PDC 2008: What did we learn today?
On Sunday night, we presented our usual series of five flashpoints: topics we expected to see addressed and questions we expected to see either resolved, or asked in some prominent manner during PDC 2008. Let's review them in the context of what we know now:
| MORE: Will multitouch change the Windows application? |
- Will User Account Control change for Windows 7? Architecturally, no. It's still there, although its number of operating settings will be increased from three to at least four (if you count "off" as one of them). There will be an easier way for general users to change those settings, although users of Enterprise editions may expect their admins to disable the new Control Panel applet for them, using group policy.
MORE: Windows 7 will add a 'volume knob' to UAC But will a virtual rheostat switch be enough to change users' attitudes? As Microsoft's Mike Nash pointed out to us on Tuesday, UAC could evolve to a point where it's far less bothersome than it is in Vista, if only third-party programs which really don't require admin privileges just to do what they do (setup routines among them) simply followed the rules that Microsoft set out for them from the beginning.
MORE: Look for the 'delighters' in Windows 7 No matter how the final Win7 handles the re-introduction of UAC, it is inevitable that Microsoft will be ridiculed for that design choice. It has worked its way into a damned-if-you-do/damned-if-you-don't situation. For instance, if enough Win7 users choose to turn UAC off, the vulnerability level of the kernel could increase. If Microsoft ends up urging users to please, please not use the off-switch for UAC, then Apple will very likely parody that plea in a commercial at some point. (It's unlikely that anyone on the Design team will be asked to incorporate any "delighters" into this feature.) If it changes its mind, then Vista may be riffed as being "unfixed."
- What and where is this cloud? Well, that question has definitely been answered. What we still don't know is how much the cloud will cost; and that's unusual for Microsoft. Multiple representatives, including public relations and developers alike, answered the cost question with almost the same pattern of response: The company is learning what elements of the cloud will be monetizable, and while it's in the discovery period, Microsoft is urging members of the community to make suggestions.
We've heard of open sourcing one's business model, but this is ridiculous. In fairness, I can't imagine the company not having some concrete choices available to it; most likely, it's testing how its services will be utilized during this experimental phase, and applying models to these utilization patterns to determine which pricing models are the most effective and the fairest. But not saying that up front, in exactly that way, is an odd decision; a company should not be asking its cloud customers to spell out how they think they should pay. This isn't Priceline.
- Do enough developers see a need for "modeling?" Um...yea. That question was answered on Monday. The Dublin approach to the Oslo platform, albeit perhaps poorly named, easily generated more outright excitement among attendees than Windows 7, based on what we noticed.
Microsoft's Julie Larson-Green introduces attendees to a theme generator in Windows 7 -- very likely one of the "delighters" being added to what's essentially (so far) a Vista upgrade.
There's some possible good reasons for that: PDC is made up of people who live and work every day in their minds, not their desktops. Windows 7 may be beautified with a bounty of "delighters," but that really won't stir interest. Besides a little beautification and the addition of control over UAC, developers don't really see Win7 as much more than what its own kernel says it is: Windows 6.1. But Oslo introduces a fresh way of thinking, which includes a language that appears less based on XML and less geared for work on ASP.NET, and more around the native syntax that database developers and system architects use every day. XML is based around markup, which is for the good of the parser; M is for the good of the developer. The response to M so far is phenomenal -- possibly the greatest wellspring of enthusiasm since Microsoft introduced C#.
- Getting third-party developers back on the same page. As Mike Nash also acknowledged to us on Tuesday, some of Vista's ills were caused by developers who didn't start testing their wares until the very end of the beta cycle, if not long afterward. Nash dispensed some of the blame to Microsoft for not getting everyone in line.
MORE: Cleaning up the desktop in Windows 7 So obviously the company recognizes the issue here, and the Windows 7 rollout cycle appears intentionally designed to address it. Though there's no kernel refresh here, third parties are being given the time and the excuse to get with the program. The "pre-beta" released this week (we need to come up with some new names for things) will alert developers as to what to look out for, even though it's really more like one of the dinosaurs from the movie Jurassic Park. If you'll recall the story, the scientists filled in the gaps in the DNA sequences with genes taken from amphibians; in the case of the Win7 pre-beta, a lot of it is actually grafted clumps of Vista. No matter, it's still something to work with, and a fresh opportunity for third parties to develop applications and the marketing to go with them, as though Win7 were the upgrade to Windows XP.
- Will multi-touch matter? In terms of gathering developer interest, clearly, yes. But we only saw Win7's multitouch API prototypes being demonstrated on a hybrid HP tablet/desktop PC platform, and while that was certainly a cool system, it's less than representative of the type of device where multitouch could clearly have the biggest impact.
MORE: Will multitouch change the Windows application? If the emerging form factor of the netbook is ever to take hold, especially in this lousy economy, it will require an iPhone-like aura of wonderment and user delight. So far, the models for multitouch programming that we saw this week appear to be evolved forms of PC event-driven models, gleaning elements from the company's now-widely-known Surface experiment, grafted into the Win32 API. So it's being approached as a way to make Windows applications touchable, rather than a way to make new touchable applications for Windows -- and the difference isn't actually so subtle. There needs to be new concepts and new functionality besides just spinning a globe around, and it doesn't make too much sense to look to "the community" for ideas there when the community is looking to Microsoft for leadership.
| MORE: What is Azure and why should you care? |
| MORE: First look at Dublin, .NET for the cloud |
| Julie Larson-Green does something we've always wanted to do with multitouch: write on a virtual postcard. |
Besides the greater-than-anticipated popularity of Oslo, PDC 2008 did not give us too many surprises this year. That's fine, because we didn't particularly need very many -- a lot of attendees would be pleasantly surprised if Windows 7 were to Vista what XP SP3 was to Windows Me.
What did come as a surprise to some, due mainly to the over-speculation about Microsoft's timing, is that with all that the company has planned ahead in the Win7 development timeline, there's no way in heck that an evolved final product will be ready for RTM as soon as May. The first Win7 pre-releases this week to attendees aren't even being called "betas" -- we were given warnings all week, which amounted to variations on the theme of, "Pay no attention to that Vista behind the curtain." If Microsoft is lucky, the operating system that's really Windows 6.1 will make it out the door by the fourth quarter of 2009. That's when retailers and OEMs would like to see it, but even then, the company has already made a late start.
The news of Office 14's inclusion of complementary Web applications was received here with kind of a dull thud. Frankly, that's not Microsoft's fault, really: Office 14 Web applications, as far as we know, are not a platform that one can develop on. Sure, Excel for the Web looks really, really good. But what developers want to see more of is the collaboration platform being used in Office 14, and Microsoft wasn't willing to show that off yet. Developers aren't nearly as interested in promises than in real things; and this year, Oslo was a real thing.
There's a little more yet to disseminate, and you'll be seeing some more of our interviews and panel discussions next week in BetaNews. While we were impressed with Microsoft's increased transparency about Vista this year, we were also impressed with the resiliency and resolve of the developers who have helped to bring that transparency about. It was a somewhat humbler Microsoft (save for Ray Ozzie fishing for fancier metaphors in the cloud), and anyone whose typical knee-jerk response is to bash the company should take heed of the fact that this is supposedly the attitude we've all been asking for.