Dear Microsoft, please use WebKit

I would much prefer that MS wise up about the special requirements of a browser engine and do as well as they can with their own. Opera is more than welcome to switch to Webkit though. :)

Things like this will serve the web better than "Hey everyone, let's all use Webkit".

In summary: "If it's a core business function -- do it yourself, no matter what."

Building runtimes (and a web browser is a runtime for web apps) is a core business function to Microsoft.

Actually, Google surpassed Apple in commits to the WebKit repository a few months ago. Google adopted WebKit with Chrome but by no means are they not worried about the rendering engine's development.

Hasn't occurred to him that maybe MS also has invested too much in its rendering engine?

I don't know, maybe you weren't doing real work in this industry back then, but things are light years beyond what they were like in 2002. We've got JavaScript implementations that actually perform well. DOM works largely the same across browsers (remember that Layers debacle from Netscape? WTF). PNGs work properly. We have largely similar rendering across all browsers. I'm talking about things at least being in roughly the same place on the screen, having mostly the same styling, whereas you couldn't even count on things being in the same general location between browsers before.

"We are all working from the same W3C specs so why can’t we at least make everything render the same way."

Have you read the W3C spec? It defines a plethora of optional features and many key required features are ambiguously stated. In other words, it's a poorly written spec. What do you think is more likely, Microsoft intentionally breaking spec or the spec being poorly worded? Considering the differences between Gecko and WebKit that you yourself have also mentioned, I'm going with the common denominator here, the spec.

Eventually, some upstart would seize the opportunity by launching a new browser with a better rendering engine.

There is another theory which states that this has already happened.

http://code.google.com/chrome/chromeframe/

Just gotta convince the users to install it.

But that's going to be way easier than convincing Microsoft to dump millions in invested development.

The whole "develop for standards" doesn't work when each browser renders things slightly differently. While coding up http://bulletxt.zetabee.com I realized that each browser handles line-height, padding, margins, and font-alignment differently, especially when applied to textarea. Sure, most of the non-IE browsers handle things in a similar way but similar is not good enough in many instances. If I do margin-top: -4px, it works fine in Chrome and Safari but I need to make it -6px in Opera and -2px in Firefox. IE doesn't even work that way so I end up doing something completely different.

I would be completely fine with there being 1 rendering engine and 1 JS engine no matter the browser. And all browsers could improve on the speed/performance of these engines without changing the output or requiring different input. If I do padding-right:10px for a float:left element with position:fixed, I want it to look the exact same in ALL browsers.

I know standards try to do that but it just doesn't work. Standards work in theory but in practice, it is the code that works. WebKit works like WebKit. If I created the spec/standard based on WebKit and implemented it based on this new standard, it would NOT be WebKit. Standards work well for protocols and communication methods but for actually rendering arbitrarily complex window elements, they don't work and last decade of failed attempts at standardization have shown us that. Think of how many websites/tutorials/articles exist solely to help deal with browser inconsistencies. Now imagine if that effort could have been made towards something productive.

Most small and medium sized businesses still run at least one web applications that was written when IE had 90% of the market and which relies on Microsoft's backwards implementations.

I suspect Microsoft couldn't make Webkit backwards compatible to those apps if it wanted to. The trident code base probably has quirks Microsoft itself has completely forgotten about. But many small web based programs rely on those quirks to run.

I don't think the author understands the nature of the "browser wars". He seems to think that none of the players have had the specific desire to have their browser dominate so that it could control, through the power of de facto standardization, exactly how HTML & kin will behave. There's one notable player, namely the one he's beseeching, who has historically been aiming at exactly that. I'll grant that it's possible that they're in the process of having a change of heart: maybe in the future their desire for market share will be a mere matter of pointing user's searches to bing instead of google. But we're still waiting on delivery of a version of IE that proves this.

For all we know, they could be stalling to switch strategies at the last moment possible. But I think it's more likely that they will strategically leave something like Canvas out and continue the same old strategy, but on a slightly different front.

What they need to do (IMHO) is to have W3C create a reference implementation of the spec (and tests) and release that open source with say a BSD like license. So the different engines can use that if they want. Also if the browser doesn't display it exactly like the reference implementation then it shouldn't be called a web browser.

This article should have made an argument about retiring old browsers faster and getting people to use the latest versions sooner.

Other than that, I can empathize with his position. Web development is hard. I guess that's why we get paid to do it.

If MS would just implement the <canvas> tag (and according to the standard) then I'd be a little less disgruntled.

I really don't care if they all use WebKit or they all adhered more closely to some standard. I want less work for me to create a website. I guess I could always use Flash. ;-)

Of course not. If a corporate IT department wanted everyone to use a WebKit browser, then they'd already have done it.

If Microsoft starts using WebKit, they'll find a way to fragment it and make their implementation of WebKit incompatible and unmergeable with the rest of the crowd.

No. Please, Microsoft, stay away from the projects I depend on.