedia usage article.

In 2005 IE had roughly 90% of the market still. If you take the time to look at percentage of the world population that was online at the time you find out that 15.9% was online. The population was then 6.46 billion, so that's 1.027 billion people online.

In 2008 IE had roughly 70% of the market... at the same time the percent of population online had grown to 23.9% and the world population had grown to 6.69 billion, so that's 1.59 billion.

90% of 1.027 billion is 924.3 million.
70% of 1.59 billion is 1.113 billion

So while "losing" 20% market share, the number of IE users grew by 189 million.

Current guesstimates put the % of world population online at 34% (given the steady trend of increasing penetration since 2005) and the world population at 6.873 billion, for 2.33 billion.

51.34% (wikipedia's number) of 2.33 billion is 1.19 billion.

Meaning that while dropping from 70% to 51.34% IE gained 77 million new users!!!

Meaning IE is just as relevant as it was five years ago or even ten years ago!

You figure in things like FF's prefetch artificially inflating it's market share, users smart enough to use other browsers getting counted more than once because of all the different devices they use to get online (like me since I get counted for Opera four times - work, home, road laptop, bedside laptop), and those percentages become even shakier.

Basically, most of the people who were using IE five years ago are still using it today... and more people are using it than ever!

DO NO LET SHARE PERCENTAGES BE USED TO PROPAGATE A LIE!

-- edit -- oops, I may have misunderstood your post. Did you mean the OS? You're not very clear on that.

Edited 2010-10-07 01:14 UTC

Reply Score: 3

XCoder
Member since:
2006-08-11

IMHO the browser-only computing is a real scenario, but it is not a too good thing. In this case the Chromium will kill windows, macos, desktop linux, etc. You can buy 10$ laptop with Chomium - but you cannot use it without a google account...

The web computing make people very-very vulnerable. With a real computer with windows or linux your depenency on software vendors limited to purchase or upgrade time. You can use pirated software (not too nice, but real alternative). With web applications your dependency is continous: the vendor can kick off from the system at any time. The web based computing is the opposite of the spirit of open-source.

Reply Score: 4

Kroc Member since:
2005-11-10

Not unless open source sees the need and changes things. Why should remotely storing and accessing data mean that you should give your data to a company? Self-hosted web apps should be the norm.

Reply Score: 1

Lennie Member since:
2007-09-22

hear, hear !

Thank you for that one.

Some say the Linux-powered $US 100 wall-plug will be that device which will do the self-hosting (with a backup with the data encrypted of course).

With the whole mobile thing now getting really interesting, I've been thinking maybe it's actually a mobile device which will do this.

Who knows, interesting times.

Reply Score: 2

Bill Shooter of Bul Member since:
2006-07-14

Self-hosted web apps should be the norm.


I agree 100% with that statement.

However much that really *should* be the case, it won't be the case. Most people would have to come to the conclusion that the convenience of using an existing service like Google's various services are not worth the lack of privacy . Unless, someone makes a cheap web server as easy to administer. Like its built into your cable/dsl modem and works as easy and painless as wifi/dhcp.

Reply Score: 2

vodoomoth Member since:
2010-03-30

Unless, someone makes a cheap web server as easy to administer.

Opera has already gone that way with Opera Unite. I can access my files from wherever I am in the world, stream my music to whatever computer I'm using (until the music industry finds out about this and criminalizes me). Of course, that server runs inside the Opera browser. I know nothing about the security of it in terms of access (I mean someone trying to read my files). I don't know whether the app and infrastructure are not doing something nasty behind my back such as logging my passwords... But I guess it's no different from entering the same passwords in the browser.

Like its built into your cable/dsl modem and works as easy and painless as wifi/dhcp.

you forgot "when they work fine" at the end of the sentence :-)

Reply Score: 1

ASP ?
by Lennie on Wed 6th Oct 2010 20:12 UTC
Lennie
Member since:
2007-09-22

Just a quick note:

'It wasn’t until the popularisation of server side processing languages (particularly ASP) that saw the first “web apps”.'

Euh, many people were already using Perl for this years before that.

Reply Score: 4

RE: ASP ?
by Kroc on Wed 6th Oct 2010 20:39 UTC in reply to "ASP ?"
Kroc Member since:
2005-11-10

But ASP brought server-side to the mindless Microsofties of the dot-com boom, which was my point. It wasn’t about who did it first, it’s who made it easy enough for unskilled people to clobber together something that works. There’s no way you could say that Perl was as peachy easy as ASP.

Reply Score: 1

RE[2]: ASP ?
by Lennie on Wed 6th Oct 2010 20:47 UTC in reply to "RE: ASP ?"
Lennie Member since:
2007-09-22

Yeah, I guess unskilled and Perl don't mix that well. :-)

I see your point.

Edited 2010-10-06 20:47 UTC

Reply Score: 2

RE[3]: ASP ?
by Bill Shooter of Bul on Wed 6th Oct 2010 22:44 UTC in reply to "RE[2]: ASP ?"
Bill Shooter of Bul Member since:
2006-07-14

That didn't stop them from mixing ;) Cleaned up a few of those messes in my day.

Reply Score: 2

Last hurdle for browsers: access to devices
by Lennie on Wed 6th Oct 2010 20:53 UTC
Lennie
Member since:
2007-09-22

The editor of HTML5 (Ian Hickson) has created a draft specification for HTML-applications to get access to devices. Which means you won't need anymore plugins (read:currently Flash) for a webapplication to get access to the webcam.

Or you can make a webapplication which can read directly from a barcode reader:

http://ajaxian.com/archives/video-conferencing-with-the-html5-devic...
http://dev.w3.org/html5/html-device/

Why not upload your photos directly from your USB-connected camera through the webbrowser into the Photoshop-like webapp ?

(yes, yes, all with: grant this web page/domain access to device-X interface ofcourse)

Edited 2010-10-06 20:55 UTC

Reply Score: 2

Kroc Member since:
2005-11-10

There’s also experimentation going on with `<input type="camera" />` and `<input type="speech" />`. This is exactly where OS integration and native-UI totally matter.

Reply Score: 1

Lennie Member since:
2007-09-22

If these tags really get added to the browser, I doubt there will be much of a GUI.

That is the whole point, the browser is the UI.

Even before any webpads (I'm not talking about the iPad, I'm talking about, long, long ago) I was thinking one device which connects to the Internet is really all we need.

Next think on my list of things that still need to be 'invented' is a good API for the user to say, please copy this file from web-application X to web-application Y.

Now that we have oAuth, etc. it might start to happen.

Edited 2010-10-06 22:08 UTC

Reply Score: 2

All this talk about cloud computing...
by koki on Wed 6th Oct 2010 21:01 UTC
koki
Member since:
2005-10-17

Maybe I am just too old fashioned, but to me web-constrained (cloud) computing is a step backwards in terms of usability and user experience (perhaps several steps). Until usability in web applications improves to a level where it can at least challenge the user experience of existing OSes, I would rather do my computing on the ground than in the clouds. In all my ignorance, I have the feeling it is going to take quite a while before that happens...

Reply Score: 4

Lennie Member since:
2007-09-22

It is probably true, it will take time. Even W3C announced today, the full HTML5-specification is still years away from being finished. Not that this won't prevent developers from using it anyway. :-)

Reply Score: 2

Why the OS Doesn't Matter
by Drumhellar on Wed 6th Oct 2010 22:25 UTC
Drumhellar
Member since:
2005-07-12

"Why the OS Doesn't Matter", an article written by a man who created an OS that, in the end, didn't matter.

[j/k]

Reply Score: 4

DeadFishMan
Member since:
2006-01-09

... there are several factual errors - more like unintended revisionism at work - in the article that are distracting.

The first thing that bugged me was the statement that ASP should be credited for the coming of the more "sophisticated" web pages when that is plain not true! As already addressed by Lennie, if anything that honor should go to Perl that, despite not being created specifically for it, served us very well during the early years of the web as the engine behind (in?) famous CGI-enabled websites.

Even though Perl is definitely not a beginner's language (no arguments there!), the web is still littered to this day with Perl scripts repositories with clear and concise directions about how to deploy pretty much anything ranging from visit counters to chat scripts to CMSes and everything in between and only as of the huge explosion of PHP circa PHP3 that trend slowed down a little. Been there at the time, done all of that. I had already passed through one or two ISPs at the late 90's so I am pretty sure that ASP had nothing to do with that.

Most hosting companies had either cheap boxes running Linux/FreeBSD or, in case of big ones like the late RapidSite (nowadays part of Verio) huge datacenters with Sun machines or, ironically, SGI machines (running IRIX nonetheless) to provide shared hosting services.

Heck, MS' own FrontPage had server extensions that were basically a bunch of Perl (and shell scripts IIRC) until the very day that it was retired!

In fact, ASP was't even a blip on the radar until ASP3 came out. And even then, ASP was running neck to neck with PHP from a features point of view but its market share/installed base/whateveryouwannacallit was nowhere near the latter's and for a while it seemed that its only distinctive advantage was the fact that it was heavily tied into MS tools and middleware. At that point the article starts to make sense as MS did invest heavily into providing easy to use tools to deploy websites quickly using its development tools (which happens to be MS forté, IMNSHO) which somehow boosted its presence on the data center, at least as far as web hosting is concerned.

(And yes, Kroc: I'll concede that ease of use was probably a major concern for the hordes of Visual Basic and VBA developers out there back then but that still sounds like a stretch to me... ;) )

The other thing is that traditional business had absolutely nothing to do with many enhancements of the web during the late nineties. Most people agree that the porn industry (yes, I said PORN!) pushed most of the things that we take for granted today (e-commerce, filesize compression techniques for web server software, demand for faster bandwidth, etc.) and that the traditional brick and mortar businesses only took a chance on the web after witnessing the profits that those pioneers companies reaped during that time.

Once the porn industry laid down the ground rules about how to do business on the web was that most companies jumped in and it became what we know today.

But I have no peeves with the rest of the article. MS should be commended for the groundwork that led to the powerful AJAX websites that we have today even though I am pretty sure that they regret that they couldn't figure a way to tie that to Windows and create a better mouse trap. ;)

EDIT: Uhhh... Sorry for the huge post... :S

Edited 2010-10-06 23:50 UTC

Reply Score: 3

Kroc Member since:
2005-11-10

I said in the article “particularly ASP” - as in, there certainly were server side systems before (I wouldn’t even begin to deny that), but that ASP changed the gears up because it lowered the level of access and was similar enough to VBScript / VBA / VB. Before ASP, server side was being done properly by proper engineers in Perl. What happened with the dot com boom is that a task force swept in of such size that there wasn’t enough IQ to go around and Perl was certainly not their language of choice.

Reply Score: 2

axilmar
Member since:
2006-03-20

For a site that is named 'OSNews', confusing the UI with the OS is a serious problem.

You could have said what you said with a single phrase:

"The more a web app looks like a native app the better".

Besides the UI though, the OS has little to do with the web. In fact, it is irrelevant. You could run a web app on a Commodore 64, provided that the interface is nicely looking and simple to use/understand.

Reply Score: 2