30 July 2005

IE7 Beta 1: The Vultures Fly High

At last, Internet Explorer 7 Beta 1 and Windows Vista Beta 1 are out and about. An MSDN subscription or invitation is required to play, and I am a grateful recipient of the latter.

Now what?

How about “Attack of the CSS Squad!” (In COLOR!) Wow, there sure is a lot of vitriol to go around, eh?

As the melee continues at full tilt, Dave, Molly, Anne, and Faruk, among others I’m sure, have weighed in with reality checks, calls for sanity, and requests for - here it comes - patience.

With that, here’s my unsolicited $0.25. (Look, it’s my bloggie and I’ll cry if I want to.)

Up first, a few official words about IE7 Beta 1, direct from Microsoft:

“In some cases, the architectural underpinnings for new features are provided for Beta 1 and user and developer-visible functionality is planned for beta updates.”

“Internet Explorer 7 includes fixes for issues with the Cascading Style Sheets (CSS) feature. Both the peekaboo and guillotine bugs have been addressed, and work on other issues is under way to provide web developers with reliable and robust CSS functionality.”

From What’s New in Internet Explorer 7 on MSDN

While this “What’s New” excerpt has already been sliced, diced, and made into beautiful julienned fries by the web standards cognoscenti et. al. … I’m going to have another go at it anyway.

No Alarms and No Surprises

Two CSS issues have been called out by name and resolved in IE7 Beta 1. Oh goody! That’s two down, um, how many to go? Well, look, it’s not like we were actually expecting all the CSS woes to be washed away, with everything tied up in a neat little bow, were we?

Were we?

Oh, there there, take this warm, forward-compatible blanket for comfort … and as Hayo said in his reply to Faruk’s post, have a cup of tea. (I think I’ll go make some for me too.)

“Architectural Underpinnings”

Now there’s a phrase you’ll never see in Scrabble®. Let’s see, I take this to mean that all the groundwork for [insert feature list here] is already present and accounted for in IE7 Beta 1. It simply isn’t 100% exposed to the end user just yet. We should expect to see more and more functionality revealed in future betas as each feature makes the grade.

I know it sounds like a convenient catch-all answer for “Where’s feature X?” but the truth is it isn’t an uncommon practice. I’ve seen it happen many times before … just not usually in beta.

In all my trips through the software development lifecycle, when an application is pre-alpha it’s not uncommon for bits of functionality to be stubbed out. There’s either a stand-in for the real feature, or perhaps the feature’s partially in place, but it’s not necessarily working 100% as intended.

Next comes alpha, where the features are all in place but the whole app runs like Bambi on ice. It’s at the alpha state that we’re feature complete and no new features are added. No goldplating allowed, but there’s plenty of bruising and bandages. Comes with the territory.

This isn’t where IE7 is though. Not if it’s a beta. By this time, everything that’s supposed to be in place has long since reported for duty, warts and all. “Be there or be orthogonal.” It’s up to the beta cycle to make it all real and put the fit and finish on things.

Even so, there might be some features that made it to beta but for one reason or another don’t make the grade for Release Candidate status. These misfits might end up dormant, hidden beneath the surface until next time (which, in the case of APIs, can lead to cries of “look - undocumented features - <gasp!>”). Perhaps a feature might be backed out completely, so long as its absence doesn’t adversely affect other parts of the app. It all depends on the circumstances. “It is what it is.”

(We interrupt this post … I’ve just been handed this link: Internet Explorer 7 Beta Goes Gold. No! IE7 goes Beta! Ahh, technicalities. Har har.)

“Reliable and Robust CSS Functionality”

Back to the topic. I can’t help but feel the distinct whiff of the “legal/marketing treatment” creeping in here. Microsoft does not explicitly commit to any CSS1 and/or CSS2 particulars, nor is IE7 actually cited as the intended target for it all. Go back and re-read it. Appearing in the “What’s New” for IE7 doesn’t count, not the way this is written. The language is more than a bit non-committal in my book.

Even so, I have to wonder, if only hypothetically: Could the “architectural underpinnings” for CSS1 and CSS2.1 already be in place in IE7, just not enabled yet? Suppose they’ve already given it the old college try. Suppose it did make it in. What if the results are so problematic at this point that it isn’t even worth flipping the switch yet, lest they get an even greater backlash from folks who could care less about CSS2.1, not to mention those who do?

Measure twice, cut once, I always say.

That’s why, in the end, I can’t really wag my finger at the development team for delivering only two CSS fixes at this point. It would be rather smug of me to think that the IE Development Team or their management are CSS-ignorant. In fact I expect they are all rather clued and talented individuals. I also expect they are limited in what they can say and do at any given time, and that’s a mighty hard line to toe. Trust me, I’ve been in that situation on many occasions, and I don’t envy their task in the slightest. This isn’t the same thing as working on a project like, say, Basecamp or Ruby on Rails. This is, by my estimation, an entirely different animal, no matter how noble the intentions and wishes may be from all involved.

Waiting For A Superman

In the end, am I still a bit sad to see such a lack of CSS progress? Absolutely, and I stated as much in my feedback. Of course I wish there was some way to help the CSS cause from an entirely different angle, which is why my attention has managed to turn yet again toward Rosetta. Not the one from Apple, I mean my proposed (and yet to be written) Browser Helper Object that tucks a supercharged version of Dean Edwards’ IE7 CSS fixer-upper inside IE5/6 and automagically does the right thing. Accent on “yet to be written.” Yes, show don’t tell.

This past Thursday I managed to (finally) get an Intel laptop with XP Professional, just in time for installing Vista. Perhaps I should hold off on that, or install Vista under Mac Virtual PC instead (fun!). Then I could leave the Dell alone, drop Visual Studio on it and start tinkering with BHOs again. Of course, having just re-read Joel’s excellent and sobering take on the Windows API, I get a li’l case of the shivers. Rosetta, in all likelihood, won’t be a managed code trip. In other words, this way lies (potential) madness. Hmm.

It is what it is.

<Pauses, breathes, sips some tea, gets up and goes for a walk. End scene.>

Posted by joe at 12:25 AM

Trackback Pings

TrackBack URL for this entry:
http://www.joesapt.net/mt/mt-tb.fcgi/26

Comments

Joe writes:

While reading the beta newsgroups, I came across a link to the IE7 Beta 1 Technical Overview. Let’s have a look …

“Supported Operating Systems: Windows 2000; Windows 95; Windows 98; Windows ME; Windows XP” - uh, whoops! Memo to Redmond: Remove all but the last one. (Yep, they’ve been kindly notified.)

Seriously, here’s where it gets interesting:

“Web developers have expressed some frustration with certain peculiarities in the behavior of Internet Explorer 6, especially in the areas of standards support. Application developers also want to take advantage of new capabilities of the Web but have been required to develop everything from the ground up to do so. …

“In Internet Explorer 7 beta 1, the browser architecture has been reengineered to address compatibility and will offer additional support for popular standards. …

“CSS is a widely used standard for creating Web pages. Internet Explorer 7 is prioritizing compliance to CSS standards by first implementing the features that developers have said are most important to them. As a result, in Internet Explorer 7 beta 1 Microsoft has addressed some of the major inconsistencies that can cause Web developers problems producing rich, interactive Web pages. The work Microsoft has done includes fixing some positioning and layout issues related to the way Internet Explorer 6 handles div tags. (More information about these bugs can be found online at http://www.positioniseverything.net/explorer/peekaboo.html and http://www.positioniseverything.net/explorer/guillotine.html.) The final release of Internet Explorer 7 will focus on improving the developer experience by reducing the time needed for developing and testing on different browsers.”

While the “What’s New” document seems rather non-committal in the CSS realm, that last sentence above actually offers something approaching a measurable goal.

But wait - it gets better. Chris Wilson posted a veritable windfall of information regarding Standards and CSS in IE (as in IE7) over on IEBlog.

This is a very big deal because for the first time we now have … specific and measurable goals! Now that’s what I call true commitment to CSS1 and CSS2. (OK, I trust that actually means 2.1 since it’s a legit revision.) Bravo.

# 30 July 2005, 12:19 PM -05:00

Comments are closed for this entry.