Timbuktu Pro 8.0 Finally Adds Encryption — I’m not a nervous nelly, but it’s always bothered me that my Timbuktu Pro connections are being sent in the clear when I’m not using a virtual private network (VPN) connection. Finally, Timbuktu Pro 8, released last week, includes standard SSH (Secure Shell) support and even uses its built-in compression to enhance speed. SSH uses public-key encryption to exchange a one-time session key. As long as you know that the public key is valid – which needs to be established by what are called "out-of-band" methods of confirming a certificate’s identity – then you have something close to absolute assurance that no one can (currently) snoop your session.

<http://www.netopia.com/software/products/tb2/ mac/>

Timbuktu Pro 8 also uses Mac OS X accounts instead of requiring separate account management, has drag-and-drop file exchange from a remote shared window, allows push installations if SSH (the Remote Login checkbox in the Sharing preference pane) is turned on, and works with Rendezvous. A new single-user license costs $95, and a twin pack costs $180 (other pricing is available for multiple licenses); upgrades for current owners are 50 percent off the new license rate. [GF]

Stream Anything to an AirPort Express Using Airfoil — Rogue Amoeba has released the first version of Airfoil, a program that can take the audio output of any program under Mac OS X and stream it using AirTunes to an AirPort Express’s audio output jack. It’s a simple piece of software that will delight all AirPort Express owners who want to stream audio from applications such as QuickTime Player, RealPlayer, or Windows Media Player. However, due to latency between Airfoil and the AirPort Express, audio and video will not be synchronized, such as when playing a DVD and sending the audio to a home stereo; Rogue Amoeba has posted a possible workaround that’s worth trying. There’s no guarantee that Apple might not step in and update their AirTunes software to disable Airfoil, but it doesn’t seem to fall into the category of things that Apple believes is detrimental to their products, contracts, or partners. The software costs $25, but Rogue Amoeba is offering it at an introductory offer of $20 through 31-Mar-05. [GF]

<http://www.rogueamoeba.com/airfoil/>

Office 2004 for Mac 11.1.1 Update Improves Stability — Microsoft has released an update to the English and Japanese versions of Microsoft Office 2004 (French, German, Italian, and Swedish versions still to come). All Office programs gain improved display of inserted PICT images and correct importing of black-and-white scans from Xerox scanners. In addition, Excel 2004 features improved performance when calculating array formulas that include a user-defined function, PowerPoint 2004 better handles opening presentations with invalid font information, and there’s a fix that could cause Office 2004 programs to freeze when users with Adobe Acrobat opened documents containing forms created in Visual Basic for Applications. The 17.4 MB update is available via the Microsoft AutoUpdate utility (look in your Applications folder) or as a stand-alone download. [ACE]

<http://www.microsoft.com/mac/autoupdate/ description/AUOffice20041111EN.htm>

<http://www.microsoft.com/mac/downloads.aspx? pid=download&location=/mac/download/ office2004/update_11.1.1.xml>

I asked Apple to loan me a new PowerBook so I could test first-hand the hardware features they added in the latest refresh a few weeks ago: the scrolling trackpad, the Sudden Motion Sensor for hard drive protection, and increased backlighting for the keyboard. You can read about these features in Apple’s marketing materials, but it’s nice to test them first hand.

<http://www.apple.com/powerbook/>

Using the scrolling trackpad is more natural than it may sound. You use two fingers to gesture across the trackpad to simulate a mouse’s scroll wheel, which works horizontally as well as vertically; the sensor has no trouble telling the difference between one finger or two. (I’ve been waiting since college for 3D gestural recognition; a scholar-in-resident spent a year working on that, but obviously we’re not there yet.)

Apple says the keyboard backlighting is up to 10 times brighter than in the previous models, and, man, are they right. In a fully darkened room in the back of my office, I kept hitting the brighter-backlight function key and the room got brighter and brighter. It’s so bright, in fact, you’ll set it below maximum for most situations.

As for the Sudden Motion Sensor, which detects quick movement and locks the hard drive heads, you may ask, did I drop the PowerBook from a great height? Hey, this is a loaner, and I’m responsible for returning it intact. So, no. But I did shake it and drop it in my hands, and it surely did pause and restart the drive without a skip. For a more entertaining test of the Sudden Motion Sensor, see Amit Singh’s exploration of the sensor’s capabilities, including software that adapts to the PowerBook’s position (such as a self-adjusting window that stabilizes itself according to how the laptop is tilted).

<http://www.kernelthread.com/software/ams/>

Have you ever pasted a graphic into a word processing document and later wanted to update it? It’s a tedious process of opening the original file, making your changes, copying the new graphic, returning to the word processing document, deleting the old graphic, and pasting the new graphic. It doesn’t have to be like this – applications can share data in a rich fashion that enables communication between the two applications. In fact, within limited spheres, some applications already do this, usually within a suite of programs from a single company. But a new open source technology, called LinkBack, promises to bring data linking to more Mac OS X programs. Such a technology would be welcome, since Apple has made various failed attempts at providing such connectivity over the years.

If you’ve been around the Macintosh world long enough, you might remember Apple’s Publish & Subscribe technology. It appeared in System 7 back in 1991 and enabled you to "publish" data – a picture, some text, a chart – from one application and "subscribe" to it from another application – in essence, to insert a live copy of the published data into another document. That way, if you changed the graphic in the publishing application, the subscribed document would automatically receive the changes. Publish & Subscribe was a nice idea, but as late as 1994, I was commenting in TidBITS that it was a failure due to a poor implementation and minimal support from developers.

<https://tidbits.com/getbits.acgi?tbart=01980>

Then there was OpenDoc, a technology Apple introduced at the Worldwide Developer Conference in 1994 and which became real in late 1995 and early 1996. Apple never did a good job of explaining OpenDoc, but in essence it enabled a document-centric interface in which small modules – potentially from different companies – combined to provide the power of a monolithic application. The theory was great: you could put together exactly the word processor you wanted by adding together the best Find module, and the best Table module, and so on, and they would all fit seamlessly into the same interface. Despite the popularity of Apple’s OpenDoc-based Cyberdog program (an integrated Internet client) and support from a few companies like Nisus Software, the reality never matched up to the theory, and Apple put OpenDoc and Cyberdog into "maintenance mode" in 1997. Interestingly, the OpenDoc community tried to negotiate a "stewardship agreement" for the OpenDoc Development Framework in exchange for Apple continuing to ship OpenDoc, but the deal fell through when the vice president who had agreed to this left Apple.

<https://tidbits.com/getbits.acgi?tbart=02260>

<https://tidbits.com/getbits.acgi?tbart=02239>

<https://tidbits.com/getbits.acgi?tbart=01487>

<https://tidbits.com/getbits.acgi?tbart=01245>

<https://tidbits.com/getbits.acgi?tbart=04711>

While Apple was working on the doomed Publish & Subscribe and OpenDoc, Microsoft developed OLE (Object Linking and Embedding), which remains in use within Microsoft applications today, and NeXT created Object Links in 1995.

<http://www.channelu.com/NeXT/NeXTStep/3.3/nd/ ReleaseNotes/ObjectLinks.htmld>

LinkBack, an open source technology jointly announced by Nisus Software, The Omni Group, and Blacksmith, is a step in the direction of providing system-wide data linking again. With LinkBack, which will appear in future versions of Nisus Writer Express, OmniGraffle, OmniOutliner, Chartsmith, and Stone Create, if you want to edit a pasted graphic, you’ll instead just double-click the graphic to edit it in the original application, after which your changes will automatically be reflected in the destination document. The pasted data doesn’t have to be a graphic; it could also be text, such as stock quotes that you want to update automatically.

<http://www.linkbackproject.org/>

Of course, it remains to be seen just how well LinkBack works, and in particular, how well it avoids problems that have bedeviled all of these other data linking technologies in the past. Obviously, widespread support is tremendously important, since users won’t even think about LinkBack unless it’s widely available. The open source nature of the project should aid in adoption, especially since developers burned by Apple in the past won’t worry that LinkBack will be at the mercy of a single company. It’s also essential that LinkBack be reliable and easy to use, or the technology will have difficulty garnering an audience.

So if you’re a developer, give LinkBack a look. Just because Apple’s heavyweight data linking technologies have failed in the past doesn’t mean the rest of us couldn’t still use a good solution now.

When Apple thrust Mac OS X upon us, it was quite a change. I remember one colleague remarked that his head was filled with all sorts of Mac OS 9 troubleshooting arcana, nearly all of which would be rendered moot once Mac OS X gained its footing. Some behaviors in the new operating system changed enough that they disrupted the flow of how we’d been using the Mac for years. Subsequently, several utilities appeared to bring back those behaviors (see "Top Mac OS X Utilities: Restoring Mac OS 9 Functionality" in TidBITS-622).

<https://tidbits.com/getbits.acgi?tbart=06763>

For the most part, my transition to Mac OS X went smoothly without relying on such utilities. To my surprise, it didn’t take long for me to adapt to the new Mac OS X order. However, one thing remained an irritation, an aspect I actually forgot about because long ago I installed a utility that fixed it: Mac OS X’s default window behavior.

Trouble Begets Frustration — When my PowerBook recently started behaving strangely, I went looking for startup items that might be contributing to the problems I was seeing. One utility I disabled was ASM (Application Switcher Menu) 2.0.2, a utility that provides an application menu like that found in Mac OS 9. Although ASM didn’t appear to be the cause of my problem, I quickly realized it offers a feature that I absolutely cannot live without: when I click a window belonging to an application (such as Eudora or the Finder), ASM causes all of that program’s windows to be brought to the front. Normally, in Mac OS X, clicking a window causes just that window to move to the front; other windows in that application remain unaffected.

<http://www.vercruesse.de/>

In fact, that’s the only reason I installed ASM in the first place; I don’t actually use the application menu. But being without this windowing feature drove me crazy for several days.

Mac OS X’s default window behavior is nutty. I keep four Eudora windows open at once: my In box, my list of mailboxes, the Task Progress window, and the Filter Report. If I’m in another application and I want to switch to Eudora, I want to see all four windows, not just the one I clicked on.

One response to this behavior has been Apple’s move toward single-window interfaces, such as iTunes, iMovie, and iPhoto. But other Apple software can’t be confined to one window: Final Cut Express, which uses at least four main windows, includes a preference to bring all windows to the front on activation – a different workaround to the problem. (Perhaps inspired by this example, other developers could add a similar preference to their applications.)

I could click the application icon in the Dock, but that’s a mouse-trip to a small target on the other side of my screen, and it’s not always what I want; clicking the Finder icon, for example, creates a new Finder window if none existed beforehand. But Mac OS X would prefer to layer windows like shuffled cards, with windows acting as separate entities instead of as groups of applications. And having a Bring All to Front command in the Window menu of every program isn’t helpful.

What’s needed is a simple preference that enables me to specify whether all of an application’s windows come forward when the program is brought to the front. In the meantime, several utilities fill that particular gap.

X-Assist — Despite my earlier comments, this isn’t an article about ASM. Although I haven’t had problems with it, the last freeware version of it is now a few years old. A 2.1 beta version is available as $15 shareware, but it currently has issues with Mac OS X Panther and doesn’t appear to have been updated in over a year. If ASM were the only solution around, I’d happily pay for it, but I don’t want to inherit problems.

Instead, I poked around online and found Peter Li’s X-Assist, which seems to offer many features similar to those in ASM, such as a Mac OS 9-style application menu and a hierarchical menu to access System Preference panes. It also features a plug-in architecture for add-on capabilities and a list of recent applications, but frankly, I turned off all these other features. X-Assist brings my windows to the front the way they should behave, and that’s all I want. Even better, the software is free, and, although its version number is 0.7, seems to be rock solid.

<http://members.ozemail.com.au/~pli/x-assist/>

Other Solutions — Shortly after an abbreviated version of this article appeared on ExtraBITS, several readers wrote in to either defend the Mac OS X window behavior or to recommend other utilities that provide the same functionality I’ve found with X-Assist. Surprisingly, I haven’t been able to find a utility that only brings application windows to the front in groups. Typically, it’s a preference added to other useful features in programs such as Proteron’s LiteSwitch, TLA Systems’ DragThing, and Peter Maurer’s Butler, among others. Most also have an option to disable the window preference temporarily if you want to use the regular window behavior.

<http://emperor.tidbits.com/.3c5a36d9>

<http://www.proteron.com/liteswitchx/>

<http://www.dragthing.com/>

<http://www.petermaurer.de/nasi.php?thema=butler &sprache=english&kopf=labor>

Ultimately, this is another example of how one person’s preference is another person’s irritation. It was pathetic that I would get angry at my Mac whenever I switched applications because of what I perceive to be brain-dead window behavior. But other people I’ve corresponded with over the past week have clearly expressed their relief that windows now operate as independent elements. To each his or her own, I suppose, and if there’s a moral to the story, it’s perhaps that Apple should, in the very first versions of Mac OS X, at least made the Mac OS 9 windowing behavior an option, if not the default, to reduce the annoyance for those accustomed to the older behavior. Alas, the horse has left that particular barn long ago, but at least there are plenty of third-party utilities for restoring this behavior.

There are many things in the world that you feel to be true, but you’re not exactly sure why. So if you’re a thinking person, you’re left with this nagging suspicion that you should be better able to come up with a better explanation than "But it’s just wrong!"

For many people, myself included, digital rights management (DRM) technologies fall into this category. Even if we have no intention of breaking copyright law by downloading music or movies willy-nilly, and even though many of us earn our livings through the production and sale of copyrighted material, we’re still offended that the entertainment and media conglomerates of the world – the Content Cartel, as one commentator has labeled them – are pushing so hard to ensure that every song, every movie, every television show, is wrapped up tight in some form of DRM that controls access to the content and use of it.

Thanks to a talk by Professor Dan Burk of the University of Minnesota Law School that was organized by Cornell University’s Information Science Department, I have a significantly better sense of just why DRM makes my skin crawl. If you’re generally interested in the topic of DRM and the law, I encourage you to read the draft paper on which Professor Burk based his talk.

<http://www.infosci.cornell.edu/about/Feb02.html>

<http://www.infosci.cornell.edu/about/burk.pdf>

Legal Rules versus Legal Standards — As Professor Burk explained, the law is broken down into two basic aspects: rules and standards. A legal rule is a specific imperative in which all the thought surrounding the details of the law takes place ahead of time. In theory, at least, with a legal rule, the body establishing the rule deliberates on specifics such as boundaries, exceptions, penalties, and so on, and for violators of the resulting law, there is no leeway for interpretation. For instance, consider a drug possession law that states that offenders caught with more than 5 grams of marijuana must serve a 3 year prison term. If some stupid pothead kid falls into that category, regardless of any other circumstances, it’s off to prison for 3 years.

Contrast that with a legal standard, which essentially posits a goal and lays down some guidelines for defining illegal behavior, but which leaves significant room for interpretation. So, instead of a rigid law stating exactly what behavior is considered illegal and mandating specific punishment, a law based on a legal standard would declare that drug possession was illegal, but would leave discretion in the hands of the judge as to whether the crime warrants a lesser punishment (in the case of the pothead kid) or greater punishment (in the case of a known drug dealer caught with a kilo of heroin).

I’m no legal scholar, but from a common sense standpoint, I think most people would prefer legal standards to legal rules. After all, laws are created by politicians; would you trust a politician – even one of the honorable ones – working with hypothetical "what if" scenarios to define a crime and a punishment? Or would you prefer that cases be decided by a judge with the actual facts of a specific case at her fingertips? Consider a law that most of you have probably broken in the last few days – the law against speeding. Would you prefer a law that said being caught driving over the speed limit was grounds for an automatic $200 fine, or one that gave the police officer and the traffic court leeway to see that driving a seriously injured person to the emergency room was grounds for dismissal?

As Professor Burk pointed out to me in email subsequently, some people do prefer rules to standards for the simple reason that the rules are predictable, so you know what to expect beforehand. He also noted that some people also become concerned about judges having too much power, although it seems to me that most of the people who complain about "judicial activism" are politicians, and are bent out of shape about having competition.

DRM: Them’s the Rules — Let’s step back a moment. Creating a law is only one of many ways that societally acceptable behaviors can be encouraged. If society’s overall goal is for people to drive more slowly and cautiously, putting speed bumps in the road would have the same effect, as would keeping the road and shoulders narrow. Of course, those strategies have other downsides, such as slowing down ambulances or making it difficult for fire trucks to maneuver, and they don’t absolutely prevent the unwanted behavior, they just discourage it. You can still drive quickly over speed bumps or along narrow roads. In this respect, such extra-legal strategies are akin to legal standards – they leave some wiggle room in the system.

DRM technologies fall roughly into this category of extra-legal methods of encouraging behavior, but there’s at least one important difference: DRM, like all technology, is an embodiment of a legal rule, not a legal standard. It’s simply impossible to create a DRM technology that can evaluate and approve exceptions, no matter how reasonable or legal they may be. If you want to play a song purchased from the iTunes Music Store without stripping the DRM, you must use an iPod or iTunes on an authorized machine; there’s no wiggle room at all.

This is a big deal because the law that DRM instantiates is copyright law, and copyright law is distinctly a case of a legal standard. Copyright law allows all sorts of exceptions, including fair use, reproduction by libraries and archives, and musical performances at agricultural or horticultural fairs (I wonder how much that last exemption cost?). Plus, in any copyright infringement case, the judge would have to take into account what was copied, how it was copied, what the intent was in copying, and the harm done to the copyright owner in the marketplace. No matter how hard the Content Cartel tries to conflate the two under the rubric of "piracy," there’s a big difference between the downloading of a song from Kazaa and the burning and reselling of thousands of DVDs of the latest Harry Potter movie.

<http://www.copyright.gov/title17/92chap1.html>

So now you can see why DRM rubs so many people the wrong way. It’s turning copyright law, which is at its heart a reasonable legal standard, into a legal rule with no ifs, ands, or buts.

Permission and Forgiveness — There’s another aspect to the way DRM stands in for laws. No matter whether we’re talking about legal rules or legal standards, you’re still free to do whatever you want and then ask for forgiveness if you’re caught. As a result, many violations of the law are never noticed, and many others never make it to court because the cost to society of enforcing them is higher than the benefit (a police officer can make the decision that it’s more important to get that injured person to the hospital than it is to enforce the speed limit).

However, the corollary to this fact is that our laws thus reach further than we intend. Exceeding the speed limit at any time is technically a violation of the traffic laws, but no one really believes that enforcing the speed limit is so important that cars should automatically inform the police whenever you are speeding. Similarly, every unauthorized copy of a digital media file is technically an infringement of copyright law, but few people outside the RIAA probably believe that every iPod owner should be hauled into court to justify copying music from a Mac to an iPod under fair use.

So in the real world, we’re used to asking for forgiveness after committing actions that are technically in violation of a law (and frankly, we’re used to getting away with a lot of violations that are too trivial to justify enforcing). In the digital world, however, DRM inverts this system, forcing us instead to ask for permission rather than forgiveness. Anyone who has ever been a teenager knows just how problematic that is – parents seldom agree to the cool stuff. When it comes to technology, the end result of being forced to ask for permission is that experimentation and innovation are stifled. If the original Napster and the other peer-to-peer file sharing networks hadn’t scared the hidebound music industry silly, do you think they would ever have agreed to Apple creating the iTunes Music Store?

Because most DRM systems start from the written copyright law and prevent any behavior that would technically be an infringement, they not only fail to account for the exceptions in copyright law, they also ignore our societal expectations about how laws should work in practice. It would be like car manufacturers outfitting all cars with limiters that could determine the posted speed limit on any stretch of road and prevent the car from driving faster than that, for any reason. Talk about grounds for a revolt!

Room to Move? In fact, there is a little wiggle room with DRM-protected content like songs from the iTunes Music Store, and that’s the fact that pretty much every piece of DRM technology has been broken. According to Professor Burk, the peer-to-peer tracking company BigChampagne has found that it takes about 4 minutes after release for a song using copy-prevention technologies to appear on the file sharing networks. So you could purchase a song from the iTunes Music Store, remove the FairPlay DRM in any one of a variety of ways, and use it in some way that would otherwise be impossible.

But there’s a problem with creating your own wiggle room by breaking a DRM technology: our old friend the DMCA (Digital Millennium Copyright Act); see "The Evil That Is the DMCA" in TidBITS-656. The DMCA distinguishes between access of content and usage of content (though it’s a relatively fuzzy distinction), and forbids any circumvention of access control technologies. However, the DMCA does not forbid the circumvention of usage control technologies; the thought is that this was the loophole Congress left to allow fair use of material that you had legally purchased. However, the problem is that the DMCA also bans the supplying of tools to circumvent either access or usage control technologies. In short, you can legally break any usage control technologies you want, but you can’t get any help doing it, nor can you create tools for anyone else to do it. Needless to say, this is a barrier which essentially no one can cross legally.

<https://tidbits.com/getbits.acgi?tbart=06997>

There is some hope that the courts have recently seen the danger behind the DMCA. In his talk, Professor Burk called out a pair of cases where appellate courts had ruled against plaintiffs brandishing the DMCA. In one case, Chamberlain v. Skylink, Chamberlain sued to prevent Skylink from reverse engineering the codes necessary to make Chamberlain’s garage doors open; Skylink was reverse engineering the codes for use in a universal garage door opener. The court ruled that Congress had no such anti-competitive behavior in mind with the DMCA. And in Lexmark v. Static Control, the court ruled that Lexmark could not use the DMCA to prevent Static Control from reverse engineering the chips necessary to create off-brand toner cartridges for Lexmark printers.

<http://www.eff.org/legal/cases/Chamberlain_v_ Skylink/>

<http://www.eff.org/legal/cases/Lexmark_v_Static _Control/>

The moral of this story, if there is one, is that DRM technologies are more subtly pernicious in their effect than may be apparent from first glance, due to the way in which they embody legal rules and eliminate the human effect in determining how copyright law should be interpreted and enforced. That realization does little to assuage the annoyance many people feel when their lives are unnecessarily complicated by DRM, but at least it puts into words why DRM is so often annoying, not to mention concerning for the future of technological experimentation and innovation.

PayBITS: If Adam’s article helped you understand how DRM
is undermining copyright law and why that’s concerning,
consider a donation to the EFF.
<http://eff.org/support/>

Read more about PayBITS: <http://www.tidbits.com/paybits/>

The second URL below each thread description points to the discussion on our Web Crossing server, which will be faster.

Shared Database Solutions — When one reader’s organization starts feeling growing pains, what’s the best way to consolidate information to avoid having several databases on multiple machines? (9 messages)

<https://tidbits.com/getbits.acgi?tlkthrd=2501>

<http://emperor.tidbits.com/TidBITS/Talk/359>

Delayed password disclosure technique — A new method of online authentication could help avoid stolen passwords on the Internet. (8 messages)

<https://tidbits.com/getbits.acgi?tlkthrd=2500>

<http://emperor.tidbits.com/TidBITS/Talk/358>

Timbuktu 8.0 Finally Adds Encryption — Readers discuss Netopia’s latest version of its remote administration software, including the new Push Install capability to upgrade Timbuktu easily. (5 messages)

<https://tidbits.com/getbits.acgi?tlkthrd=2498>

<http://emperor.tidbits.com/TidBITS/Talk/356>

Cleaning House in iTunes — Adam’s article about culling duplicates from his iTunes library prompt other solutions from other readers, including synchronizing multiple libraries and changing song information for several songs at once. (7 messages)

<https://tidbits.com/getbits.acgi?tlkthrd=2497>

<http://emperor.tidbits.com/TidBITS/Talk/355>

Simple hosted CMS like Site Crossing — Readers suggest other company-hosted content management systems similar to Web Crossing Inc.’s new Site Crossing service. (16 messages)

<https://tidbits.com/getbits.acgi?tlkthrd=2496>

<http://emperor.tidbits.com/TidBITS/Talk/354>

iPods Defeating Insomnia — Last week’s article about how Adam and Tonya fall asleep using their iPod struck a chord with a few people, including one who recommends a special pillow speaker, and another who explains the mechanism for how audio books can put you to sleep. (3 messages)

<https://tidbits.com/getbits.acgi?tlkthrd=2495>

<http://emperor.tidbits.com/TidBITS/Talk/353>

Multiple iTMS authorizations — A couple of readers run into trouble when combining songs from the iTunes Music Store purchased under two different user accounts, while others don’t seem to be affected. (4 messages)

<https://tidbits.com/getbits.acgi?tlkthrd=2494>

<http://emperor.tidbits.com/TidBITS/Talk/352>

Third-party DVD Burner — Suggestions are offered to someone looking to buy a non-Apple DVD burner, both internal and external models. (4 messages)

<https://tidbits.com/getbits.acgi?tlkthrd=2493>

<http://emperor.tidbits.com/TidBITS/Talk/351>

SMTP server while travelling — When a reader relocates to Beijing, she runs into trouble sending email through her old SMTP server. TidBITS Talk to the rescue! (15 messages)

<https://tidbits.com/getbits.acgi?tlkthrd=2492>

<http://emperor.tidbits.com/TidBITS/Talk/350>