But alas, the Webkit team – who really have a great track record of being successful with bleeding edge, one upped them by not only passing the test, but releasing the code. So behold, this is Webkit nightly for Windows, build 31368 from 2008-03-26.

We know that Safari 3.1 doesn’t and Opera 9.5 won’t pass ACID3. We know IE8 is a long way off. We know Firefox 3 is still pretty far from it too. But now we have browsers that can do it. The the big question is, who will have the first stable general release that does it? Safari 3.2? Opera 10?

It’s an exciting time in web development, and I hate to admit that I think it’s largely due to IE8. If the IE team steps it up, some of themes technologies have the potential to reinvigorate the web. No serious e-commerce site would alienate all IE users – even today, they make up 80% or so of internet users. But as things progress here, we’re likely to start seeing some incredible things in the next few years.

Update: A bug in ACID3 was apparently noticed as a result of the Webkit team’s work. This awesome detailed blog post from the Webkit site chronicles the final steps of the adventure. Note that the “animation smoothness” criteria is subjectively, and that the team is apparently giving themselves a fail, but nothing that they think they are “faster than all other browsers“. Congrats again, Webkit team. Well done!

Finally, not to have all headlines stolen this St. Patrick’s Day, Microsoft loosed Vista SP1 to Windows Update.

I have installed Safari 3.1/Win and this evening I will upgrade at home on the Mac. I am currently downloading Vista SP1 for my work PC. Reviews to follow, for certain.

Enter Acid 3. Acid 3 measure even more goodness, including these six “buckets”:

• Bucket 1: DOM Traversal, DOM Range, HTTP
• Bucket 2: DOM2 Core and DOM2 Events
• Bucket 3: DOM2 Views, DOM2 Style, CSS 3 selectors and Media Queries
• Bucket 4: Behavior of HTML tables and forms when manipulated by script and DOM2 HTML
• Bucket 5: Tests from the Acid3 Competition (SVG,[5] HTML, SMIL, Unicode…)
• Bucket 6: ECMAScript

Using recent browsers, everything fails pretty spectacularly. My Opera 9.26 install gets a 42/100. Safari (including iPhone) does 39/100. IE7 does 12/100, Firefox 2 does the most respectable with 52/100. Even IE8 only does 17/100 while Firefox 3 tops out at 59/100 and Opera 9.5 at 60/100. The current generation, even the next generation of major browsers are still far from coming close to rendering Acid 3 with any accuracy.

I have been playing, now and again, with Webkit nightlies, since Webkit is actually a really neat engine, and guess what it kicks out? This:

Webkit nightly on Windows Vista

Pretty impressive. Safari is pretty limited when it comes to extending its function – it doesn’t even support a “new tab” button. But the webkit and javascript core engines are respectable both in rendering skill and speed.

I love Camino, really I do. But recently, its limitations have been bothering me. I prefer my tabs in a very specific order and often I have several tabs open. If ever I close a tab by mistake, I cannot get that same order without doing tons of work or re-launching. Safari 3 draggable tabs.

One of the things that used to bother me about Safari was that there was no “New Tab” button available for the toolbar. There is now. It’s also got great keychain integration, private browing, the original embedded RSS, true Aqua widgets, resizable text boxes, easy PDF integration, and it’s super-fast.

Camino doesn’t support Ad-Block, but rather, stylesheet-based filtering. Safari does that too, by default, and it’s even easier to use than it is in Camino. Safari doesn’t have any Flash problems and once you add “Safari Stand” and enable the debug menu, you have a perfect drop in replacement.

My biggest complaint about Camino was the lack of development tools. It doesn’t have a Javascript debugger (ChimericalConsole never worked me for), doesn’t have a decent source viewer, doesn’t have many third party hacks to add functionality – it’s a browser for users, not developers. Without XUL, it’s tough to add features easily. And that made it tough to use for me. When I did any serious work, I’d always switch to Opera or, more recently, Safari 3. Safari 3’s Inspector is just awesome.

So… for now, I am Opera on Windows and Safari on Mac. My browser requirements are more demanding than most. I have felt for some time that Opera and Firefox on Mac just “feel” wrong, they don’t fit. So we’ll see how the Safari experiment goes.

When I started using Safari 3.0.1 beta at work, I was impressed, but not impressed enough to ditch Opera. At home, however, I am using Camino, which I love, which is based on Gecko, the underlying Mozilla engine that also forms the core of Firefox. The problem is, as much as I love Camino, it’s tough to use for development: it doesn’t support extensions, it doesn’t have a javascript debugger that works, it doesn’t have draggable tabs, or tab restore, and it’s not very easy to extend functionality. There are lots of tricks at PimpMyCamino, but even today, the most useful add-on, “CamiScript,” is billed as unstable on Camino version above 1.0. Camino 1.0 was released in the first half of 2006. We’re over a year later.

This is not a post to bitch about Camino though. I love 1.5 and it’s serving me well. The thing is, I downloaded a nightly build of Webkit recently. Webkit is to Safari what Gecko is to Camino, and Webkit comes easily packaged in a disk image that requires no installation.

Webkit nightlies are awesome. First, there’s the page inspector. From a development standpoint, this is awesome.

click image to view at full size

The inspector shows you each detail of the page load. You’ve got the entire page transfer size, as well as the page transfer time. You can break it down by element or by element type. You can view the headers sent and received. This is tremendously useful. It’s been very interesting to see what parts of requests are properly cached and compare original load to subsequent page loads.

Then we have “Drosera,” the Javascript debugger.

click image to view at full size

I haven’t quite figured out how to use this tool, but I’m excited that it exists. It’s something I’ve needed for some time on a Mac. This is all very promising.

Safari may be mostly bare, but by the time 3.0 final is released with Leopard, plus the fact that Safari exists on Windows, it, or its featureful offshoot based on Webkit, Shiira, just may be my main Mac browser.

You can get Webkit nightlies at nightly.webkit.org.

But alas, I ran Software Update and updated my Safari/Win install at work to 3.0.1. Whereas 3.0 was a major disappointment at work – fonts were a mess, pages had major problems with rendering, and the browser would crash randomly – a few minutes after install I can tell you that 3.0.1, on my computer at least, is a HUGE leap forward. The browser hasn’t crashed on me outside of one bug that existed before (maximizing on the slave screen of a dual-monitor setup), the thing is SO much better!

Safari is far from usable as my main browser. The thing is feature-barren, is far less customizable than Firefox and Opera and even Camino, and on Windows, it sticks out like a sore thumb. That said, I just love having the rendering engine on my windows machine, I love that it’s available for iPhone and Mac-friendly web development.

Kudos to Apple for porting this great app to Windows fairly successfully. Microsoft has been very slow to move to OS X and Intel; they have let RDP stagnate, they have let Office go 5 years with no update, they have no management tools that work on Mac, no IE, no WMP, not even a fully compatbile Outlook Web Access (OWA)… yet.

I am usually wary of excessive praise on Apple, but after seeing the Leopard previews pushing the evolution of the desktop and the accessibility of backups, the iPhone pushing the mobile experience, and Safari pushing web standards, I’m really feeling good about what they are doing.

I suggested that Safari should exist for two reasons: firstly, that web developers could test their apps in Safari, and secondly, to lure more users into comfort with the Mac UI and Mac apps. So, score me 50%. There is one reason and one reason only for Safari on Windows – so developers can test their stuff in Safari. Now, it turns out it’s less for web sites and web apps than it is for iPhone development, but nonetheless, iPhone apps are, in fact, Safari apps. Thus, web developers can now test their sites in Safari, whether for iPhone or not.

The interesting thing here is that Apple is in a very unique position, and I hope they don’t pull a Microsoft. Apple can now introduce new proprietary hooks into their iPhone. Let’s say they “extend” javascript or CSS or even HTML itself. What if they invent tags like <iphone:dial> or <iphone:toAddressBook> or something that has unique function ignored by normal browsers but defined on the iPhone. I dread this, and yet, it would allow for rich, powerful applications without an SDK.

Assuming, or even ignoring that possibility, Safari on Windows does all Windows-based web developers to test their sites in Safari. I just installed Safari 3 on my Mac, and found it to be fantastic; it’s faster, it’s more compatible, and thus far, it’s a far better browsing experience. That said, on Windows is was a nightmare. It doesn’t play nice with dual-monitors, it doesn’t handle fonts well on my work computer (defaulting most fonts to “Metal Lord” font, odd choice) and crashing randomly. But then… it’s a beta and a first shot, and I bet most of these bugs are fixed.

Either way, I think this was a great move by Apple to establish themselves as serious about making the Mac a first class citizen for web browsing. Currently, it’s just not. There are several notable sites, like say, the MLS, which require IE. And there’s simply no IE for current Mac users. So this is great news all around, even for the Opera-ers, Firefoxers, and Camino-ers who use Macs.

I’m not feeling especially vindicated by this announcement, because I don’t think I spotted something so far fetched – I always felt Safari/Win was a good idea. But I am thrilled to see the seeds being planted for the Mac to be considered a legitimate, affordable, enjoyable contender as a computing platform for the general public.

If Apple ports Cocoa to Windows (like they obviously have done with at least a subset of Carbon in order to run Quicktime and iTunes), they can introduce all sorts of Mac software for Windows which could very keenly familiarize Windows users with the Mac experience to help lure potential switchers, people who might be close to considering a new computer and having to face a learning curve with Vista anyway.

In other words, this is great news. I’m firmly on Opera right now, but if Safari for Windows came out, you never know…

Read on, I’ll tell you why.

Safari is the native OS X browser. It’s akin to IE in Windows: it’s been preinstalled as the default browser since Panther, and it’s the only bundled browser in Tiger, so almost every Mac user has at least tried it, if only to download Firefox, Camino, Opera, or OmniWeb.

The problem here is that you’re trying to convince people to switch from Windows, which is like a crappy old blanket: it’s got holes in it, it smells a little, the dog peed on it a few times, and it doesn’t even work that well, but it’s familiar and using it is like second nature if only because we’ve been using it so long. Since so much of the home computing experience is using the browser, getting people to feel at home in their most used application is a big step. Combine this with iTunes and you’ve got yourself one hell of a 1-2 punch.

Plus, we already know that your apps can be successfully ported to Windows. iTunes keeps pace on both platforms, and is a large part of the sales pitch you use: “Know iTunes? Then you know a Mac.” Quicktime codecs and the player itself are well synced. iDisks are compatible with Windows, iPod updaters and Apple Software Update run, so we know your engineers can program for the platform. I don’t think any of these are written in Cocoa like Safari, but I’m sure your engineers can make this happen if they were motivated to do so.

How about this: people don’t write websites for Safari. Safari’s Webkit is a branded version of Webcore, which is based on KHTML, the rendering engine for KDE’s Konquerer. But it’s only “based” on it, it’s not the same. There are changes to Webcore that haven’t been backported, so they aren’t completely compatible. As such, it’s not easy to design or test webpages to work well in Safari with direct access to a Mac. You don’t provide web developers any sort of VMWare or Virtual PC image they could use, so they just have to assume things work. If your browser is considered a second class citizen, then your whole OS suffers from obscurity syndrome.

Thus is the gist of this piece: by running Safari on Windows, you can not only coerce more web developers into providing Safari compatibility, and therefore be a true presence, rather than a fringe app, but you can also introduce users to the Mac experience even just a little more. Plus, now that IE 7 has been totally reconfigured and is once again foreign to users, you have a chance to introduce users who aren’t planning on upgrading the Vista just yet to Safari and Macs, and possibly even make a play to be their next purchase, rather than a tired, old, unexciting PC.

Maybe it’s making a leap of faith, and yes, the browser market is one where making a noticeable entrance will be challenging, but the less of a jump into the deep end buying a Mac is, the easier it is to make your Apple brand accessible, available, and not scary. The best way to start? Safari on Windows.