Fog Creek Software
Discussion Board

"New API is HTML", but "IE is dead"

Joel says "None of this bodes well for Microsoft and the profits it enjoyed thanks to its API power. The new API is HTML, and the new winners in the application development marketplace will be the people who can make HTML sing."

But, before, he said "It's not that Microsoft didn't notice this was happening. Of course they did, and when the implications became clear, they slammed on the brakes. Promising new technologies like HTAs and DHTML were stopped in their tracks. The Internet Explorer team seems to have disappeared; they have been completely missing in action for several years. There's no way Microsoft is going to allow DHTML to get any better than it already is: it's just too dangerous to their core business, the rich client."

So, how can HTML be made to sing if IE development has been stopped, since IE is the de facto browser standard? I'm confused.

Chris Ryland
Wednesday, June 16, 2004

I think you miss the point; the new API is HTML and Microsoft doesn't want that...  so of course, browser development has stagnated.

Now, you can still build pretty decent applications (lets say an entirely new class of application) with HTML as it stands right now.  The web browser itself will continue to evolve (witness Firefox).  At some point, Internet Explorer will either get displaced or they will have to add to their browser.

Almost Anonymous
Wednesday, June 16, 2004

I see that Joel, as he is a microsoft fan, didnt breath the dreaded "JAVA" word.

Java, an OS in its own right, and can run on most platforms, it should be the logical choice to develope that rich client application.

No wonder Bill Gates is so scared of JAVA.

Wednesday, June 16, 2004

Almost anonymous says:

"I think you miss the point; the new API is HTML and Microsoft doesn't want that...  so of course, browser development has stagnated.

"Now, you can still build pretty decent applications (lets say an entirely new class of application) with HTML as it stands right now.  The web browser itself will continue to evolve (witness Firefox).  At some point, Internet Explorer will either get displaced or they will have to add to their browser."

No, I think I got the point.

But your second point seems to assume that people will use some other, more capable browser as it evolves (since IE isn't evolving). I don't think that's the case--the 399,000,000 people Joel's thinking about (after he dumps all the Mac users ;-) don't know a browser from their elbow. They just click on the "globe icon" and they're off.

Chris Ryland
Thursday, June 17, 2004

Why you should dump IE:

fool for python
Thursday, June 17, 2004

What proof is there that MS stopped IE development because it threatened their API?

If IE already controls a 96% market share then what is the point in fixing it?

I believe that MS stopped developing IE because there was no serious competition and they could maintain dominance without spending money on innovation.

Who on earth keeps working on an app that already controls a 96% share?  Better to spend your resources elsewhere until a serious threat appears.

Thursday, June 17, 2004

Andrew, I think you're missing the point by asking for proof; Joel is speculating.

Besides, why would Microsoft have continued work on Windows?  Your 96% market share argument applies just as much to that situation.  Microsoft was willing to continue work on its core products so that competition wouldn't start to gain momentum, but the IE group seems to have run out of steam ever since their spectacular defeat of Netscape.

Speaking of Netscape, I wonder what Joel's opinion of XUL is, as applied to this context... it could go a long way towards solving his HTML usability complaints, if only people would develop for it.  Which they more than likely won't.

Thursday, June 17, 2004

> No wonder Bill Gates is so scared of JAVA.

Bill isn't scared of Java. Nobody is scared of Java anymore, nor exited for that matter.

Daniel Tio
Thursday, June 17, 2004

Man this is the best topic that has come up on JoS in a long time.  I can't stop spewing meaningless comments, and I've got to add another uninformed opinion here.

I think Microsoft might be really screwing up by not putting out IE 7 before Longhorn ships or upgrading IE on older platforms.  It is a buggy security problem.  We've tried integrating WinInet into our application twice only to find show stopper bugs both times.  It used to be handle leaks, now it is locking up while making ADO connections simultaneously.  This is the basis for IE.  The fact that virsus are now spreading via IE has to scare those who actually care about security.  The other thing is Firefox is passing IE in usability and reliablity.  If companies start specing Firefox for accessing their intranets, Microsoft could get in trouble real fast.  Both the client and the server become commodities.  This could get interesting. 

Another thing that Joel misses is that it doesn't matter how much money Microsoft has in the bank or what resources they have at their disposal, if they start missing their numbers it could start to get ugly fast.  Corporations aren't set up to deal with shrinking bottom lines, and let's be real, Microsoft really has no place to go but down. 

christopher baus (
Thursday, June 17, 2004

"I don't think that's the case--the 399,000,000 people Joel's thinking about (after he dumps all the Mac users ;-) don't know a browser from their elbow."

Way back in the day everybody knew what Netscape was.  IE was the underdog, even bundled with the OS, that nobody used.

Lets compare with instant messenging:  MSN Messenger is bundled with Windows, it's there when you first bootup, it even asks you to create an account the first time out.  What's the most popular IM?  AOL!!  Newbie don't-know-nothing-about-computers go out and download and install that software.

If you, and I, and everyone else started swithing to a superior browser than eventually even Joe Newbie would switch.  Everyone wants to have the best experience and when the difference is wide enough they'll do whatever it takes.

Firefox is already starting to take over for one simple reason: security.  Newbies are especially troubled with adware and viruses.  Tell them that Firefox blocks popups and doesn't ask you every 5 minutes to install some hostile ActiveX control and they are very greatful.

Almost Anonymous
Thursday, June 17, 2004

>So, how can HTML be made to sing if IE development has
>been stopped, since IE is the de facto browser standard?
>I'm confused.

You write really good web-based applications using good 'ole HTML that browsers have been doing for years.  You don't need to upgrade IE to support kicken HTML based apps, IE does HTML just fine.  There is a lot of untapped power left in HTML that people haven't exploited yet.

That said, Microsoft will soon be stuck in a place where they either don't update IE and have a new, more capable browser take over or upgrade it to support new technologies.  It sounds like Mozilla is about to create a new way to provide "rich client" functionality in web-based apps.  I have a feeling things are about to get fun for everyone but Microsoft.

Eric Galrand
Thursday, June 17, 2004

I was reading and agreeing all the way down the line with Joel's original article until it came down to this left turn into the cul-de-sac.

Yes, I agree that Microsoft has lost its hold on the API and that .NET doesn't give the traction to tie developers and applications into their platform in anything like the same way.

But HTML is not the new lingua franca API.  If there's any result out of this at all, its that there is no broad common API any more.  There are a multiplicity of APIs, some relying on heavier clients or heavier runtimes and some relying on back end servers and runtimes.

Yes HTML will be one of those, though XHTML and XML will probably completely overtake plain old HTML.

Will there ever be a platform specific API again?  Will Microsoft lose actual market share in the application segment because there are now so many virtual platforms to host products on?  Only time will tell.

Up to now product marketing's strategy at Microsoft has not been about being best in class but most ubiquitous.  Not necessarily a USP argument more that products are positioned such that the customer or specifier has to argue against buying a Microsoft product rather than for what may be a better product.  Just because of ubiquity.

In a market where there is no dominant O/S API that applications have to write to Microsoft are going to have to out develop out market a gazillion little guys, but this is much simpler than having to out market one or two similar size heavyweights.

You might argue that this could be a turning point in Microsoft's fortunes, that it could be the beginning of the slow decline in revenue and market share.  But even in this scenario of multiple APIs Microsoft still has most of the advantages, it can afford to develop in multiple APIs, even if taken as a whole that seems crazy in bottom line terms across the organisation it makes sense in an organisation that has separate competing divisions.

And that I think is the real change in Microsoft, not that they've lost the API, but that they've evolved into competing internal organisations because the external competition is so weak.

You might characterise the two philosophies as classical and romantic (q.v. Zen and the Art of Motorcycle Maintenance), the Raymond Chen (classical), MSDN (romantic).  There are signs that whole product divisions are proceeding along these separate tracks making whatever deals they can along the way with the other side in order to get their product out the door.

Whilst in most companies that have a single core product family (whether software or not) the concentration in product marketing is in getting people to buy your product, in companies like Microsoft the concentration and effort is in getting your product out at all in competition with all your company's other products.  The politics of product marketing in such companies can be very nasty.

In short, it looks like Microsoft is going to become a real large company, not a small company with lots of people.

Simon Lucy
Thursday, June 17, 2004

I've been reading this thread with amusement since:

a) I just got hired by microsoft, and
b) I have been doing DHTML-based web apps professionally for quite some time.

Let me tell you, I don't know about simple-style web applications like google, but the entire ecosystem around advanced-style web applications with DHTML and the like is drying up faster than Lake Powell (1).

There used to be a thriving community around DHTML at places like and Those places still have users, but it's nothing like the good old days. Many of the professional DHTML developers I know have moved on to more profitable development like Java, PHP, and - surprise surprise - .NET.

I don't know where the developers for these clever applications are going to come from. From my vantage (and it should mean something, I think) there aren't that many clever javascript programmers left. To be honest, I think I can count them on one hand. Who is going to write these apps? If you haven't noticed, it's really not all that easy (2).

I think the salient point that is missing from both the original article and these various comments, is that developers don't want to screw around with extremely low-level mundane details when they're trying to draw a freaking menu (3). That is the reason the win32 api must be replaced and that is the reason DHTML is withering (4).

Somebody wondered why microsoft's 96% share on the desktop wouldn't protect them in the same way that it protects IE. The answer is that no developers are running to go develop for alternative browsers *in a way which prevents users from using IE*. On the other hand, developers were, in droves, moving to Java on the server because they couldn't freaking stand COM or ASP classic.

Microsoft has always understood that it's real power is their developers, and they still do. I see .NET as quite simply their latest attempt to get developers really excited about programming for windows computers. No more, no less.

And from my perspective as an ex-ASP/DHTML (5) nerd whose friends have all moved to ASP.Net or Winforms (and are particularly excited about XAML), it seems to be working.



(3) and



Aaron Boodman
Thursday, June 17, 2004

To me, this talk of HTML being the new API just sounds like all the BS that came out of the dot-com days.

Steve Jones (UK)
Thursday, June 17, 2004


Excellent rebuttal.  I'm not sure I agree entirely but it is well thought out and well written.

name withheld out of cowardice
Thursday, June 17, 2004

As far as it goes for people switching away from IE...  I think the internet has turned a corner is getting outright hazardous to the average computer user, even if you are somewhat careful.  I've had my browser hijacked, and now I don't want to use IE anymore.  And I've heard from other people who have had to completely reinstall after their IE gets trashed by some trojan.  As these stories get more and more frequent, people are going to be advised that it isn't safe to use IE anymore, and the big companies will be switching away from IE in their offices for security.  Though the tide hasn't quite turned yet, I really think IE is about to lose the browser wars.

Keith Wright
Thursday, June 17, 2004

In most cases, though, the right security settings can prevent spyware problems when browsing.  And if you run something like Spyware Blaster, then you're protected against most spyware and hijackings already.  You should be running something like that regardless of your browser.

It has yet to be shown that alternative browsers like Firefox are secure.  All we know is that they don't have the same obvious security holes as IE.

Thursday, June 17, 2004

Joel answered the IE stagnation question himself. Microsoft has enjoyed 90%+ browser market share for several years. A large number of public web sites and an even larger number of intranets and web applications are coded for IE and know about its quirks and tricks. Upgrading or replacing IE at this point is tricky and dangerous for Microsoft.

If Joel is right about Microsoft taking backward compatibility seriously (and I believe he is right), MS is probably reluctant to do anything to IE that will cause hundreds of thousands of web applications to break or change appearance or behavior. Regardless of what MS can say about better W3C compliance or new security features, people stuck with broken web apps will blame Microsoft. It's a no win for them.

So MS has to move ahead very slowly and carefully, which is exactly what they've done from IE 4 to IE 6. And they've had to deal with a nonstop plague of security problems which chew up lots of their resources, have no ROI, and only make their customers more wary of changing anything.

I advise clients and family to use Mozilla/Firefox (if I can't persuade them to use a Mac), but as a web developer I can't dismiss or ignore IE. It's not going to go away no matter how good Mozilla gets; as one poster pointed out people use IE because they just click on the globe. When I replace IE with Firefox I use the IE globe icon in the quick launch toolbar, and usually customers don't even notice anything different.

.NET gives Microsoft one way out. By moving more intelligence to the back end and getting web programmers away from twiddling HTML and Javascript, MS will eventually have some breathing room to rev IE.

Greg Jorgensen
Thursday, June 17, 2004

*  Recent Topics

*  Fog Creek Home