With the release of QuarkXPress 7 for the Mac OS X, the move toward Universal Binary — applications which contain bits for both Intel x86 and PowerPC processors — is progressing very well.

In studying my own software usage with Apple Activity Monitor, the only business-critical non-x86 binaries on my iMac are Microsoft’s Office 2004 suite and Adobe’s Creative Suite CS2 . Those continue to run through Rosetta, Apple’s translation layer. With Rosetta, applications consume huge amounts of virtual memory, and bog down to poor garbage collection.

Considering that Apple’s move to Intel processors was only announced in June 2005, and that the first Intel hardware began shipping in January 2006, this is a lightning fast transition.

Apple’s Safari browser is not as successful. Many Web applications (including the blog-writing system on blogger.com) don’t work correctly under Safari, and require Firefox to run on the Mac. This is particularly true of AJAX applications, and those which rely upon Flash. Safari’s RSS capabilities are dreadful; I use NewsGator’s NetNewsWire. I hope that Apple does a better job with Leopard.

Am still waiting for the Sun SPOT kit. The contact at Sun who offered to send me a kit doesn’t have an ETA. And from the comments on Sun’s mailing list, the company’s SPOT team doesn’t have an ETA either.

Postings on that list indicate problems regarding the RoHS compliance of the hardware, which means that the eval kits can’t be shipped into the European Union. RoHS is a new rule, which took effect in July 2006, that limits the amount of lead, cadmium, mercury, hexavalent chromium, polybrominated biphenyl and polybrominated diphenyl ether in electronic devices. But why that’s affecting distribution of the equipment in the U.S. is unclear.

>> Update 11/3 AM: Sun hasn’t responded to several requests for a schedule update.

>> Update 11/3 PM: Sun’s PR team told me, “At this point, it’s on hold.”

Yesterday, BT acquired Centerpane Internet Security, best known as the home of Bruce Schneier, one of the top computer-security experts in our industry today. Bruce, who founded Counterpane, will stay on as CTO. I’m sure that was a key part of the deal. While Centerpane has an impressive gallery of services clients, the real asset is Bruce himself.

Bruce’s classic, Secrets & Lies, is required reading for anyone concerned about software security. If you’re into crypto, his Practical Cryptography (co-authored with Niels Ferguson) is also good, although it’s biased toward Bruce’s own algorithms and against more commonly used ones. A better book for C programmers is his Applied Cryptography. I’ve not read his latest, Beyond Fear.

As someone who started his academic career intending to study astronomy, I can’t help but admire Brian May, former guitarist from Queen. I had no idea that he’d been a Ph.D student in physics/infrared astronomy at Imperial College, London. May’s new book, “Bang! The Complete History Of The Universe,” came out this month.

Homer’s Iliad has two Greek heros named Ajax. One is the giant king of Salamis, known as Ajax the Greater. The other is the prince of Locris, and called Ajax the Lesser (though, presumably, not to his face).

It looks like Microsoft’s Atlas project (named after another mythological figure) is more like AJAX the Greater. Atlas, now formally called the ASP.NET AJAX, is finally available in as a real beta. It’s been a long time coming.

To me, at least, a lot of AJAX development seems to be overly complex, with a lot of that complexity involved with plumbing, not buiness logic. Since it’s all new ground, there aren’t best practice, and the tools and runtimes are mainly “1.0” products, at best. Automation is rudimentary at best on most platforms.

Atlas, by contrast, cuts out a lot of the complexity, largely because it’s based on ASP.NET. The lack of portability cuts down the plumbing requirements. That’s one reason why ASP.NET, in general, makes Web app development straightforward. While it’s less scalable than Java EE for big projects, ASP.NET AJAX a heck of a lot easier to work with, because so much is already done for you on the tightly integrated Microsoft stack.

Assuming that you’re happy to work with ASP.NET, it’s been hard to find fault with the Community Technology Preview versions of Atlas, which have been more solid than Microsoft’s usual early efforts.

In fact, Microsoft’s initial support for both IE and Firefox has been praiseworthy. With the first formal betas, they’ve now extended their non-Microsoft browser support to Apple’s Safari, which is a welcome move, albeit a minor one.

What’s available today are two separate downloads from Microsoft. The first download contains the core ASP.NET AJAX 2.0 extensions and accompanying AJAX library, which is a fairly mature toolset — you can work with it today and feel confident that it’s not going to break when Atlas goes G.A. The other is a CTP of a lot of additional library features that are useful, but not required (see the feature matrix in Microsoft’s white paper).

Atlas — ahem, ASP.NET AJAX — has justifiably been criticised because it’s taking a long time. But then again, everything from Microsoft (from SQL Server 2005 to Team System to Windows Vista) takes forever. Inch by inch, step by step. But despite the delays, I think it’s a robust implementation, and worth the wait.

Microsoft has issued several security updates for Office 2003 since Oct. 6, but they won’t install on my 64-bit Windows system. (It’s clearly not my weekend for working with software.)

My big HP workstation (two single-core 2.6GHz Opteron processors, Win XP Pro x64, SP1) just won’t accept the seven “high priority” updates issued for Office 2003 SP2.

I’ve tried automatic background updates, running Microsoft Update in the foreground, and even downloading the patches to install them manually. No dice, and nothing about this in the Microsoft Knowledge Base. Grrr.

Today (Friday, Oct. 20) is the last day for discounted registrations to the Software Test & Performance Conference — tomorrow, the full conference passport price increases by $200. We’ve got a great conference for development and test/QA professionals, Nov. 7-9 at the Hyatt Regency Cambridge. There are lots of timely classes, including new tracks on test management, performance management and security. There’s a keynote from Rex Black. Plus, the second annual Testers Choice Awards. Don’t miss it.

After Red Hat purchased JBoss and created an integrated offering with Linux and a commercial-grade open-source Java EE app server, it was only a matter of time before Novell did the same. The answer has a singularly uncatchy name of “Integrated Stack for SUSE Linux Enterprise,” and launched this week out of a Novell partnership with IBM.

The Integrated Stack combines IBM hardware (such as blade servers or standard x86 boxes) with SUSE Linux, WebSphere App Server Community Edition, DB2 Express-C, and the Centeris Likewise management suite (which I’m not very familiar with). The system is sold by both Novell and IBM.

Red Hat, by contrast, offers several versions of its Red Hat Application Stack, with Red Hat Enterprise Linux, JBoss App Server, JBoss Hibernate, MySQL and PostgreSQL, and the Apache Web Server. No management tools.

The Novell/IBM pricing starts lower than Red Hat’s, at $349/year for the software. It’s unclear how much support Novell/IBM provide at the entry-level price point. Red Hat charges from $1,999 to $8,499, depending on the number of CPUs in the server, and the desired level of support services.

Sun has a comparable open-source stack, which can be run on Linux or Solaris. The Solaris Enterprise System, which consists mainly of Sun’s own software plus the PostgreSQL database, is a strong offering that doesn’t get as much exposure as it deserves. That’s largely because the tightly controlled Solaris operating system isn’t as popular with the “Anyone But Microsoft” crowd as Linux. Watch for the new Sun Java Composite Application Platform Suite, coming soon.

My column in today’s edition of SD Times News on Thursday discusses two new subsets of the Rational Unified Process — OpenUP, which is implemented in the Eclipse Process Framework, and EssUP, developed by Ivar Jacobson for use with Visual Studio Team System.

I’m anticipating that some people will ask, “Why didn’t you mention the Enterprise Unified Process?” That’s not an oversight. Scott Ambler has done a tremendous job with the EUP, but its goals are different.

OpenUP and EssUP were designed as simplifications of the heavyweight RUP: functional subsets that retained the principles of the Unified Process framework, but which were streamlined for agile development.

By contrast, Scott’s EUP is an extension to the RUP, bringing the production and retirement phases of the software development lifecycle into this well-defined process. If anything, the EUP makes the RUP even less agile — but leaves it more complete, and better suited for serious requirements-driven development.

Evans Data, citing poor attendee registration numbers, has cancelled its first-ever Development Products Conference. The conference, scheduled to be held in San Jose this Thu. and Fri., was billed as “If your job involves planning new technology products for developers to use, or positioning and marketing those products, this is the ONE conference you can’t afford to miss.”

This is unfortunate: I’d been looking forward to attending, and BZ Media was a sponsor of the event. Ted Bahr, president of BZ Media, was a scheduled speaker on the program, along with folks like BEA’s Bill Roth, TIBCO’s Ram Menon, IBM’s Bernie Spang, Programmer’s Paradise’s Jeff Largiader, Telelogic’s Brian James and Red Hat’s Bryan Che.

You won’t find any mention of the DPC on the Evans site; the company has scrubbed it away.

Last spring, my good friend Andrew Binstock and I agreed upon a simple wager: Would we be able to purchase a terabyte hard drive, in a 3.5-inch form factor, by the end of 2006? At that time, 500GB drives were readily available at places like Best Buy and CompUSA. I believe that the 750GB drives were just coming out as well.

Forget Moore’s Law: The pace of innovation in hard drives is incredible. As I write this, there still aren’t 1TB drives commercially available, but the price of 500GB and 750GB 3.5-inch drives is falling fast. You can also get a 160GB laptop drive (in a 2.5-inch form factor) for under $200, which is equally amazing.

Twenty years ago I’d not have predicted this… the future belonged to optical media, remember. When the CD-ROM was introduced in the mid-1980s, a typical hard drive had somewhere around 30MB of storage capacity. An ISO-9660 CD-ROM, by contrast, had 650MB capacity, or about 20x that of the hard disk.

Imagine if 12cm compact disc technology had kept pace with magnetic media: we’d have 15TB discs! But optical didn’t keep up, and capacity has grown very slowly. The first move, in the mid-1990s, was to the 4.7GB DVD, and then the 8.5GB dual-layer DVD. Now, finally, there are 20-50GB optical discs in the Blu-Ray spec coming out in lte 2006.

Given the pace of change, the future clearly belongs to magnetic hard disks.

The OSDL has released Portland 1.0, its set of common interface for GNOME and KDE. Because Portland will be found in many Linux distros, such as Debian, Fedora and SUSE, it could help solve some of the forking problems that we’re seeing on the desktop. Let’s hope that the Linux community embraced Portland, and that other distros climb on board — because that, in turn, will encourage third-party app development.

I spend a lot of time on airplanes — not as much as many of my colleagues, but it’s plenty. My default carrier is United Airlines, which has a hub in San Francisco, and which also has an e-mail flight notification system, called EasyUpdate.

(A digression: At SFO, United promotes the service by boasting, “Only United Offers EasyUpdate.” Well, EasyUpdate is a trademark of United Airlines, so of course only United has a service with that name. Is that a lame slogan or what?)

EasyUpdate sends two e-mail messages to you, your loved ones and your administrative assistant about each flight segment:

* Before takeoff, it confirms the projected departure time and gate
* Before landing, it confirms the projected arrival time and gate

But, interestingly, the EasyUpdate system is not updated if you change flights, and therefore, information transmitted by the EasyUpdate service can be obsolete.

To wit: Sometimes I arrive at the airport early, and squeeze onto an earlier flight. More rarely, sometimes my flight is cancelled, and the airline places me on an alternative flight.

EasyUpdate never knows about any of this.

To use a real-world example, today I was scheduled for flight #931 to LAX, but arrived at SFO early enough to hop onto flight #1171 instead. Before that, I had already received the first EasyUpdate e-mail, that flight #931 was scheduled to depart on time from SFO’s gate 84.

However, EasyUpdate never knew that I changed flights. So, half an hour after my flight #1171 had landed, EasyUpdate dutifully sent me and my loved ones (alas, I have no administrative assistant) a message that flight #931 was scheduled to land on-time at LAX’s gate 69A:

** UNITED AIRLINES ARRIVAL MESSAGE **

The following flight is scheduled for arrival:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Flight Number: 931
Departing From: San Francisco California (SFO)
Traveling To: Los Angeles (LAX)
Date: October 10
Gate: 69A (Gate information is subject to change)
Estimated Arrival Time: 8:12 p.m.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Flight times are subject to change. Please check the flight information monitors at the airport.

Thank you for choosing United!

No message was ever sent about flight #1171. That’s pretty worthless for a proported travelers’ real-time information system. From an altitude of 30,000 feet, though, it sounds like an easy problem to fix. Web services, SOA, real-time database orchestration, all that sort of thing.

Let’s see how long it takes.

Colin L. Powell, Gen. (Ret.), U.S. Army, former Secretary of State of the United States, is a very funny man.

Gen. Powell had the audience in stitches during his keynote address at Dreamforce 2006, the user conference held by software-as-a-service pioneer Salesforce.com today in S.F. The general riffed on a number of themes: he mocked President Bush’s “When I looked in [Russian president ] Putin’s eyes, I saw his soul” statement, pined after his government-issued Boeing 757 aircraft, laughed about his BlackBerry, and shared his wife Alma’s advice on working with the government.

The keynote address, which lasted about an hour, meandered. The general shared his views about the benefits of corporate philanthropy, the state of the world, conflicts in Europe, Asia, Africa and the Middle East. He reminisced about his training as a young lieutenant, and his first days at the State Department. He talked about looking into Mr. Putin’s eyes, and seeing the KGB. He talked about the economy, about terrorism, about security, about war, about peace.

I took a lot of notes and photographs during Gen. Powell’s talk (that’s my own picture above), but what struck me most were his comments regarding leadership.

Leaders, he said, don’t merely motivate their followers. They inspire their followers — and inspiration makes them motivate themselves, which is obviously desireable. In fact, he said (repeating advice given to him early in his career), if you’re a good enough leader your followers will follow you if for no reason other than curiosity about where you’re going to take them… with trust that the journey is worthwhile.

Gen. Powell talked about the attributes of a good leader. A leader, he says, has to have vision about what the enterprise is doing, and why it’s doing it. That doesn’t just mean the CEO’s overarching vision. A department head, a business-unit manager, a shop steward, even a team leader within a workgroup, all need vision if they’re going to lead.

But the leader needs more than vision (and the ability to communicate that vision). A leader has to be able to give his followers the tools they need to do the job, as well as the training needed to use those tools. Why? Because it’s the followers who get the job done.

A general might know that the army needs to take a certain hill, or protect a certain city, or set up camp in a certain valley. But it’s the troops who take that hill, protect that city and set up the camp.

A leader also needs to be willing and able to discipline the followers when they need it. Why? It’s important for morale and inspiration. If the good followers see that someone is being a bad follower and nothing is being done, then the good followers will say ‘why should I bother?’ and then it all collapses. “If you’re not doing that, you’re not a good leader,” the general said.

And finally, you need to do the right thing. “Everyone wants to be part of an organization that has high standards,” he said. I have no doubt that any organization that Gen. Powell is involved with will have the highest standards possible.

But would someone give this guy an airplane?

From: email hidden; JavaScript is required
To: email hidden; JavaScript is required
Subj: Request to mailing list Ximian-mono-list rejected

Your request to the Ximian-mono-list mailing list

Subscription request

has been rejected by the list moderator. The moderator gave the following reason for rejecting your request:

“[No reason given]”