Comments from Frankston, Reed, and Friends
Thursday, February 28, 2002
BobF at 3:54 PM [url]:
OK, one more peeve: Submarining!
I have a whole raft (no, this is not a marine pun) of other complaints about silly UI's that are unnecessarily clunky.
But there is one that I find particularly bad in Windows systems -- the submarined window. This is my own term for modal dialog boxes that are hidden well behind other windows. A modal dialog is a window that you must respond to before you can continue. In fact, the application often refuses to respond to any events until you've satisfied the demands of the dialog box.
This would be simple enough except that this dialog can be buried under other windows. For example, if you are sending a letter in outlook and open another application you might not see it asking about a spelling error. The only symptom is that Outlook and, perhaps other parts of Microsoft Office seem to be stuck.
Yesterday I was puzzled about why I couldn't view PDF files in the browser. I eventually noticed that Acrobat had posted a dialog box asking if I wanted to check for a new version and was waiting for a response. Even if you know there is such a dialog box, actually bringing it to the front challenges even trained professionals with decades of experience.
This is a very old problem with Windows and there are some obvious fixes such as forcing such windows to the front whenever you touch any dependent window. But the biggest problem for me has been getting people to even know there is a problem and to care enough to fix it.
I promised just one peeve so I�ll write about the tyranny of the OOBE in a separate note. What is the OOBE? Well, sorry, I�m only taking on a single task in this note though you might try and look for the OOBE directory if you have an XP system.
But for now, I'll just worry about finding Waldo
DanB at 3:23 PM [url]:
More forms complaints and the effect on transcription errors
I second Bob's comments below. My pet peeve is forms that ask for credit card expiration dates and give the months as alphabetic names, not numbers. All the cards I have list the date as MM/YY or MM-YY. When you are copying from your card, you see "03" or something, not "March", and want to visually match it with the "03" in a drop-down. Why make it so hard? Are you showing off that you can do an alphabetic list? Are you thinking you are being nice to convert it to a name, when mostly what it really is for this purpose is another identifying key? What is the error rate for people converting numbers to months, especially months that aren't their birthdays or "this month"?
With respect to cadence, I remember that when I used to take credit cards over the phone at Software Garden years ago, when I read back a credit card number I always tried to read it with a different cadence than the person gave it to me. Usually I ended up reading it grouped two digits at a time. As Bob points out, that always throws off the listener, and sometimes required reading it twice. I found, though, that if I didn't do that, reading it back in the same cadence was more influenced by repeating what I thought I heard and not by what I wrote down on the charge slip. Even if I transposed some numbers when writing the order down, I'd repeat correctly what I heard from memory. The listener would probably also be checking partially with auditory memory of what they said. Changing the cadence forced me to read what I wrote and forced the listener to listen carefully. It resulted in fewer errors.
BobF at 2:23 PM [url]:
Dash it all or Cadence Counts!
I just tried to type in my zip code and then realized that I had typed a "-" and the form assumed I wouldn't do such a thing. A US zip code is officially of the form 00000-0000. And my American Express card is of the form 3727-123457-182112. Yes, I know, those hyphens are ignored by the computer. But if I have to type a long string of digits and you want me to type them correctly, then you want me to type those hyphens in order to make sure I get it right. Humans need to chunk information in order to see or understand it.
Cadence matters. In 4/4 time (in music) the strong emphasis is on the first beat, and the secondary is on the third. American phone number are 000-000-0000, Europeans often use 00-00 00-00 etc. If you use the wrong cadence people have trouble hearing the number.
This is all very obvious and very well known.
So why do most web forms insist that I don't type a hypen or other punctuation?
Much of the blame lies with the ungeek. The ungeeks are sufficiently intimidated by technology that they take "you can't" for an answer. It is nothing more than illiteracy! And, more to the point, it is offensive. Offensive to the user who is annoyed with such nonsense. It is offensive to companies that have to deal with these errors. And it is offensive to those of us who are real geeks and understand how to make technology work for people and find our reputation sullied by those who blame computers for their own ignorance and laziness.
Punctuation is important! If you don't understand .replace(/-/g,"") you shouldn't be allowed to use powerful tools nor interface with humans.
Wednesday, February 27, 2002
BobF at 11:15 PM [url]:
Jack Valenti replies to Lessig, other "professors"
This is a response to a recent Washington Post OpEd written by Jack Valenti, Chairman and CEO of the Motion Picture Association "What's keeping the movie industry from making its creativity theft-proof?"
I'm fascinated by the concept of "theft-proof creativity". Need I say more? (perhaps not but I can't resist) I presume that any effective mechanism will also prevent remakes of old movies.
If I were to write a full response I'd point out that the whole movie industry is premised on a particular idiosyncratic model that supports very high costs per production and then attempts to recover the costs over a very wide audience by tightly controlling distribution and limiting choice.
I'm much more excited about other models. As we become better able to generate movies on personal computers (I find 405, The Movie and Animation Master interesting harbingers) it will become increasingly hard to justify holding creativity hostage in return for a complete lack of choice.
If $200,000,000 movies are so important, then Valenti should ask Congress for a direct payment. It's certainly better than creating the necessary mechanisms to police the use of bits.
Just like with e-books, the old ways of doing business don't necessarily translate well into the new arena and it may be a while before I can purchase the latest movies online. Small price to pay for freedom.
BobF at 10:41 AM [url]:
I have a new essay on the future of ICANN.
While I believe that the ICANN members are trying very hard to serve the public good, perhaps they are trying too hard. Unfortunately most of those concerned with ICANN want it to expand on the current policy of providing commercially meaningful names.
The biggest threat to the Internet is the confusion between technology and social policy. It is unfortunate that so many people confuse DNS with a trademark or keyword system. Those are services that can only be resolved by the marketplace. There is no technical solution to the problems of ambiguity and human perception. The DNS must be nothing more than a mechanism for implementing these policies.
Asking that the DNS be a better implementation of these social policies is like trying to make the Internet operate just like the phone network when the power of the Internet comes from giving us the ability to use connectivity in ways that the phone network does not allow because it is overdesigned for one purpose.
The purpose of the DNS is very simple. It is supposed to be a source of unique and stable handles. The Web depends on the stability of these handles in order for the linkages (URLs) to stay valid.
As long as the DNS is forced to act as very bad and naive implementation of trademark and keywords all the handles must expire and worse, they can be repurposed without notice as many schools have discovered when students click on links that have gone from innocuous to erotic. Far worse, however, is that the names come with commercial terms and conditions that force those innovating on the Internet to ask permission in order to be allowed to innovate. This assures that the Internet is not annoying or vibrant. It also makes the assumption that the Internet is about e-commerce when, in fact, the Web (just an application itself) was the product of individuals experimenting and not corporate initiative.
While we can't rewrite history, we can provide a path out of the quagmire by asking ICANN to do nothing more than fulfill its simple mission of providing globally unique identifiers. We should create the last TLD (.DNS) and use it as a source of globally unique handles that are never reused and are not fettered by the terms and conditions and not beholden to commercial interpretation (perhaps just use long integers). Very simple and a start towards assuring that the Internet is a source of innovation even if it does threaten those who fear change.
Whether or not one agrees that the current .COM mania makes sense or not, the opportunity to opt out and choose stable identifiers is very important. Note, however, that the release of the seven new gTLDs has been met with utter and profound indifferences. If you type name.COM and it doesn't work, you go to a search engine and directory rather than trying all the possible TLDs and all the variations in spelling. People know this and don't really depend on these commerical names (nor on JohnSmith.NOM). Adding more TLDs just adds more confusion which emphasizes the point that the whole ".COM" thing is not very important except to those who believe there is the one true phone book that has everyones one true phone number and one true name.
ICANN's failure is not due to a conspiracy but simply due to its attempt to do the impossible and replace complex and intrinsically ambiguous mechanisms with hard and brittle technology. The technology can provide the means of implementing policy. And, as the Internet has demonstrated, these means should support the implementation of many policies, old and new, rather than being highly tuned for the existing (old) policies. In the case of the DNS this is especially important given that it can't even implement the current policies and their rich complexity. But trying to do so does real harm to the ability to create new value.
For more on this you can read the essay.
Sunday, February 24, 2002
BobF at 11:30 PM [url]:
Worlds in Collision: Consumer Electronics meets generic computing and networking and ...
Note that though I mention real companies and products, I'm not endorsing them. They are just examples.
It's interesting to contrast two products. NECXDirect is selling ViewSonic's combination of a 15" LCD monitor and a TV tuner. The tuner itself is available for about $100 and can display on a standard computer monitor (though I presume one would use an LCD). The bundle is $450. Or you can go to Best Buy and get the Sharp 15" Aquos Flat-Panel LCD Stereo TV for $1300. I presume the Sharp has better styling but there isn't necessarily a quality advantage.
But for better quality, you can buy a PC for a few hundred dollars with a TV tuner and benefit from all the investment in support for video games. You can even outfit the PC with neon accents and other effects if the right (or wrong) look is important.
In January I was at the Consumer Electronics Show (CES) in Las Vegas and was surprised that computer people were so accepting of consumer electronics products, such as TiVo. I just felt frustrated at the inflexibility of the products and the difficulty in making them do anything more than they did the day you bought them.
This feeling was reinforced when I visited You-Do-It Electronics last week. The store is vaguely like Radio Shack with a second floor selling an interesting selection of consumer electronics products and the downstairs selling electronic components. It's different from Radio Shack in that it sells to professional installers and the consumer electronics section has more high end gear.
But what struck me was the plethora of connectors and wires and special little doodads that compromised the world of analog electronics. Every wire had a single purpose and had to be set up just right with seas of connectors on some of the devices. And with analog signaling the quality of the wire has to be just right otherwise the signal degrades. Just to be safe you hire experts and pay for premium gold plated (literally) wires. Companies like Monster Cable cater to this market and older "speaker" companies like AR are getting into the business. And HDTV, which used to look so futuristic, is well below the level of a standard computer monitor (and Congress still is planning to force us to adopt it in 2006 when it will seem very quaint).
We also have some of this complexity in computer systems with USB (1.0 and 2.0), SCSI, FireWire (1394), Video (VGA and DVI), parallel, serial and phone wire. The confusion here is similar to that of the analog world. Monster Cable is willing to sell you a $60 printer cable just to be safe. I don't begrudge them taking advantage of the opportunity as long as I can easily choose to buy a $3 cable instead.
But, as we see with the ViewSonic products, we are at the start of a rapid conversion to digital. Gigabit Ethernet is about to become affordable and available just like 100 megabit Ethernet did and before that 10 megabits. When I tried to buy an 8 port 10/100 switch LinkSys just threw in some PCI Ethernet boards for free. For those not familiar with the technology, a "switch" is much more powerful than the old hubs and there is no longer a price difference.
This means that I can just connect everything in my house (and beyond) with commodity connectivity and redefine the behavior simply by changing some software settings and I can buy new features and capabilities and "download" them. The components I buy become more valuable over time (or less expensive). I can control my lights or my TV from anywhere and easily add a rule to dim the lights when I turn the TV on. OK, not easily yet, but far easier than hooking up a special device for that one purpose.
We've already seen this happen when the telecommunications industry tried to reenter our homes (after being pushed back to the "demarc" (demarcation point) in the 1970's) with the residential gateway. This was to be the platform on which they would lease out "space" for running services in our homes. The simple router box that connects our home network to the rest of the Internet put an end to that dream (as I had intended). And it will soon put an end to the set top box. Instead of displacing the PC with a "dumb browser" the set top box lives on as little more than a futile last stand against the video streams being Napstered.
It is interesting to see how the various industries try to cope with these changes. Just as with telecommunications we'll see many attempts to preserve the old familiar way things are "supposed to be". But the new opportunities are so compelling and exciting that such attempts to hold back change are futile, dysfunctional and expensive.
DPR at 11:03 AM [url]:
Blogs vs. discussion groups
Dave Winer wrote today in his Morning Coffee Notes a really excellent capsule summary of why weblogs work better than network discussion groups, from a social perspective:
it's clear that some of the people there don't understand why weblogs are superior to discussion groups. Briefly, DGs are like mail lists. All it takes is one stinker to grind the whole thing to a halt. If everyone has their own weblog, people can flame all they want in their own space, but mostly they just attract other losers. And people are less likely to whine in their own lonely space. It's got their name on it (unless they do it anonymously which is even more boring, probably just a competitor without the guts to say so), and it reflects poorly on them, more than it does on the people they're complaining about.
I think this really hits the nail on the head. Though it would be nice if you could hit somewhere in the middle, letting people add "commentary threads" that can be used to link blog entries into a discussion.