Recently in Computers and Technology Category

Kevin O'Keefe recently related that virtual storefronts in Second Life for real-world enterprises seem to be shuttering (well, not literally — in Second Life closed businesses just vanish, leaving bare ground behind).

No one who has actually spent time in Second Life should be surprised. The creations of Second Life's more ordinary users are far more interesting than those of real-world companies.

One of the more intriguing features of Second Life is that it allows users to design, program and trade objects amongst themselves. Everything from avatar accoutrements (clothing, hair, jewelry) to buildings, vehicles, plants — the whole Second Life world is populated with the creations of its users, fueling an entire in-world economy lubricated by a so-called "limited license right to ... participate in the virtual economy" of Second Life called the Linden Dollar (L$).

The creations of Second Life users are meant for use in Second Life. They advance the stories that people play out when they use Second Life. They add to the social context of Second Life. By contrast, real-world companies' presences in Second Life tend to look like nothing more than corporate advertising. And not just any advertising at that, but advertising that has no social role in that context. IBM may buck the trend in that it gets a lot of traffic in its Second Life regions, but that's primarily because the software running those areas runs on high-powered, well-connected servers, features Second Life residents can use to their advantage when designing and creating objects. It doesn't make much sense to venture into an artificial world to advertise products that are useless there.

Here's an interesting idea — a keyboard without letters printed on the keys. The idea is to force the user to look away from the keys and thus memorize their positions, making it possible to type faster.

Well, that's what they say the idea is, but I'd guess that most of the purchasers will be computer users who would take especial pride in the ability to type using a letterless keyboard.

I wouldn't have predicted it, but Apple Computer has released a public beta trial of a dual-boot system tentatively named Boot Camp that can launch Windows XP or Mac OS X on the new Intel-based Macintoshes. It includes Mac-specific drivers for Windows. Of course, you still have to provide your own copy of Windows XP SP2, because Apple isn't going to sell you that.

This means Apple is listening to its customers, especially those "power users" who until now were experimenting with ways to make Windows run on Intel Macs themselves. I don't know if we'll ever see Apple providing technical support for those who want to run Windows on Apple computers, but at least it's interested in providing the capability.

[Via several sources more or less simultaneously, including Blurbomat and Pages in the Ether.]

Tim Bray ponders switching away from Apple equipment. His major gripe is with Apple's secrecy policies, but he also weighs the technical pros and cons of Unswitching that he would be likely to encounter.

For now, I'm still planning to buy an Apple Powerbook for my next notebook computer, but Apple's secrecy and legal policies have caused some serious cognitive dissonance for me.

A VC (that's the name of the blog) observes, "Apple survived the WinTel dominance in the PC business by becoming a "we" company. … But in the past couple months, Apple has made some very "they" company decisions." He outlines a few.

A commenter on that post responds,

As a long time Apple developer back in the early 80's, the common comment has always been, "Love the computers, hate the company" when it came to Apple. You might think that Apple switched from a "we" company to a "they" company, but those close to the company have always thought of it as a "they" company. … Apple's continued pursuit of the bloggers is a direct reflection of how Apple (Jobs) ISN'T in tune with their user base out in the real world. Apple has never understood the passion of it's users, though some employees over time have. (ie. Guy Kawasaki)

There's more. Have a look.

Meanwhile, Robert Scoble loathes the notion that Microsoft is a "they" company. His perception from inside the company is that they're not.

I've long felt that Microsoft was focused on itself and took an exploitative attitude toward its customers. But recently, I've encountered a remarkable counteragent to that perception in the form of the many bloggers at Microsoft. They help humanize this consumer's perception of an apparently monolithic corporation that wields great power over a great many consumers. I'm still skeptical of much of Microsoft's management and some of its tactics (I mean, come on — patenting storing word processor files in XML?), but it still helps to hear from developers about their genuine enthusiasm for putting tools in the hands of other developers and consumers.

The Mercury News is running an editorial about (Cal.) Superior Court Judge James Kleinberg's ruling in Apple v. Does just over a week ago. [Via Donna Wentworth.]

Let's be careful with the extent to which we call this case a "precedent." It has no authoritative precedential force within the law, because it comes from a trial court. But there is a possibility that an important appellate court could agree with the reasoning, establishing more authoritative precedent. Or, other trial courts could simply find the reasoning persuasive in the absence of other more binding authority.

When should website operators who post information they receive about business activities be able to claim either:

• A privilege not to disclose the identity of their sources?
• An immunity from civil suit or criminal penalties for disclosing information?

I'll call the first of these a "journalistic privilege" and the second "journalistic immunity." In this post I'm just going to talk about privilege, but it helps to remember that we are talking about at least two kinds of protections. Still a different kind of privilege protects publishers from prior restraint in most cases.

Ernest Miller (The Importance Of....) has an idea for an answer to those questions above. It's pretty simple. But it may be too simple, and efforts to refine it tend to lead to some of the more complicated inquiries that I think he'd prefer to leave out of the test.

Mr. Miller argues in this post for a standard that would make it very easy to obtain legal protection for disseminating information one receives from business or government insiders. On Mr. Miller's view, we should not focus on the speaker, the content of the speech, or the medium of communication. See his post for a quite articulate summary of that part of the argument. (Mr. Miller cites Linda Berger, Shielding the Unmedia: Using the Process of Journalism to Protect the Journalist's Privilege in an Infinite Universe of Publication, 39 Hous. L.R. 1371 (2003)). Mr. Miller and Prof. Berger argue for approaches where what counts is whether the speaker is "engaged in the process of journalism."

Miller and Berger argue for different standards. Prof. Berger provides a pragmatic (but still a bit fuzzy) outline of what sorts of behaviors and attributes characterize the "journalistic processes" that law should protect. Mr. Miller argues that Prof. Berger's standard should really be stripped down to just two or three elements. I'm going to focus on his argument below and leave Prof. Berger's aside, mainly because I haven't had time yet to give her article the closer reading it deserves.

Prof. Berger's article also provides a very handy historical summary of journalistic privilege (or its absence) under federal and state law. Note that she and Mr. Miller are arguing for reformulations of privilege law — that's really another way of saying "new law". I'm not "dissing" either's argument, which may be sound and valid from the perspectives of First Amendment values — specifically, the free flow of information — and jurisprudential efficacy. I'm just saying their arguments are not what the courts have been doing so far.

So how should we tell who is "engaged in the process of journalism?" Mr. Miller writes, "The 'press' and journalism boils down to two things: gather information and publish it publicly." So, he asks, "Isn't the only process we need to know about is that information was gathered and then it was publicly published (or there was intent to publicly publish)?"

If I'm reading Mr. Miller's post right, the elements of the journalistic privilege would go like this:

If

• Insider communicates information to Outsider; and
• Outsider
  • publicly publishes the information or
  • intends to publish it publicly;

• Outsider cannot be compelled by law to identify Insider.

It looks simple, but there are still some tricky parts.

For example, there's still some question what it would mean to "publish" information. Assume that any widely disseminated medium counts as publishing — including publicly accessible websites. That does away with the blogger/journalist issue. But the bigger problem word in this test is "publicly." Is it "publicly publishing" if Outsider only shares the secret information with his or her family? What if Outsider puts together a "publication" that consists of a for-profit newsletter distributed only to a trade group? Or only to certain competitors of Insider's employer? When does the further dissemination of the information become a "public" publishing?

Mr. Miller's test also has a subjective element, in that Outsider's subjective intent to publish would qualify Outsider for the privilege. Suppose we have an Outsider who has engaged in some of the more questionable practices described two paragraphs above. When Outsider gets a subpoena (because yet someone else leaked Outsider's newsletters or a copy of Outsider's private website), is it enough for Outsider to say, "I planned to release that information to the public at large next week"?

The hunt for a description of "public" smells to me like a search for the kinds of journalistic processes that we consider to be more for the benefit of the public at large than for select groups. If we go down that path, it dredges up the questions of disclosure, journalist track-records, and other scrutiny of Outsider that Mr. Miller's proposed test aimed to avoid in the first place.

Even if these questions show up difficulties in Mr. Miller's proposed test, he could quite aptly point out that it's often better to have one hard-to-apply word or element in a law than to have two, five, or ten, especially when that law is designed to protect the flow of information and public discourse. Maybe that's true, or maybe this topic is trickier than it looks so that we may have to put up with more complexity. Based on what I've read so far, I suspect Prof. Berger's approach admits of more of that complexity. On the other hand, her approach may be more difficult to apply, and she may overly favor "established media" outlets.

All that, and we haven't even reached the core policy issues, though I'm trying to illustrate some of them with the hypothetical questions I've asked in the last few paragraphs. You can find more detailed discussions elsewhere (Prof. Berger's article might be a good place to start).

Notice that Insider still gets no special protection. That's the same as in current law. If Insider's employer can identify Insider through other means, Insider is still in trouble unless some kind of "whistleblowing" law applies. Regardless of how we treat the journalist, I don't think the rules that apply to Insider should change. Except in extreme circumstances, no business should be compelled to employ someone who leaks private business information.

If you're interested in the topic and you haven't read Mr. Miller's post already, have a look at it and think about it for a while. If you're really interested in the legal background, also look at Prof. Berger's article.

Peter Burrows, BusinessWeek's Computer department editor, has a column describing why Apple's choice to sue fan-driven websites is a bad business call that "could tarnish the Apple brand." ("Memo to Apple: Lay Off Your Fans")

There's a certain irony to this. Remember the Orwell-inspired ad that ran during the 1984 Super Bowl, that one that depicted Apple as the free-thinking company breaking the Big Brother dogma of the computer industry? Through conscious marketing efforts, Apple has long stood for innovation and creativity, and has been embraced creative sorts like artists and writers -- not people who usually care for big corporations pursuing lawsuits against 19-year-olds. That may vastly oversimplify the nuanced legal issues Apple's lawyers are trying to address, but it could well be the overriding impression that results.

A large QuickTime .mov file of the 1984 advertisement is available here. There are smaller copies floating out around the internet.

By now, everyone who reads law and technology related weblogs is aware that a California court has permitted Apple Computer, Inc. to issue subpoenas to weblog author-editors and ISPs. Apple aims to discover who leaked information about an Apple project called "Asteroid." The court order rejecting the webloggers' motion to quash the subpoenas is here [via EFF].

To comply with the subpoenas and order, the webloggers will have to disclose the sources underlying the articles they published. The webloggers are not currently defendants in the lawsuit — the defendants are the people from whom the webloggers received the information in the first place. Apple might, under the circumstances, be able to join the webloggers as defendants for disseminating trade secret information. But it currently seems that Apple's primary goal is to rout out its employees who leaked the information, firing them and maybe (or maybe not) maintaining its suit against them.

Here's a very non-exclusive list of what some people have had to say about this case in the legal and technical blogosphere since the court released its ruling:

• Denise Howell (Bag and Baggage): Links and information.
• SoCal Law Blog: Brief comment. (I would not normally endorse a weblog so drowned in advertisements that one can scarcely find the text. –tph)
• JMoore (JurisPundit): JMoore was hoping for some analysis on the role of blogger-as-journalist, an issue the judge sidestepped by holding that no journalistic privilege at all applied in the case.
• News release, Electronic Frontier Foundation. The EFF provided legal support to the webloggers.
• Donna Wentworth (Copyfight): Donna criticizes the judge's determination that no First Amendment journalistic privilege applied in the case, and links to—
• An exchange [one] [two] between Susan Crawford and Ernest Miller on the proper scope of First Amendment journalistic privilege.
• Dan Gillmor worries that if this ruling is upheld on appeal (assuming there's an appeal), business reporting will become too dangerous for reporters.
• The Mercury News's Mike Langberg asks Apple to reconsider its legal approach toward enthusiasts because it communicates arrogance and attacks people who love Apple and its products.

It seems to me that the main thing that really bothers Apple enthusiasts about these subpoenas (as well as some other claims Apple has been pursuing against enthusiasts who spread leaked information) is that they convey the impression that Apple wants to punish webloggers who are enthusiastic about its products. This impression relates to Langberg's argument in his editorial. As I've followed the news and commentary, I've been wondering about Apple's business decision to proceed with these lawsuits more than the nuances of the legal doctrines that apply (though the latter are important). I have to admit that my first impression of Apple's lawsuits was that they seemed eager to stifle the very kinds of enthusiasts who provide valuable word-of-mouth marketing for their products — the kind of marketing that you just can't buy.

Apple hasn't even tried to justify its strategy to the public, and if it tried to do so using one of those "we just have to protect our intellectual property in all cases" arguments, I'd be sorely disappointed. Still, I've tried to imagine why Apple might want to pursue these cases knowing that there is some risk that it might alienate some enthusiasts. What might the people in charge of this policy at Apple be thinking? My first impression is that Apple likes word-of-mouth and media hype for its products, but it really wants to be able to control that hype itself. Is that just the symptom of obsession for control, or can one justify the desire for control?

In some circumstances, Apple might want to conceal new product plans because it wants the maximum lead time in the market before competitors could develop a competing product. But this case doesn't look to me like a situation involving intense competition in a particular market space. The secret information at issue in Apple v. Does relates to an improvement to theGarageBand music sequencing software in the iLife software suite.

Maybe there was a major concern about competitive advantage for Apple in this case — I don't pretend to know all about the market for music sequencing software. But it looks to me like Apple is trying to enforce a blanket policy regardless of whether this particular leak is comparatively a "big deal."

Concerns about employee discipline provide one potential justification for pursuing "leakers" even if the leak didn't cause severe damage. Assuming that Apple has a blanket policy of secrecy — and it has had just such a policy since the 1970s — it probably wants to be very sure that it employs only people who can honor that policy, and it probably wants to fire those who don't honor it. It's probably wrong to label such firings "retaliation," since Apple has reasons other than spite for discharging people who break their agreements. For example, distributing certain information about forthcoming Apple products may in fact cause Apple harm if it tips off competitors. Also, leaking information about the company's projects may hurt morale, given that Apple has many employees who do act very cautiously to protect information about its development efforts. They probably don't want to feel that their own co-workers are disseminating the very information that they are trying to protect. I don't know how much of a problem that is because I don't know anyone who works at Apple.

I think Apple should have some right — probably a right with a lot of enforcement leverage — to demand confidentiality from its employees. I also think it should be able to use legal tools to do so. I'm just not sure where the line should be drawn as a policy matter. Maybe those tools should be limited, or maybe they should have a broad reach except in those cases that would classify as "whistleblowing."

I've seen a lot of discussion about whether the webloggers affected by subpoenas in Apple v. Does should be treated as journalists. Most people assumed the court's ruling would focus on that question. But the court concluded that regardless of webloggers' status, disclosures by Apple's employees about Apple's forthcoming products didn't implicate a journalistic First Amendment privilege at all.

That conclusion raises the question, When do communications to a journalist (however that word may be defined) about a company's business or research activities obtain a privilege such the journalist cannot be compelled to identify the source of the information? That question is probably more important than the journalist/weblogger question, and it's the core of a conversation that's started over the weekend (see, e.g., Crawford and Miller), and it's an important one.

I'm not sure where I stand on that issue, and if I have enough time this week I'll play around in this space with some possible arguments related to that question. I don't want a policy under which companies designate everything under the sun as "trade secrets" to prevent dissemination, but I also don't want a policy under which employees can feel free to share whatever company secrets they want as long as they're talking to a journalist.

In the meantime, I'll look forward to reading more arguments about the legal standards that apply and about Apple's policy in particular. I'll be especially curious to read any arguments people might test out in support of Apple's policy, since I didn't exactly work philosophical or rhetorical wonders in its favor earlier in this post. Before we can really evaluate the legal standards that apply, I think we need to try to understand why Apple might want to draw close the cloak of secrecy. If we don't give serious attention to that viewpoint, the perspective of the business that generates the information that it then seeks to protect, our discussion of the legal standards will be under-informed and imbalanced.

I recently bought on eBay a Hewlett-Packard Scanjet 6200c scanner. A consignment shop was selling it, but I knew who the owner was and trusted that the equipment would probably be in good shape. It was, when it finally arrived. I can't say much in favor of the consignment shop; they took five days to get the product out the door after I'd paid for it, and they didn't respond to e-mail inquiries. I also think they could have marketed the product better, though any weakness there worked out to my advantage.

The only problem is that Hewlett-Packard's driver and scanning software for the 6200c is horrid. So I'm going to hold it up as an example of bad design from installation through (a brief cycle of) operation to failure, focusing on human-computer interface and error reporting. I will also criticize HP's support website for the product. I only have limited experience in which it has actually worked, so that will be a very short discussion. I'll be focusing mostly on — you guessed it — failure conditions. (I'll note that the hardware itself appears, from what little I have been able to accomplish with it, to be quite decent.)

Installation

I downloaded the driver package from Hewlett-Packard's support website. It's a self-extracting ZIP file with the very descriptive filename "sj673en.exe." I double-clicked on the file. I got an error message.

Problem: The self-extracting ZIP file automatically tries to extract to C:\sj673en. I don't have a C: drive. (My system drive is M: and my data drive is D:. Don't ask why, because I don't remember.) Even if I did, what gives HP the right to decide where on my system the file should be extracted, especially when it doesn't delete the extracted files afterwards? (And I wouldn't want it to delete them if I had control over the extraction.) What HP should do: If you're going to distribute in the self-extracting ZIP format, don't bypass the dialog that lets the user decide where to extract the files. What I did: Temporarily mapped a shared folder to drive C:.

The self-extractor automatically ran the installation program for the drivers and "HP PrecisonScan Pro" software (a program that exhibits few if any "professional" attributes). This was a poorly-configured Wise installer. By default, the software installs to the SCANJET folder on the main system drive. Bad call. Since Win95 — that's more than 10 years ago, people — Microsoft has prescribed a standard filesystem layout in which applications should be installed in the "Program Files" tree. Why does HP think it's special and its program files should be installed on the system root? What I did, but shouldn't have had to do: Browsed to an alternate installation location in my Program Files folder.

Finally, the software installed.

Next, I connected the USB cable from the scanner to the computer and plugged in the scanner, which has no "on/off" switch of its own. Windows detected the new USB device, located the driver, and started to install it. After a few moments, I saw an error message: The driver installer needed the program hpsjrreg.exe. This program was not included anywhere in the downloaded ZIP from HP. I found it on a third-party repository of printer drivers. It made me a little uncomfortable having to get an EXE file from a third party.

The installer completed. I started the "HP PrecisionScan Pro" software and tested the scanner, and it worked.

Operation and Failure

It worked for about 5 minutes. After five minutes, I had a paper jam in the Automatic Document Feeder (ADF). Following instructions in the ADF manual, I extricated the paper and prepared to try scanning a different document in the ADF.

Problem: The "HP PrecisionScan Pro" software could no longer communicate with the scanner. I got the following error message: "Sorry, scanner could not be initialized. (Scanner not found.)"

Troubleshooting and a slog through a badly designed and poorly written technical support information database

I checked the USB connection by looking in the Device Manager. It was good. As far as Windows XP was concerned, the scanner was there.

I unplugged the scanner, let it rest for a few minutes to discharge capacitors, and plugged it back in. This did not alleviate the problem.

I checked the documentation for the scanner and ADF. There's no "troubleshooting" section.

I went searching on HP's support website. Here I found a difficult-to-navigate website for this particular model of scanner. The search function proved useless, so I browsed around and found a page offering help with error messages.

As you may have guessed from the error message I quoted earlier, the "HP PrecisionScan Pro" software does a lousy job explaining what's going on when it presents an error message. I've seen a few variants on the "could not initialize scanner" error message, and none of them have given me a clue about where things have gone wrong. Moreover, one can't tell the difference between them. They're the kind of error message that could only help the programmer, and those kinds of error messages infuriate end-users.

The website offered no fewer than six pages related to initialization:

1. "General Troubleshooting for Initialization Errors and/or the HP Scanjet Scanner Stops Working When Using SCSI or Parallel." Well, I'm using USB, so this one doesn't apply.
2. "'Cannot Initialize Scanner' Message Appears after the First Page Feeds Through the Automatic Document Feeder." You can't tell from the description, but this one also applies only to SCSI connections.
3. "An Error Message of 'Scanner Not Found' or 'Sorry Scanner Could Not be Initialized'." That looked promising, too, but this one only applies when you're trying to use a network connection to access a scanner that's attached to another computer.
4. "'Sorry, Scanner Not Found; Scanner Not Initialized'." Hey! This one looks almost exactly like the last article title, but it has different advice. This article suggests manually adding some keys to the Windows registry. I tried it. Nothing happened. Rebooting didn't help.
5. "Resolving 'Scanner Could Not Be Initialized, Scanner Read Error' Message." Amazing how all these documents with similar titles can have different advice. Like the first few, this article only applies when the scanner connects to the computer using SCSI.
6. "Resolving USB Scanner Initialization and Related Communication Error Messages." Aha! I connected the scanner using USB! Maybe this is a real solution!

I looked at the last article. What I found there was an absolutely atrocious un-install everything and start over procedure that required the user to edit the registry manually.

This kind of recommendation is not a real solution. It's giving up and starting over. It shows that somewhere in the software there is a bug so severe that giving up and starting over is the best hope. It's an admission: "Our program sucks." In the course of the years since the program was released (in 2002), no one has figured out what's wrong and tried to fix it. Instead, HP asks the user to take twenty to forty minutes, engage in risky computer management activities, and start over from scratch.

This should shock anyone selling or buying hardware and software. You ought to be appalled, and HP ought to be ashamed. It should never, ever get that bad. Especially not because of a paper jam.

In which I try to take HP's advice

I picked my jaw up off the floor and thought I'd give HP's advice a try.

1. Uninstall the software. The uninstaller didn't work. After I clicked on "OK" in the "are you really sure you want to remove this software" dialog box, the box disappeared and nothing further happened. I manually deleted the files from my Program Files folder.
2. Delete other files the uninstaller wouldn't have deleted anyway. Done.
3. Uninstall the scanner drivers in the Device Manager. Done.
4. Edit the registry. Done very carefully.
5. Reboot and repeat the installation process. Done.

After all of that, lo and behold, the scanner worked!

Operation and Failure (again)

The "HP PrecisionScan Pro" software works with the ADF, but it doesn't save Class F multipage TIFFs. It saves each page as a separate image. Useless. Its dialog boxes are obscure and its settings obscurer.

This morning, I was using the computer when the power failed. I wasn't operating the scanner, but when later I tried to scan a letter I'd received from a friend, I got an error message:

"Sorry, scanner could not be initialized. (Scanner not found.)"

Unbelievable. Truly unbelievable.

Since then, I've tried the uninstall-reinstall process again and it didn't work. I may try it once more later today.

Things to Appreciate and Lessons Learnt

What do I take away from all this?

I'm glad the previous owner took good care of her equipment. If she ever got this error, though, it'd be a wonder if she ever got to use the scanner. I hope the software behaved better for her.

I will think twice before buying again from the consignment shop the seller used. There's no excuse for a five-day shipping lead time or for failing to respond to polite e-mail inquiries.

I probably will never again buy a Hewlett-Packard product.

I will take solace in the fact that the Macintosh software for the scanner is produced by someone other than HP. Their website could be better organized, but when I get a PowerBook I'll give that software a test drive. It'd be hard for it to be any worse.

In the meantime, I've read that the open-source SANE Project (Scanner Access Now Easy) supports the Scanjet 6200 on USB. I'll boot into Linux, load it up and try out some SANE-enabled applications to see if it works over there. It would be a pain to have to use Linux for all of my scanning, but if it works that'll probably be the lowest stress solution until I get the Mac.

I still think I got a really good deal on a scanner. I just have to get it working, and I'm glad I still have a few options through which I can try to achieve that end. But I'm truly appalled at HP's bad software and disorganized web-based support.

When I rebuilt my desktop computer about a year ago, I bought a new power supply to feed the upgraded central processor. I avoided spending much money on the power supply, though. After about a year, I think the fan bearings aren't quite as smooth as they used to be, or something else there has become less-than-perfect. Increasingly often, I hear a faint sound:

wrrn wrrn wrrn wrrn wrrn

A vibrating sound, about 150 beats per minute, like the fan is just slightly out of balance.

Except for a few days a week, I'm in front of this computer or in the same room for most of the day. Wrrn wrrn wrrn wrrn is starting to drive me nuts.

I'm hoping that soon I'll be using this computer a lot less, partly because I won't be here during the day and partly because I'll be able to use a new notebook computer when I am here. And then eventually I'll replace the case and power supply on this computer.

Until one of those things happens...

wrrn wrrn wrrn wrrn wrrn wrrn wrrn wrrn....

That msnbot sure is a drain on bandwidth.

If you're concerned about the privacy implications of information technology and you don't already follow Sabrina Pacifici's weblog beSpacific, this is a great time to start.

Today, beSpacific includes a post with links to resources on how to find out what ChoicePoint knows about you. ChoicePoint is a conglomerate of database services that collects all the data about individuals that it can get its hands on. Then they sell it.

Not exactly comforting. I think I'll have to submit a request to see what they've got on me.

Novell announced yesterday that it's releasing "core components" of its NetMail "collaboration server product" under a combination of two open-source licenses, the GNU Lesser General Public License (LGPL) and the Mozilla Public License (MPL). The spun-off open-source project is named the "Hula Project." Opinions are beginning to circulate about whether "groupware" makes for a good open-source project. Some of the comments out there made me wonder why we don't have better business software.

Jamie Zawinski, former Netscape programmer turned nightclub owner, suspects that the project will founder, and expounded on his belief in characteristic fashion on his LiveJournal. Clay Shirky quoted some key passages over at Many-to-Many, and I'll quote some of them below, too.

Zawinski says Hula won't work out as an open-source project because it won't make its users happy, and independent coders won't be motivated to jump in and be part of the project. Possibly true.

If you want to do something that's going to change the world, build software that people want to use instead of software that managers want to buy.

When words like "groupware" and "enterprise" start getting tossed around, you're doing the latter. You start adding features to satisfy line-items on some checklist that was constructed by interminable committee meetings among bureaucrats, and you're coding toward an externally-dictated product specification that maybe some company will want to buy a hundred "seats" of, but that nobody will ever love. With that kind of motivation, nobody will ever find it sexy. It won't make anyone happy.

Zawinsky's concern is that the Hula project will not attract open-source coders because it lacks an intrinsic sexiness or coolness factor. "With a product like that, there was going to be no teenager in his basement hacking on it just because it was cool, or because it doing so made his life easier. Maybe IBM would throw some bucks at a developer or two to help out with it, because it might be cheaper to pay someone to write software than to just buy it off the shelf."

Beyond that, though, I want to ask why we settle for "enterprise" software that's hard to use and bogs its users down. Jamie Zawinsky's rant promotes software that "makes people happy" by helping them do things that they already want to do. I would extend that principle to say that using business software should be easy. Business software should also be engineered to make the user as "happy" as possible from a usability perspective. I have had few moments more frustrating than when I've been working on a project and poorly-designed computer software only got in my way.

If I had employees, I'd want to have tools that they would like using. I'd propose that enterprise business software should, at a minimum:

• Help the user get something done — communicate, schedule an appointment, solve a problem, think clearly.
• In the process, also make the user feel like she or he is getting something done. (Hint: this may be best accomplished by reducing, rather than increasing, the number of interface cues. Successfully completing extra steps arbitrarily imposed by software or interface design does not make me feel like I've accomplished anything other than fighting software.)
• Stay out of the user's way — the program's design and human interface should be carefully engineered to facilitate the purpose of the program, but never to interfere with it.
• Prevent information overload. Information overload comes in many forms, of which an overcrowded interface is only one.

The more I think about it, the more those propositions all seem like corollaries of each other. This is my first foray into thinking about business software that way, so I'm sure I've left something out in this over-general, amateur software design rant. Perhaps my friend studying for his Ph.D. with a concentration in human-computer interaction will have something to add. But I'll leave it at that for now.

I've read a lot of pitches for enterprise software that's supposed to improve employee efficiency, trim the "bottom line," and improve communication and planning within an organization.

That's a fine place to start, but I really want software that can deliver on a pitch that goes something like, "Not only will this software make your employees more effective, but they'll like using it. You won't have to cajole them or force them to take training courses because they'll want to use the software anyway, and it's so well-designed that they won't really need (much? any?) training. They won't have to waste time figuring it out. Give this to your employees, and they'll leave at the end of the day feeling like they got more done. That's because they did. And they'll come to work a little happier the next day."

Who could turn that down? Is there anything out there that can make that kind of sales pitch, let alone deliver on it? Shouldn't that be the business software designer's goal just as much as for any other software writer?

Zawinski says, "If you want to do something that's going to change the world, build software that people want to use instead of software that managers want to buy." I say, managers ought to want to buy software that people want to use at work; I'm just not sure who makes it.

See also: Tim Bray on collaboration tools.

The Background

I last acquired a new notebook computer in 2000. That one was a replacement for a glitchy one I'd bought in 1999. They were both Intel-based machines that arrived from the store running Windows 98SE. I chose those computers because I had built my own Windows and Linux desktop computer, I wanted to be able to dual-boot Windows and Linux on the notebook, and I knew that the software I would want for law school either ran only on Windows or had Mac versions that lagged the capabilities of the Windows releases.

I stopped using my notebook computer during my third year of law school because it was too heavy to lug around. I had several pairs of jeans with a small hole where my computer bag brushed against my leg and wore through the denim. I also wanted to write my notes more selectively instead of transcribing classroom events.

In the meantime, I've continued to upgrade my workstation at home, but the notebook is sitting unloved because its piddly 64MB of RAM can't handle anything that's currently out there. Once I load Windows (or Linux and a desktop environment), running just one worthwhile application starts the hard drive a-grinding as the computer constantly relies on its swap space. And the battery is dead.

So as soon as I've re-established positive cash flow, I need a new notebook computer. But this time I plan to make a different choice. I plan to buy an Apple PowerBook.

Why?

Several reasons. I could probably analyze it out into seven to ten reasons. But instead, let me tell a story.

I like playing with different operating systems and interfaces. I've been a computer hobbyist since I was about seven years old. My experimentation has taken me from Apple ProDOS (on the //c) to MS-DOS to OS/2 to "old" MacOS to OpenVMS to Windows XP and Linux — and those are only some of the steps along the way. Since 1999, though, I've used Windows for all my "real work," because that's what law school and my firm required.

For years, my hobbyist background meant that user-interface quirks and technical glitches didn't bother me as much as they bother most people. I usually had enough time that I could treat a computer problem as an interesting puzzle to be solved rather than a brick wall across my path.

Law school, and law practice especially, have robbed me of that point of view. When I have free time, I don't want to spend it solving computer problems — unless, of course, I've chosen to spend hobbyist time on them. Several times over the course of the last two years, I've found myself hacking away at a computer problem instead of finishing my "real work" while under deadlines.

I have several friends who have used Windows, Linux, and OS X. They agree that OS X has given them the greatest amount of flexibility with the least amount of frustration. One of them said, "I didn't use Linux because I wanted a challenge or because I necessarily needed it for something; I used it because I wanted a computer that would do what I wanted it to, and I was willing to invest some effort in that. The Mac does what I want it to, and I don't have to put any effort into it. It just works."

Of course, I'll grant that a computer won't just work all the time, even in the Mac world. But it just works more of the time, and it works more intuitively. That translates to increased efficiency.

Interface design in the Apple world also plays a big role in my decision. Apple, and people who design software to run on Apple equipment, do really impressive things with interface design. I want to take advantage of that. I've spent too much time in the last year cursing programs for getting in my way.

Even better yet, the Mac will still let me experiment if I want to. Part of the guts of OS X is based on a BSD-flavored UNIX called Darwin. I'm already familiar with a few flavors of Unix, so once I figure out some of Darwin's unique quirks, I'll be in a good position to play around with it a bit.

There's still one area where software tends to be a little sparse on the Mac in a way that affects me. It remains true that a lot of specialized legal software runs on the PC only. My next firm will probably require me to use Windows XP. They'll probably keep time with Timeslips or TABS3, or maybe even Amicus Attorney. CaseSoft doesn't run on the Mac, and I haven't yet found any reports of whether it works under Virtual PC. So I'll have two computers at work. But that's worked well for an increasing number of people (Ernest Svenson is my model in this regard). I'll need to cough up four to five hundred dollars for Microsoft Office. But then I'll be set.

I'm heartened to see that Apple is again targeting the legal market, particularly small and mid-sized firms where the costs of transition are lowest and the necessary software is most robust. If I were to start out on my own, I'd probably do it like these guys. It's good to know that law practice management software is out there for the Mac.

I replaced my old Macintosh Performa with a PC in 1997 because I wanted to experiment, because I wanted to run Linux. At the time Apple looked like it was in a slump. Now I want to spend my time getting things done, making things happen. So I'm going to buy a Mac.

(Now that I've gone and said all these things, I guess I'm really going to have to go through with it.)

Adina Levin has this interesting post raising questions about electronic social networks and notions of public identity.

Many2Many

Christopher Allen's post that started it.

Ann Althouse has this interesting this post about the article "Our Godless Constitution," in The Nation. She shows how the article incorporates shoddy scholarship and, more importantly, explains how the article's inflammatory approach will utterly fail to explain the importance of separation of church and state to religious people. There are many arguments, including those adopted by the founders of the United States, that would demonstrate the social and religious importance of the separation of church and state.

Andrew Raff's IPTAblog has an article following up on the controversy that blew up last month when Martin Schwimmer asked Bloglines to remove his Trademark Blog from the Bloglines service. He frames and summarizes the issues well, I think.

My jaw dropped a bit, though, at the Sieblogs site that Andrew linked to. Sieblogs aggregates content from what must be a collection of hundreds of weblogs and mass media sources. It then displays that content by category. There's nothing on the main page or entry page that indicates the original content author. That's a violation all of the Creative Commons attribution licenses, and the sheer copying and re-display of the content is a violation of general copyright law, although it is something many weblog authors would permit if asked. Even though there are no advertisements yet, the utter failure to credit the author and source weblog really surprised me. They at least link to the original article.

Slashdot linked a story by the BBC reporting that the U.S. hosting service The Planet terminated the contract of the Iranian Student News Agency (ISNA).

According to the article, ISNA is "semi-official" and may have some connections to the Iranian government. (If I recall correctly, the fundamentalist revolution in Iran called itself a "student" revolution, which may have some connection to the name of the organization.) Assuming so, the policy concerns, from the perspective of the United States, are different from a termination of a more liberal website.

There may be no effective way for a U.S. ISP to distinguish between the two kinds of websites, though, nor is it at all clear that the U.S. government should systematically engage in content-based discrimination between them. From a practical standpoint, ISP operators can't read Farsi and are unfamilar with the culture and politics of Iran, and can't be expected to make those calls anyway. It's not surprising that some ISPs are taking a bright-line approach and terminating accounts of Iranian entities and Iranians not resident in the United States, even if it silences some voices that the United States would probably like to amplify.

Assume an Iranian national resided legally in the U.S. and used a webpage hosted in the United States to write in favor of the existing Iranian government. Can the ISP terminate the contract because he is an Iranian national? Can the government require the ISP to do so without running afoul of the First Amendment? What if the Iranian national moves back to Iran?

Hossein Derakhshan reports, citing the BBC's Persian edition, that Iranians who have hosting or domain registration in the United States are being refused or cut off by their U.S. hosting providers.

Iran is on a U.S. blacklist that is intended to prevent it from receiving technology that it could bend to military uses. For example, it is particularly illegal to transmit encryption technology to Iran. I'm looking for a copy of the law or regulation now but I haven't found it yet.

If U.S. providers won't host Iranian web pages, reform-minded Iranians lose out, as do U.S. efforts to support them. They can't host their web pages on Iranian providers. I don't know whether it is at all easy for an Iranian to host a web page with, for example, a provider in an European country.

So I doubt that the U.S. ought to apply its restrictive technology policy to web-hosting. But even if the Department of State or other agencies are not putting pressure on hosting providers, those providers may themselves restrict access in order to avoid a conflict with the government. If that's the case, then what is needed — as a first step — is communication from and among the various government agencies that could potentially be involved, making it clear whether they intend to make life difficult for hosting providers who provide access to Iranians. A policy of prosecuting hosting providers would be unfortunate for reform-minded Iranians and contrary to the policies announced in President Bush's inaugural speech.

I have seen some comments conjecturing that this is part of some action leading up to a conflict between the United States and Iran. I suspect it's just the application of a technology blacklist that's been in effect for decades. But like most embargos, it harms the people the U.S. wants to help; and in this case, it harms them in a politically powerful way.

I hope it is just ISPs being overcautious, but even if it is, then someone in the government needs to tell them that they can relax. But governmental communication, especially on points such as these, is sadly too much to hope for.

[Via Dan Gillmor]

Thanks to this post by Jon Armstrong and the "bitpim" software he refers to, I finally have a way to extract photographs from my Sanyo PM-8200 phone without zapping them across SprintPCS's digital data service.

I had already bought a USB data cable for the phone, so all I needed were current USB drivers for the phone and bitpim. I did struggle for a while with the hardware until I realized that the caller ID program on my computer was grabbing access to the phone and locking bitpim out. Once I disabled the caller ID software, bitpim worked fine.

Late last year, the Creative Commons project announced that it had prepared several form content licenses designed to allow people who publish on the internet and in other media to publicly license their work. The Creative Commons organization aims to increase the amount of creativity that the public can share and draw upon in further creation. "Taking inspiration in part from the Free Software Foundation's GNU General Public License (GNU GPL), Creative Commons has developed a Web application that helps people dedicate their creative works to the public domain -- or retain their copyright while licensing them as free for certain uses, on certain conditions." [link] Nearly four months later, the experiment is still only just beginning both for the Creative Commons organization and for the artists who license their work under Creative Commons licenses.

U.S. Rep. Zoe Lofgren has reintroduced her act to curtail the overreaching potential of the Digital Millennium Copyright Act. Rep. Boucher is co-sponsoring. The bill is 108 H.R. 1066 and is being called the "Benefit Authors without Limiting Advancement or Net Consumer Expectations (BALANCE) Act of 2003." It has been referred to the House Committee on the Judiciary.

Mercury News article: here
Peter Suber's comments at Free Online Scholarship News: here
Bill text in PDF: here; status here.

This CNet article by John Borland says that several companies plan to avoid building digital rights technologies into chips in hopes of attaining the necessary functionality through software.

Despite a move away from building the rights-management tools deeply into chips, chipmakers' strategies remain widely varied. Given the long lead time in designing and building chips--often 18 months or more--this is one sign that DRM support is likely to be scattered and haphazard for some time to come.

Giant Texas Instruments has long eschewed hard-coding DRM technology into its chips, for example, despite the potential speed and memory gains.

"Our philosophy has always been that DRM should be software," said Randy Cole, chief technologist for Texas Instruments' Internet audio business. "The advantage to that is that it's changeable in the field."

What that means is that if a consumer is able to break through the antipiracy technology on a device such as an MP3 player, it can be restored automatically the next time the device is connected to the Net, Cole said.

This doesn't mean "no DRM," but it may mean "more flexible DRM," and allows for the possibility that companies could revise bad DRM technological standards without obsoleting recent equipment. However, the article notes, major players like Intel and Microsoft still plan to pursue hard-coded DRM implementations.

David Weinberger says that Larry Lessig told the SXSW conference that the Creative Commons project will release new kinds of licenses tomorrow. I'm looking forward to seeing them.

Speaking of which, how about that -- a mention over at the Creative Commons! Thanks! I guess that means I'd better finish up the next revision of my earlier posts on the topic.

Declan McCullagh reports in a CNET article that U.S. Rep. Zoe Lofgren (D-Cal.) has or soon will re-introduce legislation that would limit the DMCA's anticircumvention provisions (12 U.S.C. § 1201) so that they would not apply to circumvention for purposes that constitute fair use or noninfringing use under ordinary copyright law. Her legislation would expressly permit "in order to perform or display the work, or an adaptation of the work, on a digital media device, if such performance or display is not public." The legislation would also expressly apply the first sale doctrine to digital media, though it's not clear that doing so would negate restrictive provisions in digital media licenses. Link to the draft she submitted in the last Congress here.

Meanwhile, Dan Gillmor's column in the Mercury News reports that U.S. Sen. Ron Wyden (D-Ore.) wants DRM-secured and other copy-protected media to bear prominent labels to that effect.

Last year, I began using Microsoft Word 2000 because the journal and the law clinic both used it. I recently installed WordPerfect because I remembered that it would give me more control over the little details (hooray for Reveal Codes!), and the version of WordPerfect 9 I had was able to generate PDF output.

It's been a bit of a bumpy road. I'm using WordPerfect 9 (2000) with Service Pack 3 running on Windows XP. My printer is a Samsung ML-1450 laser printer. The first quirk I encountered was that WordPerfect didn't seem to think that I had any printers installed. It pops up a dialog saying "Cannot print without printer installed." Windows reveals that there are three printer drivers installed, and my printer is selected as default. It did this even when I told it to print to PDF.

Then I discovered that I could get it to print if I first put it in Print Preview mode. I could also get it to export to PDF by doing that first. However, the resulting PDFs looked awful. The spacing between letters varied dramatically for no apparent reason, even sometimes appearing to split words in two. I did not find that at all att ractive.

Then, thanks to a recent entry on Ernest Svenson's PDF for Lawyers blog, I remembered that Ghostscript exists. Ghostscript is a general, multipurpose tool for working with PostScript files. I'd used Ghostscript on Linux frequently, but I'd never used it on Windows. I went to Samsung's webpage to download a PostScript driver for my printer. (That was a nightmare because Samsung's driver downloads page doesn't have a listing for my printer even though they carry drivers for it. Google found it.) While my printer won't actually print PostScript, the driver generates a .ps file according to my printer's metrics. I set it up to print to a file, then used the Ghostscript GSview program's "Convert" function to convert the PostScript file to a good-looking, good-printing PDF. What's more, I should be able to do that with the output from any Windows program. I'll have to test how it handles different fonts.

In any event, I at least have a good PDF copy of my resume, which is what I was after the whole time.

Someday, I'll buy Acrobat. But for what I need to do now, this arrangement will get the job done.

Creative Commons has added comics that aim to explain in simple terms the options that the Creative Commons licenses present for creators.

It's good that they try to spur people's imaginations a bit more about how the various Creative Commons license rights and reservations might work for them. This is very important to anyone who might consider using a Creative Commons license. On the other hand, I've still found nothing on the Creative Commons website that thoroughly addresses the question "what if I change my mind?"

As predicted, today's Federal Communications Commission meeting resulted in newly relaxed versions of the regulations that require incumbent telephone providers to lease their network capabilities to competing companies. Ben Charny has a CNET article on the matter.

The Bells will still have to open up their phone lines to local competitors--at the steeply discounted rates the regulations require. But the rules will be off-limits to any new broadband networks the Bells plan to build, the FCC decided.

Also, the states will get more clout in deciding whether the Bells should be required to open up their networks, and in what locations. The FCC had been the only agency with the authority to handle requests to offer service in new markets. Under the new rules, the state’s Public Utility Commissions will do most of the deciding instead. States had been clamoring in the past few weeks for just such a decision, and insiders believe their push for power helped topple what seemed like a Powell victory.

Now everyone involved has to take the rest of the day to read the FCC's report and figure out the details of the regulations.

FCC website with press releases and Commissioner statements: www.fcc.gov.

CNET reports that Palm Computing and AT&T Wireless will soon have the Tungsten W on the shelves. The Tungsten W resembles the Tungsten T, but it doubles as a cellphone.

Well, it doubles as most of a cellphone. It seems to lack a built-in speaker and microphone. Instead, the user has to plug a hands-free earpiece in. Maybe they thought that would add to the device's appeal because you can talk while looking up or editing data on the handheld device. However, that deprives the user of the all-important ability to just grab the device and talk. I'm glad they decided to include a thumb-keyboard, though.

The Tungsten W also lacks the faster PDA technology of the Tungsten T. Instead of the T's ARM processor and Palm OS 5, the W runs Palm OS 4.1.1 with a 33 MHz Motorola Dragonball VZ processor.

I won't buy the product for three reasons: I don't have the money, it doesn't run on CDMA1x networks (I'm locked in as a SprintPCS customer), and I really need something I can hold to my ear and talk into. Having an earpiece jack is a good idea, but the earpiece should be an optional accessory.

Microsoft is introducing a peer-to-peer technology called Threedegrees to allow people to chat in small groups while "listen[ing] to music available in a common play list."

For NetGenners, Microsoft learned that using the Internet for socializing is a way of life. So the company focused on technologies that would help 'get groups formed and have activities they can do,' Savage said. 'We wanted things that paralleled our customers' priorities, which was hanging out with your friends and having fun....' ...

Microsoft used the dinner party as the model for developing the size of the social group and the way music is shared within it....

Group members can create play lists of 60 songs, or about the equivalent of six CDs. The songs are played from the participant's hard drive, rather than being illegally swapped. Songs can be in Windows Media Audio, MP3 or WAV formats.

[CNET story via The Shifted Librarian]

Well, that removes the file copying problem, at least as long as no one makes a way to dump the music data stream into a file as it's played. If users can share the music files simply by, say, changing the file extension so that the system thinks it's a different kind of file, that might be more of a problem.

Here are two questions: does playing the music over the chat system constitute a performance by digital transmission of the music, which happens to be an exclusive right under U.S. Copyright law? Is it a "performance" limited by 17 U.S.C. § 106(4)?

I doubt that using the proposed system would infringe those rights. This activity is just like having friends over to my house and playing some music together. We don't need ASCAP, SESAC, and BMI licenses to do that. The digital audio transmission right in 17 U.S.C. § 106(6) and the general performance right in § 106(4) both include only the right "to perform the copyrighted work publicly" (Emphasis added). That's why I don't infringe the § 106(4) performance right by playing music for some friends at home or in the car. The proposed system isn't like an audio stream that's open to all comers. Microsoft restricts the peer-to-peer chat group size to ten people, which helps limit the possibility that the audio transmission could be considered "public." It also doesn't infringe the right to copy, because nobody walks away with recordings they didn't have before the gathering started.

Legal matters aside, record companies will probably realize that Microsoft is offering something that will facilitate word-of-mouth marketing for their recordings without allowing the proliferation of copies that the record companies so deeply loathe. Like people gathering to play recorded music together, this encourages interest in music. This sort of activity is the only way that people like me who don't listen to the radio discover new music that we like. The next question is whether Microsoft's offering will catch on.

Update: Newsweek online at MSNBC has a longer story on the new system. Cory Doctorow (BoingBoing blog) has a link to the Newsweek story along with a few normative comments on the implementation.