I was intrigued by the fact that they're using a scanning approach to projection, something I've been tinkering with (in my head only) for years now.
I've wondered if it shouldn't be possible (or even easy) to set up a mirror that vibrates on two axes, one horizontal and one vertical, and then to point tri-color LEDs at that mirror and focus their converged beams on the wall.
Basically, I figure a 3x3 array of bright LEDs should be more than bright enough, once focused, to illuminate a small pixel space on the wall. Get three of these, one for each color (now that you can find blue LEDs easily), and you've got a full-color pixel. I'd be reluctant to use lasers, both for obvious eye-danger reasons, and also for the (presumed) expense of blue and yellow beams.
In my imaginings, I figure that the horizontal and vertical synch signals could (if properly shaped -- can anyone point me to a site that might show me waveforms for such signals?) even drive a mirror deflector directly -- just hook one corner of a mirror to a pivot point, and hook each of the opposite two edges to a small speaker, and connect the speaker coil to the sync signals (or to re-shaped signals, if necessary).
I'm sure I'm missing something here, but what if I'm not? Could it be possible to build a home-brew HDTV projector for under $300 in parts? Anyone out there care to investigate this further?
One problem that I've always wondered about is how to reliably "sign" a credit card over the phone. The best I've been able to come up with is to have a small "computer," basically a credit card-sized calculator, like a SecurID token. Then, when processing a transaction, you enter pertinent unique information about that transaction into the computer, punch in your pin (which, of course, is not your birthday), and then write the result down on your document (or give it over the phone).
For large documents, obviously, you're not performing a signature operation on all the text. But, maybe at the bottom of the page, you put your unique public key ID (which is then used to find your public key in a big database), then another line with, say, date, cost, and PO #, then the calculated result. If you lose your little card, you simply go down to the post office (or somesuch), get a new one, and they invalidate the old one for any new use after date X.
Obviously, the big problem here is the public key database, but that's been the bugaboo all along. But the advantages of something like this are:
it's fairly easy to use
you don't have to worry about format mangling (spaces, mis-coded characters) changing the message digest
it works, with no "playback" issues (for non-duplicated input data), on paper or over the phone
Disadvantages:
big database
cost of giving people this card
fear of the "national ID" card (which is rapidly becoming moot as all non-national IDs are linked together, anyway, by data-mining techniques)
You could (and, I'd argue, should, with proper back-end privacy features) put proximity technology into this and use it as your gas speedpass, grocery-store bonus card, office key, and gym pass. I like the ideas of the prox-cards (with authentication) for checkouts (like the Mobil Speedpass) but am loathe to put a dozen dongles on my already too-heavy keychain.
Anyway, does anyone like this idea? Can anyone point me to a better way to do secure authentication/validation on paper or over the phone? (yes, I'm ignoring for the moment the possiblity of loss/theft of the card and/or PIN).
The reason why a reboot is required is that the various configuration are made at boot time, based on info extracted from the netinfo database. He probably could skip the reboot by relaunching the correct scripts.
And there's a script ("Build your own location manager" or somesuch) that does exactly this on MacOSRumors.com.
The rollout of broadband data solutions to end users has been an unqualified disaster (example: I live in Fairfax county, Internet capital of the world, but still can't get cable modem or ADSL service). I have long advocated the need to create a quasi-governmental, not-for-profit agency to run fiber to every doorstep, then sell bandwidth (on a cost-recovery basis) to ISPs, Telcos, cable-TV, etc. Barring the issue of cost, would you advocate such a step? Returning to cost, would you endorse "creative solutions" in working with industry to help finance and build such a network?
A related question hinges on technical standards. As part of my suggestion above, I'd like to see a standardized box (that I can buy from any manufacturer, at WalMart) that will convert the fiber signal into TV, Telephone, Data, or whatever. Then, if I want to change my cable provider, I simply call my current provider, say "drop me," call a new provider, say "I want to give you my money," then walk downstairs and it should be switched over. There's no technical reason why this cannot happen today -- except that there are too many competing and incompatible standards.
At many times in the past, the government has refused to endorse a single technological standard (AM Stereo, HDTV (until late), and Cell Phones are examples). The rationale behind this is comendable: "Let the market decide." The end result, however, is that we have several competing and incompatible standards out there, and the industry is unable to settle on any single one -- saddling users with hardware supporting legacy sytems, the inability to switch between providers without buying a new phone, and so forth. How would you rectify this situation (both for new needs, such as the "one connection" box I describe above, and to bring the US into the 21st century to finally catch up with the Finns and the Japanese)?
I noticed that virtually none of the Digital UNIXes are on there. There's Digital Unix, the outgrowth of OSF, but nothing about Ultrix (unless I missed it). Hardly a fringe variant, Ultrix got as high as 4.0 and was on Vaxen, workstations, RISC workstations, all sorts of stuff. Being as how this is the UNIX I learned on, I'm kinda miffed to see it missing.:-(
It's also interesting to note that, as recently as a few years ago, DEC was still using their standard character generator for their PCI-based Alpha machines -- now that's what a UNIX console is supposed to look like!
-- Offtopic --
The quote at the bottom, as I read this:
We are sorry. We cannot complete your call as dialed. Please check the number and dial again or ask your operator for assistance. This is a recording.
Reminds me of the time in school when I found a phone number that was one digit off from mine, but not currently in service. I recorded the intercept message, put it as my outgoing message, and added "but that's not my number, and I'm not here anyway." Best message I ever did, but really pissed off all my friends and family...:-)
I know that several organizations (the Washington Post, for example) have printed photographs (or partial photos) of the DeCSS t-shirts. Has anyone thought about collecting those various photographs together, with proper attribution, and putting them on a web page? I'd bet if you get enough of them any reasonably industrious person could piece together the original code...
What would happen to the trade secret issue then? Collected legitimately from multiple legally published photographs?
What information do you think is fair for Web sites to posess on an individual, and how far do current e-Commerce sites cross that line?
I'd say that any on-line store should require only the same information that is required at a real physical store. That is, if I'm paying for groceries with cash, then when paying with an accredited anonymous cyber-cash-like operation, I should need to provide no information at all. If I'm paying with a credit card at a store, then all the online-store needs is my number and signature. What? They can't get my signature over the wire? Okay, then, I guess they need whatever my credit card agreement says they need -- usually, an address.
In many cases, this is going to be up to the financial companies (banks, credit card companies, etc.) to find alternative ways of validating and authenticating transactions, without divulging address, telephone numbers, etc., to online merchants. Not sure this is ever going to happen. 'course, there's stuff like PayPal, but who knows how long until *they* start doing something with their information.
Of course, we have to remember that in many cases, the business model of online companies may actually include revenue from information collected during the transaction. You see this in bricks-and-mortar stores at, say, supermarkets, with the "special discount cards" that they give people. To be very literal about it, a certain loss of some amount of privacy (some shopping/clicking habits, etc.) are the true price we pay for discounted prices, vast inventories, and free overnight shipping. Don't like the loss of anonymity? Go to your local store. Sucks, yeah, but that's the way the internet works.
In this particular instance, as someone else pointed out, it's likely the problem is that they want to auto-generate their pages to match your browser. Here the problem isn't privacy, but a closed-mindedness as to what browsers are out there. This used to be such a nasty problem when I was surfing from my NeXT that I had to pretend I was Netscape (which was a built-in feature of the browser for just this issue!). They really should have some way of providing a general, simple HTML interface that anonymous browsers can read. Or maybe we need new brower identifications that don't ID the brower, but instead define the browser's capabilities.
Nice timeline (haven't been able to get the emulator going yet), but it incorrectly identifies the first web client as a command-line one, and the first graphical client as Mosaic.
As someone who was lucky enough to try the original browser, probably within days of its release, I find this annoying.:-)
Okay, I've read about a dozen different alternative servers to Napster, and the nicely-logoedNapigator. The one problem I have, however, is that you still need to:
Download a list of servers from a central location. What happens when that server list gets shut down?
Select a single server from that list. How does the user know which is the "best" server out there?
One of the great advantages of Napster (and Gnutella) is that you're accessing a single, common community. With dozens of alternate napster networks cropping up, that community is going to be splintered into many different ones. And people will have to search, possibly, many of the different servers before they find one that's got what they're loking for.
Wouldn't a better approach be some kind of client that combines the best of Napster and Gnutella? Connects to a list of servers, downloads a list of alternative server-list-servers (in case the current server list gets shut down), and talks to one of those servers? Then that one server could farm the request out to the other servers it knows about. Sort of like gnutella at the core, napster at the leaves.
Is there any effort in place now that will bring all these alternatives into a seamless whole, either like I suggest above or in some other way? Or a Napigator-like interface that will search multiple servers at once for me?
Seems that Uranus has even more moons that Jupiter or Saturn. Should have looked deeper into the site.
So, for completness' sake:
Mercury: 0
Venus: 0
Earth: 1
Mars: 2
Jupiter: 17
Saturn: 18
Uranus: 21
Neptune: 8
Pluto: 1
Also, I believe that all moons (except Luna) are named after Roman mythological figures, except those of Uranus, which are from more modern literature (Shakespeare and Pope). This is not a recent phenomena, the first discovered moon of Uranus was named in 1787 (by Herschel). So he's the visonary who bucked the trend.:-)
Not to be picky, here, but I believe that Saturn has more moons than Jupiter. Of course, the poster didn't say that, but I just wanted to be sure that nobody inferred incorrectly.:-)
I'm sure I knew this at one point, but have forgotten -- what good is multi-regional compatibility if the discs you get aren't compatible with your home countrie's TV spec? Specifically, I saw some cheap(ish) DVDs while in the airport in Frankfurt recently, but they were labelled "PAL." I'd always thought (apparently, wrongly) that DVDs were "specification neutral," that is, that the player converted the DVD stream to whatever local video format you were on. Am I way off base here?
Put another way, what's out in other regions that I can't find here (in the US) cheaper? Different, out-of-print movies? Better featured discs? I can't help but notice that whenver I hear about imported DVDs (and LDs, to a lesser degree), it's usually about Anime (which is fine, I suppose, but not my personal cup of tea...)
The one thing that my (admitedly quick) perusal of responses here (and a survey of co-workers) shows is that nobody believes they'll get away with it.
Don't doubt for a second that the IOC can win this. They've successfully sued long-established businesses from using the term "Olympic" (as someone did mention -- Atlanta 1996).
The truly frightening thing, and I've asked about this before, is the fact that the understanding that "Trademarks are only for a certain domain of products" is very quickly withering away. Pilot pens suing over PalmPilot is but one example.
The big question, in my mind, is why is that principal tenet of Trademark Law disappearing? And why doesn't it (apparently) apply to the internet? Fact is, I *should* be allowed to register, say, coke.com, even if I have *nothing whatsoever* to do with CocaCola. As long as I don't sell a cola called 'Coke,' I'm not violating any trademark.
Just because it's stupid doesn't mean it can't happen.
This is the first of the two pulled stories, as found on DejaNews (http://x57.deja.com/getdoc.xp?AN=6442993 79). The second story, reportedly, was only up for a very short time and I couldn't find that one online, at least not this morning. Apple's "Cube" desktop Mac confirmed
After months of obscure reports and unreliable sources, two contacts with extremely solid track records have reported sightings of one of Apple's best-kept secrets -- its next-generation Desktop enclosure and the changes to its product line that will come with it.
While one of these new reports claims intently that the machine -- a near-perfect cube about 2/3 the size of a modern-day iMac -- will be a complete replacement for today's iMac line-up, the other is not so sure....and the remaining body of evidence is none too clear on whether this enclosure will be applied to the PowerMac or iMac spaces exclusively, or in both.
Setting that important detail aside for the moment, there is much exciting news about this new Cube that is of much greater reliability:
[UPDATED] Approximately 14 inches to a side.
Clear polycarbonate plastics similar to today's Macs offer views of the machine's innards, while opaque colored panels sport a large Apple logo and provide a sleek look.
Although multiprocessor G4 applications would likely require more powerful cooling, prototypes are fanless. Cooling is provided by numerous large vents on the top and bottom of the enclosure, allowing heat to naturally rise upward and out of the cube.
All six faces of the cube are featureless, aside from the power cable and an almost-invisible ports panel on the "back," which handles USB, Firewire, Audio I/O, Ethernet, and the built-in Modem's phone jack.
A small tab on one of the "sides" allows for that side to be opened for access to the machine's internals.
The entire package, including motherboard and all components, weighs approximately ten pounds.
There are significant signs that this may be the long-rumored monitorless iMac; for example, the prototype sources have reported on does not appear to have external ports to accept PCI expansion cards. However, there does appear to be enough internal room for them if the external ports were added.
A related but as yet unconfirmed rumor states that Apple is moving away from including PCI slots by default in PowerMacs, instead wiring the Universal Motherboard Architecture's PCI controller to a small connector which would support an external PCI enclosure with any number of slots. Note that this would be much less expensive than a full-blown PCI Expansion Chassis, which connects a single internal PCI slot to any number of add-on slots via a costly PCI bridge chip and associated hardware. This scheme would merely move PCI expansion outside the default PowerMac enclosure to allow for more innovative small-footprint designs as well as support more than three PCI slots for those who need them.
For now, all but the details of the Cube enclosure itself are to be considered highly speculative. We will be watching developments in this story very closely -- if you believe you may be able to clarify matters, drop us a line!
"France's laws on privacy are very strict and in a world where one's rights of privacy are being challanged all the time, it's good to see one country taking a stand."
Before we all get on the "France is standing up to government intrusion" bandwagon, let's remember that France is well-known for "borrowing" travelling executives' laptops at customs, long enough to copy the entire hard drive. And they have, at least once, completely trashed the hotel rooms of major aerospace executives during the Paris airshow to gather information.
France has been in the forefront of government-sponsored economic espionage for some time, so this whole "Echelon is bad" is fairly hypocritical.
Of course, I'm not sure this answers the question posed -- it just shows how frequencies are used, but doesn't show how much "bandwidth" is available.
I'm not sure of the easiset way to answer that question, anyway -- think about telephone lines, for example. Used to be, everyone figured that they had an "audio bandwidth" of about 3000 Hz (or am I way off here?) So you might figure that means about 3kbps total maximum throughput. However, we're getting 56k (or so) over those same lines, through clever use of multiple channels, multiple bits per baud, etc, etc.
Put another way, 1 MHz of radio bandwith does not equal 1 million bits per second, at least not as far as my limited knowlege implies.
So, maybe, the question is really this: If we scrapped all existing modulation systems (FM, AM, whatever), turned all communications into digital bits, and selected the best (most efficient, best range, etc.) scheme for modulating and encoding those bits, what's the maximum bandwidth available? Interesting question, but basically academic, 'cause I don't see everyone throwing out all their TVs, radios, and cell phones for a maximum-efficiency digital system.
And, besides, isn't sub-space communicaiton right around the corner?:-) david.
So at this point in time, remote administration of a Mac OS X machine needs to be done either with a destablizing, single-user remote control program like Timbuktu, or with the Unix command line
Not true. NeXT's Display Postscript (DPS) allowed for remote redirection of any program's input/output, from Day One (1989, IIRC, it's been a while.) It worked like a champ, too.
Now I know that MacOS-X uses PDF-based technology instead of DPS, but I assume it still has the remote viewing capability.
There's fourty-some-odd fields of enterprise, each of which has its own trademark namespace, so to speak.
Yeah, ideally, but then why is that aspect of trademark law being ignored so often lately? Like for domain names? Or the Pilot Pen versus Pilot PDA issue? one's a pen, one's an electronic organizer, I really don't know why the heck there'd be a tradmark issue, but the pen company sued Palm, and Palm backed down.
It just seems to me that we've been moving to a "global trademark space" lately, and that scares me.
Worms, on the other paw, are self-contained programs which contain nothing but the worm itself.
But I always thought a worm had to be self-mobile, too -- that is, the LOVEBUG needs human intervention to become active on a system, while (in my mind) a worm would simply connect directly to that machine and immediately begin executing.
The Morris Worm worked that way -- once released, it had a life of its own without any need for anyone to click on an attachment. If you consider a replaced.jpg file or a faked attachment to be executables (they are, just with no other functionality), then this is really much closer to the classic "virus" def.
As I get my mpeg server up and running at home, I've been seriously considering building a voice inteface to it. However, one problem that I'm running into is that, in my mind, all computers are supposed to sound like the Star Trek computer (voiced by Majel Barret Roddenberry -- the only person to be in *all* Star Trek incarnations (well, mayby not the animated ones..)).
Anyway, has anyone built (or does Paramount sell) a set of phonemes with a decent text-to-speech converter that'll make my computer sound like the star trek computer?
Interesting that this has been predicted to "be pretty nasty," as I seem to recall reading a while back that the Milky Way is already colliding with a smaller galazy, located somewhere off in Sagittarius.
The Astronomy Picture of the Day has a picture (sort of) at http://antwrp.gsfc.nasa.gov/apod/ap9 60204.html. Astronomers now believe that this galaxy [the Sagittarius Dwarf]is slowly being torn apart by the vast gravitational forces of our Galaxy.
Customer Service: I understand that Netpliance "got their model beat" by this hack, but I still feel a good deal of animosity against the company for the way that they handled the entire situation. So, specifically, I'd suggest that they react in a careful, planned, methodical fashion to future issues -- rather than the panicked, haphazard, harsh way that they did. Also, I'd definitely invest in improving customer service in general -- long waits on the phone, non-response to email, customer service computers crashing, etc., don't endear geeks like me to a company, so how will Grandma deal with problems? If Netpliance wants to be the Amazon of internet access, they'd better get Amazon-level service.
Hardware costs: Okay, so maybe it's worth more than $99. But you'll never sell it to grandma for $400 (at least not if they've got grandchildren like me who'll advise them to buy a real computer). I'd agree with another poster who suggested a multi-tier structure for purchasing: free for 3-year contract, $50 or $100 for 2-year, $200 for 1-year, $350-$400 for no contract. Honestly, if Netpliance had told me, upfront, after the story broke on April 11, that they'd require a contract, but I could buy it outright with no contract for $300, I'd still have bought it with no hesitation. You simply can't buy this kind of system, anywhere else, now (at least not easily and pre-assembled).
Hardware Improvements: It's looking like I'm going to spend an additional $300 on the box just to add a hard drive, ethernet, and a pair of serial ports (via usb). This is not to mention hardware-level hacks, like a drive LED, external audio ports, and a quiet fan. I'd suggest providing a "broadband" version that includes ethernet instead of a modem (or perhaps have the modem be a module, replaceable by an ethernet module, like on some network switches). I'd also suggest building in audio plugs (line in/out, headphones, microphone) for general use. Serial ports are important to me (IR control of Winamp and a Palm cradle), so adding those makes sense, too. Several people have suggested redesigning the heat sink so the heat flows upwards and out (the current design tends to trap heat) -- an efficient heat sink is always preferable to a noisy fan, but there may not be much you can do. Perhaps use the Transmeta chip instead of Intel clones, for power and heat benefits. Also, maybe consider a larger-screen version (would cost considerably more, unfortunately...)
Software Changes: I really like the idea of the deadicted QNX software with automated updates, etc. However, I do wonder if having a local filesystem (requiring a hard drive) might be good -- I think of the example of Grandma, receiving pictures of her newest grandchild, not wanting to have to download that from the server every time she looks at it. Better to have that local. But you don't want to go too far towards a full-featured computer (unless you wanna support, say, java-office or something), simply because that's not the point of the box (from Netpliance's point of view--some of us may go closer to that, but not all the way).
All in all, I'm impressed with the hardware as it is, but think that ethernet is a real requirement, especially as people upgrade to broadband services. A better, bigger screen would be welcome, but may not be cost effective. Offering a no-contract option at 300-400 dollars would be great for people who want to push the envelope with local, "thinner" computing. And, finally, Customer Service is of paramount importance -- companies live or die on how well they treat their customers, and, face it, Netpliance has not treated anyone -- hackers or grandmothers -- well this past month.
I honestly did not expect Netpliance to last the year, based on the service and bumbling responses we've seen. But attempts like this to get ideas from the internet community gives me new hope. We all need a company like Netpliance, and I wish you the best. Keep your minds open and your hardware/software accessible, and I can promise you the internet community will contribute ideas and fixes back to you.
That's the book that everybody told me to get when I took a compiler class, back in '91. However, I didn't get it (and ended up re-taking the class in '92, but that's another story:-) ).
I got a terrific book, but I can't remember the title/author to save my life. It's a large format (maybe 9x12"), blue cover, and thin (maybe 100 pages), but it was a godsend -- lots of good examples, etc. Not so much theory as practical applications and examples.
I'll try to remember to look for it tonite at home.
I haven't been able to play with DP3 (anyone wanna send me a CD? I promise not to tell!:-) ), so this may or may not apply. But...
...one really nice thing about NeXTSTEP was that the/usr,/adm,/etc, etc. folders were all actually symbolic links to/private/XXX. That way, all configuration files that made that machine "unique" were located in a single place, that could then be remote-mounted from an NFS server, for truly thin (even diskless, ignoring the need for swap) clients. Also made backup a lot simpler...
Slashdot Mirror
I've wondered if it shouldn't be possible (or even easy) to set up a mirror that vibrates on two axes, one horizontal and one vertical, and then to point tri-color LEDs at that mirror and focus their converged beams on the wall.
Basically, I figure a 3x3 array of bright LEDs should be more than bright enough, once focused, to illuminate a small pixel space on the wall. Get three of these, one for each color (now that you can find blue LEDs easily), and you've got a full-color pixel. I'd be reluctant to use lasers, both for obvious eye-danger reasons, and also for the (presumed) expense of blue and yellow beams.
In my imaginings, I figure that the horizontal and vertical synch signals could (if properly shaped -- can anyone point me to a site that might show me waveforms for such signals?) even drive a mirror deflector directly -- just hook one corner of a mirror to a pivot point, and hook each of the opposite two edges to a small speaker, and connect the speaker coil to the sync signals (or to re-shaped signals, if necessary).
I'm sure I'm missing something here, but what if I'm not? Could it be possible to build a home-brew HDTV projector for under $300 in parts? Anyone out there care to investigate this further?
For large documents, obviously, you're not performing a signature operation on all the text. But, maybe at the bottom of the page, you put your unique public key ID (which is then used to find your public key in a big database), then another line with, say, date, cost, and PO #, then the calculated result. If you lose your little card, you simply go down to the post office (or somesuch), get a new one, and they invalidate the old one for any new use after date X.
Obviously, the big problem here is the public key database, but that's been the bugaboo all along. But the advantages of something like this are:
Disadvantages:
You could (and, I'd argue, should, with proper back-end privacy features) put proximity technology into this and use it as your gas speedpass, grocery-store bonus card, office key, and gym pass. I like the ideas of the prox-cards (with authentication) for checkouts (like the Mobil Speedpass) but am loathe to put a dozen dongles on my already too-heavy keychain.
Anyway, does anyone like this idea? Can anyone point me to a better way to do secure authentication/validation on paper or over the phone? (yes, I'm ignoring for the moment the possiblity of loss/theft of the card and/or PIN).
david.
And there's a script ("Build your own location manager" or somesuch) that does exactly this on MacOSRumors.com.
A related question hinges on technical standards. As part of my suggestion above, I'd like to see a standardized box (that I can buy from any manufacturer, at WalMart) that will convert the fiber signal into TV, Telephone, Data, or whatever. Then, if I want to change my cable provider, I simply call my current provider, say "drop me," call a new provider, say "I want to give you my money," then walk downstairs and it should be switched over. There's no technical reason why this cannot happen today -- except that there are too many competing and incompatible standards.
At many times in the past, the government has refused to endorse a single technological standard (AM Stereo, HDTV (until late), and Cell Phones are examples). The rationale behind this is comendable: "Let the market decide." The end result, however, is that we have several competing and incompatible standards out there, and the industry is unable to settle on any single one -- saddling users with hardware supporting legacy sytems, the inability to switch between providers without buying a new phone, and so forth. How would you rectify this situation (both for new needs, such as the "one connection" box I describe above, and to bring the US into the 21st century to finally catch up with the Finns and the Japanese)?
I noticed that virtually none of the Digital UNIXes are on there. There's Digital Unix, the outgrowth of OSF, but nothing about Ultrix (unless I missed it). Hardly a fringe variant, Ultrix got as high as 4.0 and was on Vaxen, workstations, RISC workstations, all sorts of stuff. Being as how this is the UNIX I learned on, I'm kinda miffed to see it missing. :-(
It's also interesting to note that, as recently as a few years ago, DEC was still using their standard character generator for their PCI-based Alpha machines -- now that's what a UNIX console is supposed to look like!
-- Offtopic --
The quote at the bottom, as I read this:
Reminds me of the time in school when I found a phone number that was one digit off from mine, but not currently in service. I recorded the intercept message, put it as my outgoing message, and added "but that's not my number, and I'm not here anyway." Best message I ever did, but really pissed off all my friends and family... :-)
I know that several organizations (the Washington Post, for example) have printed photographs (or partial photos) of the DeCSS t-shirts. Has anyone thought about collecting those various photographs together, with proper attribution, and putting them on a web page? I'd bet if you get enough of them any reasonably industrious person could piece together the original code...
What would happen to the trade secret issue then? Collected legitimately from multiple legally published photographs?
Just a thought....
I'd say that any on-line store should require only the same information that is required at a real physical store. That is, if I'm paying for groceries with cash, then when paying with an accredited anonymous cyber-cash-like operation, I should need to provide no information at all. If I'm paying with a credit card at a store, then all the online-store needs is my number and signature. What? They can't get my signature over the wire? Okay, then, I guess they need whatever my credit card agreement says they need -- usually, an address.
In many cases, this is going to be up to the financial companies (banks, credit card companies, etc.) to find alternative ways of validating and authenticating transactions, without divulging address, telephone numbers, etc., to online merchants. Not sure this is ever going to happen. 'course, there's stuff like PayPal, but who knows how long until *they* start doing something with their information.
Of course, we have to remember that in many cases, the business model of online companies may actually include revenue from information collected during the transaction. You see this in bricks-and-mortar stores at, say, supermarkets, with the "special discount cards" that they give people. To be very literal about it, a certain loss of some amount of privacy (some shopping/clicking habits, etc.) are the true price we pay for discounted prices, vast inventories, and free overnight shipping. Don't like the loss of anonymity? Go to your local store. Sucks, yeah, but that's the way the internet works.
In this particular instance, as someone else pointed out, it's likely the problem is that they want to auto-generate their pages to match your browser. Here the problem isn't privacy, but a closed-mindedness as to what browsers are out there. This used to be such a nasty problem when I was surfing from my NeXT that I had to pretend I was Netscape (which was a built-in feature of the browser for just this issue!). They really should have some way of providing a general, simple HTML interface that anonymous browsers can read. Or maybe we need new brower identifications that don't ID the brower, but instead define the browser's capabilities.
As someone who was lucky enough to try the original browser, probably within days of its release, I find this annoying. :-)
Check out this web page http://www.w3.org/People/Ber ners-Lee/WorldWideWeb.htmlc .gif.
or this nice screenshot: http://www.w3. org/History/1994/WWW/Journals/CACM/screensnap2_24
Note that this wasn't just a browser -- it was a "Browser - Editor!" Very advanced for its time, eh?
(sorry...its just always ticked me off that credit isn't given where it's due on this one...)
david.
One of the great advantages of Napster (and Gnutella) is that you're accessing a single, common community. With dozens of alternate napster networks cropping up, that community is going to be splintered into many different ones. And people will have to search, possibly, many of the different servers before they find one that's got what they're loking for.
Wouldn't a better approach be some kind of client that combines the best of Napster and Gnutella? Connects to a list of servers, downloads a list of alternative server-list-servers (in case the current server list gets shut down), and talks to one of those servers? Then that one server could farm the request out to the other servers it knows about. Sort of like gnutella at the core, napster at the leaves.
Is there any effort in place now that will bring all these alternatives into a seamless whole, either like I suggest above or in some other way? Or a Napigator-like interface that will search multiple servers at once for me?
So, for completness' sake:
Also, I believe that all moons (except Luna) are named after Roman mythological figures, except those of Uranus, which are from more modern literature (Shakespeare and Pope). This is not a recent phenomena, the first discovered moon of Uranus was named in 1787 (by Herschel). So he's the visonary who bucked the trend. :-)
Not to be picky, here, but I believe that Saturn has more moons than Jupiter. Of course, the poster didn't say that, but I just wanted to be sure that nobody inferred incorrectly. :-)
[Thanks to Students of the Exploration and Development of Space (www.seds.org) for the reference. Check out for a list of all the named moons for both.]
Put another way, what's out in other regions that I can't find here (in the US) cheaper? Different, out-of-print movies? Better featured discs? I can't help but notice that whenver I hear about imported DVDs (and LDs, to a lesser degree), it's usually about Anime (which is fine, I suppose, but not my personal cup of tea...)
The one thing that my (admitedly quick) perusal of responses here (and a survey of co-workers) shows is that nobody believes they'll get away with it.
Don't doubt for a second that the IOC can win this. They've successfully sued long-established businesses from using the term "Olympic" (as someone did mention -- Atlanta 1996).
The truly frightening thing, and I've asked about this before, is the fact that the understanding that "Trademarks are only for a certain domain of products" is very quickly withering away. Pilot pens suing over PalmPilot is but one example.
The big question, in my mind, is why is that principal tenet of Trademark Law disappearing? And why doesn't it (apparently) apply to the internet? Fact is, I *should* be allowed to register, say, coke.com, even if I have *nothing whatsoever* to do with CocaCola. As long as I don't sell a cola called 'Coke,' I'm not violating any trademark.
Just because it's stupid doesn't mean it can't happen.
After months of obscure reports and unreliable sources, two contacts with extremely solid track records have reported sightings of one of Apple's best-kept secrets -- its next-generation Desktop enclosure and the changes to its product line that will come with it.
While one of these new reports claims intently that the machine -- a near-perfect cube about 2/3 the size of a modern-day iMac -- will be a complete replacement for today's iMac line-up, the other is not so sure....and the remaining body of evidence is none too clear on whether this enclosure will be applied to the PowerMac or iMac spaces exclusively, or in both.
Setting that important detail aside for the moment, there is much exciting news about this new Cube that is of much greater reliability:
There are significant signs that this may be the long-rumored monitorless iMac; for example, the prototype sources have reported on does not appear to have external ports to accept PCI expansion cards. However, there does appear to be enough internal room for them if the external ports were added.
A related but as yet unconfirmed rumor states that Apple is moving away from including PCI slots by default in PowerMacs, instead wiring the Universal Motherboard Architecture's PCI controller to a small connector which would support an external PCI enclosure with any number of slots. Note that this would be much less expensive than a full-blown PCI Expansion Chassis, which connects a single internal PCI slot to any number of add-on slots via a costly PCI bridge chip and associated hardware. This scheme would merely move PCI expansion outside the default PowerMac enclosure to allow for more innovative small-footprint designs as well as support more than three PCI slots for those who need them.
For now, all but the details of the Cube enclosure itself are to be considered highly speculative. We will be watching developments in this story very closely -- if you believe you may be able to clarify matters, drop us a line!
Before we all get on the "France is standing up to government intrusion" bandwagon, let's remember that France is well-known for "borrowing" travelling executives' laptops at customs, long enough to copy the entire hard drive. And they have, at least once, completely trashed the hotel rooms of major aerospace executives during the Paris airshow to gather information.
France has been in the forefront of government-sponsored economic espionage for some time, so this whole "Echelon is bad" is fairly hypocritical.
For further information, check out Secure Computing's April 1998 issue (the first hit I got from google, I'm sure there's plenty more out there): http://www .westcoast.com/securecomputing/1998_04/cover/cover .html#French.
david.
Of course, I'm not sure this answers the question posed -- it just shows how frequencies are used, but doesn't show how much "bandwidth" is available.
I'm not sure of the easiset way to answer that question, anyway -- think about telephone lines, for example. Used to be, everyone figured that they had an "audio bandwidth" of about 3000 Hz (or am I way off here?) So you might figure that means about 3kbps total maximum throughput. However, we're getting 56k (or so) over those same lines, through clever use of multiple channels, multiple bits per baud, etc, etc.
Put another way, 1 MHz of radio bandwith does not equal 1 million bits per second, at least not as far as my limited knowlege implies.
So, maybe, the question is really this: If we scrapped all existing modulation systems (FM, AM, whatever), turned all communications into digital bits, and selected the best (most efficient, best range, etc.) scheme for modulating and encoding those bits, what's the maximum bandwidth available? Interesting question, but basically academic, 'cause I don't see everyone throwing out all their TVs, radios, and cell phones for a maximum-efficiency digital system.
And, besides, isn't sub-space communicaiton right around the corner? :-) david.
Anyone know of a good spectrum allocation chart on the web?
Not true. NeXT's Display Postscript (DPS) allowed for remote redirection of any program's input/output, from Day One (1989, IIRC, it's been a while.) It worked like a champ, too.
Now I know that MacOS-X uses PDF-based technology instead of DPS, but I assume it still has the remote viewing capability.
Yeah, ideally, but then why is that aspect of trademark law being ignored so often lately? Like for domain names? Or the Pilot Pen versus Pilot PDA issue? one's a pen, one's an electronic organizer, I really don't know why the heck there'd be a tradmark issue, but the pen company sued Palm, and Palm backed down.
It just seems to me that we've been moving to a "global trademark space" lately, and that scares me.
But I always thought a worm had to be self-mobile, too -- that is, the LOVEBUG needs human intervention to become active on a system, while (in my mind) a worm would simply connect directly to that machine and immediately begin executing.
The Morris Worm worked that way -- once released, it had a life of its own without any need for anyone to click on an attachment. If you consider a replaced .jpg file or a faked attachment to be executables (they are, just with no other functionality), then this is really much closer to the classic "virus" def.
david.
Anyway, has anyone built (or does Paramount sell) a set of phonemes with a decent text-to-speech converter that'll make my computer sound like the star trek computer?
The Astronomy Picture of the Day has a picture (sort of) at http://antwrp.gsfc.nasa.gov/apod/ap9 60204.html. Astronomers now believe that this galaxy [the Sagittarius Dwarf]is slowly being torn apart by the vast gravitational forces of our Galaxy.
Customer Service: I understand that Netpliance "got their model beat" by this hack, but I still feel a good deal of animosity against the company for the way that they handled the entire situation. So, specifically, I'd suggest that they react in a careful, planned, methodical fashion to future issues -- rather than the panicked, haphazard, harsh way that they did. Also, I'd definitely invest in improving customer service in general -- long waits on the phone, non-response to email, customer service computers crashing, etc., don't endear geeks like me to a company, so how will Grandma deal with problems? If Netpliance wants to be the Amazon of internet access, they'd better get Amazon-level service.
Hardware costs: Okay, so maybe it's worth more than $99. But you'll never sell it to grandma for $400 (at least not if they've got grandchildren like me who'll advise them to buy a real computer). I'd agree with another poster who suggested a multi-tier structure for purchasing: free for 3-year contract, $50 or $100 for 2-year, $200 for 1-year, $350-$400 for no contract. Honestly, if Netpliance had told me, upfront, after the story broke on April 11, that they'd require a contract, but I could buy it outright with no contract for $300, I'd still have bought it with no hesitation. You simply can't buy this kind of system, anywhere else, now (at least not easily and pre-assembled).
Hardware Improvements: It's looking like I'm going to spend an additional $300 on the box just to add a hard drive, ethernet, and a pair of serial ports (via usb). This is not to mention hardware-level hacks, like a drive LED, external audio ports, and a quiet fan. I'd suggest providing a "broadband" version that includes ethernet instead of a modem (or perhaps have the modem be a module, replaceable by an ethernet module, like on some network switches). I'd also suggest building in audio plugs (line in/out, headphones, microphone) for general use. Serial ports are important to me (IR control of Winamp and a Palm cradle), so adding those makes sense, too. Several people have suggested redesigning the heat sink so the heat flows upwards and out (the current design tends to trap heat) -- an efficient heat sink is always preferable to a noisy fan, but there may not be much you can do. Perhaps use the Transmeta chip instead of Intel clones, for power and heat benefits. Also, maybe consider a larger-screen version (would cost considerably more, unfortunately...)
Software Changes: I really like the idea of the deadicted QNX software with automated updates, etc. However, I do wonder if having a local filesystem (requiring a hard drive) might be good -- I think of the example of Grandma, receiving pictures of her newest grandchild, not wanting to have to download that from the server every time she looks at it. Better to have that local. But you don't want to go too far towards a full-featured computer (unless you wanna support, say, java-office or something), simply because that's not the point of the box (from Netpliance's point of view--some of us may go closer to that, but not all the way).
All in all, I'm impressed with the hardware as it is, but think that ethernet is a real requirement, especially as people upgrade to broadband services. A better, bigger screen would be welcome, but may not be cost effective. Offering a no-contract option at 300-400 dollars would be great for people who want to push the envelope with local, "thinner" computing. And, finally, Customer Service is of paramount importance -- companies live or die on how well they treat their customers, and, face it, Netpliance has not treated anyone -- hackers or grandmothers -- well this past month.
I honestly did not expect Netpliance to last the year, based on the service and bumbling responses we've seen. But attempts like this to get ideas from the internet community gives me new hope. We all need a company like Netpliance, and I wish you the best. Keep your minds open and your hardware/software accessible, and I can promise you the internet community will contribute ideas and fixes back to you.
Thanks for listening!
That's the book that everybody told me to get when I took a compiler class, back in '91. However, I didn't get it (and ended up re-taking the class in '92, but that's another story :-) ).
I got a terrific book, but I can't remember the title/author to save my life. It's a large format (maybe 9x12"), blue cover, and thin (maybe 100 pages), but it was a godsend -- lots of good examples, etc. Not so much theory as practical applications and examples.
I'll try to remember to look for it tonite at home.
david.
I haven't been able to play with DP3 (anyone wanna send me a CD? I promise not to tell! :-) ), so this may or may not apply. But...
/usr, /adm, /etc, etc. folders were all actually symbolic links to /private/XXX. That way, all configuration files that made that machine "unique" were located in a single place, that could then be remote-mounted from an NFS server, for truly thin (even diskless, ignoring the need for swap) clients. Also made backup a lot simpler...
...one really nice thing about NeXTSTEP was that the
david.