I'm assuming that the cache memory is used to hold the translated information. The amount varies by processor - the Crusoe 5600 in the Loox T has twice as much as the 5400 in the Loox S, even though they're at the same speed. I'm not sure offhand how the 3200 for mobile devices compares.
Similarly, I suspect (because it seems obvious) that the processor keeps track of how often translated code is used so it can discard the less-used translations when it starts to run out of space. It's like virtual memory, but instead of getting paged out to disk the less-needed translations are discarded since they can be redone if they're needed again. It sounds bad to repeat work, but the impact of re-translating is likely to be fairly small, because the code that needs it doesn't get called as often as the code that's kept.
One other thing that might be interesting is what sort of changes might be made to the code morphing software to improve the optimization. Perhaps it doesn't do a good least-recently-used discard of optimized code, but a processor firmware update adds that capability. What kind of impact will that have on system speed? I don't know if updates like that would be free, but would you pay $20 for a 10-15% improvement processor speed on your system?
It's interesting that Transmeta has four areas of potential improvement where Intel and AMD have only three. The first is improving processor speed, the second is improving the cache size and/or speed, the third is architecture improvements, and the third is better code morphing software (which is a form of architecture improvement). Intel has its BIOS update to let it address fixes in the same way that Transmeta could, but that's not going to make a difference in the overall performance of a processor.
I suspect that internally Transmeta is continuing to test processors to find out things like how much cache memory is needed to keep frequently-used translations from being discarded. I'm sure we'll never publicly see those, but eventually it'll probably be clear from the point at which Transmeta stops increasing the amount - when the amount of local cache stops growing it means they've either hit the point where most of what gets discarded wasn't going to be used again anyway or they've hit the point where indexing into the cache becomes a problem.
Infoworld is an industry rag targeted at business readers, and UCITA is out of the worst nightmares of most businesses.
I haven't read through it lately so this is from memory, but in addition to all the controls on liability, lawsuits, etc., I believe there's a difference in how licensing is handled for individuals and corporations - individuals have the option of returning the software for a refund if they don't like the click license. Businesses get no such protection.
It looks like Netpliance has given up on their original plan of selling it as a loss leader and requiring a long-term contract. They're now $299, with the service at $21.95/month, cancellable anytime.
How would something like this be handled with a public company or one looking at an IPO?
A court case like this clearly has potential to impact future business, however if the company was prohibited from mentioning it in filings would they end up with a choice of being screwed by the judge for violating the gag order or being screwed by the SEC for misleading filings?
One thing that I've wondered about:
Since WML is likely to remain restricted to mostly text because of the nature of the display devices, would a PC-side WML viewer that worked with screen readers or Braille devices be practical?
I'd think this would avoid the increase in graphics use (and abuse) that's caused some problems for blind users of the Web.
I believe that it's not unusual for the commercial news providers to cap bandwidth for users; perhaps SBC is doing the same? I suspect they're running the same software as some of the commercial/non-ISP providers.
If they're doing the same for email I'm surprised, but stranger things have happened. I'd expect them to cap message size first, though - how many mail clients out there are set up to automatically split and join multipart messages?
I was wondering not too long ago whether something like this could be done by libraries - kind of a cross between community moderation by library staff and the MAPS or ORBS lists. There could be both positive and negative rankings, and presumably would be a variety of categories.
My mental image of the whole thing is a library with controlled and uncontrolled (or minimally-controlled) terminals, where uncontrolled terminals were only available to adults and controlled terminals were available to children. Uncontrolled terminals could access pretty much anything; controlled terminals could access sites already in the database with a ranking (within an category) above a level set by the local staff. Unlisted sites being accessed from the controlled terminals would be bounced to someone else for approval (I'm imaginging this going to someone supervising the children's area of the library). Alternatively, children's terminals could be in one of two modes: private but limited to approved sites or monitored but allowed pretty much uncontrolled access. That way if little Johnny is looking at porn/bomb howtos/drug recipes/whatever, someone shadowing the terminal can see it and either shut down access to the site or go have a talk with Johnny.
On the moderation side, scoring would be a bit more complex than/.'s - moderators could be ranked, and moderators with higher ranks would have more influence on the scoring. Similarly, moderators with low ranks or moderators who abused the system could be given less influence and eventually dropped. By doing this, you could hopefully prevent "poisoning" of the system by people with agendas. Ranking could be handled by something similar to/.'s meta-moderation, but probably more automated. If librarian A puts a site in a category and librarians B, C, and D all disagree with that assesment, A's ranking (overall or within that category) goes down. There could even be a way for the general public to "nominate" sites for moderation with suggestions as to categories.
The biggest headache with the whole thing, I think, would be who does the moderating, particularly early on - it's not like most libraries have people sitting around doing nothing who could just take it on as an additional duty. Unfortunately, it's not like there's a good base list out there that could be used as a seed.
One interesting thing is that if such a system was developed it could have commercial potential as well - perhaps a version of it that did not include the ability to moderate, but which did allow the use of frequently-updated copies of the score files from libraries. By providing it to libraries for free, a company would both help libraries avoid community pushes for censorware with an agenda and gain access to a pool of site raters.
Overall, there are all sorts of potential approaches to the problem of protecting people from disagreeable information, the problems are that there are so many different ideas of what's considered disagreeable and there are so many different ideas of who needs to be protected. On one end, you have people who believe that what's appropriate is no restrictions at all; on the other you have people who believe that all access should be tightly controlled; in between you have the bulk of the populace. In the US the general tendency is probably toward less control, but the tight-control people are noisy. Assuming that somewhere in the middle (limited control) is where things are going to end up, the problem becomes one of ensuring that the controls that go into place are not too restrictive, and that's where I think approaches like this could come in.
Finally, before people jump all over this as advocating censorware, I lean toward the less-control end personally, but I think that there are some things that need to be discussed with children before they're exposed to them unsupervised. Children who think and parents who teach them to do so are the way to go, but in our current society I'm not sure how much we can depend on that.
FP: Any check printed by a consumer on a home PC is going to be spit out like this
B: Why do you say this? Is the tolerance of the positioning tighter than 1/300th of an inch? (I don't have the spec handy...)
The tolerances are tight, but it'd be impossible to get them that tight. The reason consumer-printed checks will spit out is because the MICR lines aren't magnetic.
These days that may mean that they just dump into another system where they'll be OCRed, but a few years ago that wouldn't have been an option and it may not be now. While I know there are check reading systems out there that do imaging, the ones I know of are geared toward businesses rather than banks - I have no idea what the central banks use or how fast their technology changes.
MICR ink (actually almost always toner) is special, in that it contains much more iron than regular toner.
This is important because check readers aren't actually doing any kind of OCR - when they started using MICR lines OCR was a dream even on a very primitive level. The numbers (and the 4 special characters) look that way because when you run them past a reader the waveforms / signals they produce are very identifiable. Think of them as being an early version of a magnetic stripe, but non-writable.
Position is also very important - individuals don't see it, but for any business printing its own checks in any significant quantity it's critical that the position of every character in the MICR line be correct. If the positioning isn't correct, the (often poorly adjusted) readers at the central processing banks will spit it out as a damaged check, and it gets special handling. Consumers never really see this, but for businesses their bank can actually be charged for the extra processing, and they're willing to pass that charge along to the business printing unscannable checks.
Any check printed by a consumer on a home PC is going to be spit out like this, but I'm not sure the banks are set up to deal with extra fees for those - they probably have the cost of a certain percentage of checks being "damaged" factored into their fees. This is also the reason that they prefer that you not fold checks - folded checks are both less likely to read and more likely to jam in the readers.
Could this be in part the judge's way of trying to get Microsoft to consider negotiating a settlement with some actual teeth?
If they settle, his opinion doesn't really matter.
If they don't settle and he takes the DOJ recommendations, an appeals court may ship it back for consideration of other options.
If they don't settle and he breaks MS into three companies, an appeals court may ship it back for consideration of other options - things like breaking it into only two companies.
If they don't settle and he doesn't break MS up at all, maybe MS doesn't appeal.
Whatever happens, we'll all get to hear about it eventually....
...the i-opener or the hackable webtv box - this one doesn't come with a monitor, and it probably doesn't output NTSC video either so you can't just use it with your TV.
I've seen reports that it uses a stripped down version of Linux, however....
Personally, I'd love to have a Palm V-sized system (or slightly larger) that could do limited voice recognition.
This could be something as limited as understanding the months, days of the week, numbers (possibly as high as the millions), years, AM/PM, morning/afternoon/evening/night, and maybe a few commands or other keywords like "with, alarm, minute(s), hour(s), todo, folder" and the like.
I even know how I'd use it - hit the record button, record a quick note or reminder ("Meet with Scott on Project X"), hold the record button, tell it where to store the recorded item: "Calendar May fifteenth at one pm with fifteen minute alarm" or maybe "Third Monday of every month at nine AM". Or maybe record "Listen to album by Artist X", then store it with "Folder music," or maybe "Todo purchase."
An awful lot of functionality could be managed with just a little bit of voice recognition and a small pre-defined dictionary of words. With the hardware available, enhanced dictionaries would be an obvious add-in, and as faster processors came along the number of distinct words that could be recognized would go up.
Overall I'd classify this as a win - if I was looking to purchase a new PDA right now I'd probably get a Palm Vx for its size, but I'd be tempted by some of the lighter CE handhelds for the recording capability. After all, I'm carrying a little digital voice recorder right now for notes when I can't write things down - should I carry two small pieces of electronics that rack up to about 8 ounces or one larger piece that weighs the same?
To purport that Netpliance could possibly hope to turn a profit off the ISP business is preposterous, to say the LEAST. ISPs lose more money every year than any other business I've seen. You know those $9.95/month unlimited ISPs? Ever wonder why it's slow and you get busy signals? Because they can't turn profit on that. That's why ISPs offer webhosting and such.
Figure that ISP XYZ in Sometown has 48 modems, and 200 subscribers, giving them roughly a 4:1 user to modem ratio - fairly good. However, ISP XYZ has had to invest [...]
I wouldn't be surprised if Netpliance hasn't invested much money at all in a dialup network - they're probably using resold service from UUNet, WorldCom, AT&T/IBM Global Network, or some other service. They'd pay a premium for it, but it avoids most of the up-front costs, scalability problems, and maintenance. It wouldn't even be all that difficult to build in remote-updating capability on the boxes so at some point they can start downloading new phone lists to people to switch to their own service.
As for the numbers tossed around in the previous article, there's a big difference between a $9.95/month ISP that's taking a loss and a $22/month ISP that might actually make a few dollars profit on each customer. Even if they're running their own dialup pools, they may be able to get away with a much higher ratio than 4:1, since they're not targeting the users who are likely to stay connected for long periods - why start a big download if you only have a few MB of flash to store it in and can't install new programs anyway? Basically, there are all sorts of ways to squeeze money out of a setup like this, including providing secondary services like shopping - Netpliance's stores would be the first place most of their users would probably go.
In terms of other savings, how much does your typical ISP spend on support? How much would it be if all customers were using only known hardware and software, didn't have significant download and local storage capacity, and didn't even have to dial in to get mail because it downloaded automagically 6 times a day?
In fact, they may have a single 800-number pool running right now for those mail checks; depending on the level of hardware control they have it should be possible to get message counts in a single call lasting less than a minute - don't bother with a data connection! Just call up, send a string of touch tones with some check digits to identify an account, and get a series of tones back indicating a message count. Much faster than negotiating a data connection.
They should see if they can find countries or cities willing to pay them to schedule burnouts right around the time of fireworks displays....
Their market is probably a bit limited since I believe they'll be bringing them down in the Pacific, but for parts of Asia they might get some real interest!
Well, my old Toshiba Portege 300CT rus at 1024x600 on a 10.4" diagonal screen with no problems at all. People are even running Linux and X on it.
There have also been a few other small systems (mostly from Toshiba, I think) that run at unusual resolutions - Librettos are the most common of them. I think one (a small system with a camera built into the edge of the screen) may even use 1024x480.
Resolution has always been my biggest concern with the newer lightweight laptops - for a long time they were only available at 800x600, and I really wanted that extra width.
Several years ago I was involved with a project for an AOL storefront; the customer service people for it all used female names because from practical experience they'd found that customers who'd had problems with orders were less likely to go ballistic on a woman than on a man.
This was particularly significant because the items being sold were personalized, and there was a policy in place that the personalization could not include certain selected words. If a customer used words from the "banned" list, they got a polite email saying that their order couldn't be completed and giving a reason why. I suspect that a lot of nasty responses were averted because the email came from "Cheryl" or "Sue."
For the indignant, the reason for the banned list was that the company in question didn't want to have products out there that visibly included both their name and even mild profanity. They were concerned about tarnishing their image, and given their wholesome reputation I can understand why (though I thought and still think their list was a bit too restrictive).
From just a preliminary reading of this, it sounds like it's along the lines of criticizing (example not based on actual holes) SuSE for selling Linux CDs that have a buggy version of Apache bundled on them.
It's a problem with a bundled software package that installs by default - how many Linux distros have been put together, then stayed on the shelves after someone found a hole or significant problem in a bundled package? Heck, how many of them have been sent to manufacturing then had something crop up after they started pressing discs and printing manuals?
You'd have to purchase that exact chip, read the information from the old PROM, change the serial number, write the new PROM, desolder the old chip, and solder in the new chip.
Worse than that, unless you're only using the copier for "nefarious" purposes you'd probably have to solder in a socket instead so that you could alternate between the original chip (or a copy of it) and your hacked chip. After all, at some point there's a chance that an identifiable document from that machine will be scanned and the numbers correlated; if that happens at most company locations, when they ask "Well, who might know how to do this?" everyone will immediately think of one person. Even worse, some of these copiers may have a connector that techs use during maintenance to pull information - things like the ID, number of copies, error logs, all sorts of interesting* stats.
So, the old chip has to be removed and read, a socket has to be mounted (using surface-mount tools that most people don't have available), new chips have to be burned, and it all has to be done without attracting attention by having the copier down for a significant amount of time so nobody calls in the techs. Oh, and that socket? Best make sure that board isn't something that gets looked at when the service guys are in, they might just notice something odd...
Overall while it's probably possible to remove the serial number or corrupt it into unusability, it's probably not feasible to do so.
As for the privacy issue... wouldn't such a encoding method be proprietary to the manufacturer? So what happens if I first copy the color image on a Xerox machine, and then take the copy over to a different machine, and copy that. Assuming the quality was not lost, the hidden ID code would not be decipherable by any (one) decryption algorythm.
Actually, what would probably work better is making a copy, then making a second-generation copy of it on either the same machine or another machine that uses the same ID method. That way there'd be some chance that the two sets of tags would interfere with each other. Of course, this would depend on what method was used to generate them and how they were placed on the output - it's conceivable that the methods used wouldn't be affected by this at all for some manufacturers.
As for the thought of using two different machines to generate the second-generation copy, that would help conceal which machine made the first-generation copy, but would probably still allow identification of the second-generation copy machine, which gives a location. If the second copy machine is going to be identifiable, why bother making a first-generation copy? For most situations there'd be no advantage to doing so - the primary times when it would be needed is if access to the original was limited, so a copy was needed for later distribution.
Sony sells its minidisc almost completely on the basis of its ability to make copies. Every single commercial i've seen for the minidisc consists of nothing but the people making copies of minidiscs. [...] But the RIAA has no problem with the minidisc.
There are two reasons why the RIAA doesn't have problems with minidiscs:
Sony has clout
Minidisc systems have copy-protection built in
If I'm remembering it correctly, minidisc copy protection works like this: there's a two-bit field on each disc that indicates what kind of disc it is. When you make a direct digital copy from a CD or another MD, that field is modified so that the copy can't be used as a source for another digital copy.
Of course, none of this keeps you from making analog copies of MDs, but because MD audio is actually compressed every additional generation adds additional "noise" to the copies. What this does is keep people from distributing first-generation copies of music - exactly what happens when someone rips a CD into MP3s and sends them out.
Switzerland not being part of the EU is perhaps not a huge issue; it seems that while the server is hosted there the primary members of the group are scattered around several countries.
In fact, this is an improvement in some ways - if there's a basis for a countersuit in either Switzerland or one of the EU countries, it might make sense to check the laws in both. One item that might make waves (at least PR, if not legal) would be an EU-based suit that presented eToys as attempting to force a European artistic group out of existence using US laws. Depending on the local feeling about US domination of the Internet, that could turn into a serious impediment for eToys. The delay by the US trademark office probably wouldn't hurt this either, given the relative startup times of the two groups.
Note that I'm assuming here that eToys will manage to at the least stretch this out for a substantial period by virtue of having a carefully selected judge in a court local to them - if they can exert enough influence they may win the case despite the lack of legal merits, simply because it's difficult for etoy to reach the location to provide an active defense.
A countersuit of the sort I'm talking about would certainly have legal merit at least as strong as eToys' case, but the larger issue might be the PR question - if they manage to build a reputation in Europe as a bullying US megacorporation that beats up local artists it could seriously impact any plans they might have to expand their business beyond the US. If the etoy folks can appear to be charmingly eccentric in court this would be even better - "They may be kooky artists, but they're our kooky artists!" Charm from eToys' suited lawyers is unlikely to be an issue.
The PR issues might be just what was needed to get eToys to discuss settling, assuming that etoy is interested in doing so rather than in fighting it out to the bitter end. A settlement would probably end with etoy agreeing to have a relatively inoffensive front page that indicates that it is not associated with eToys (and probably with a link to the eToys site). If that sort of thing does happen, I personally think that etoy should push for it to be reciprocal - a note at the bottom of the eToys front page that says "This site is not affiliated with the etoy.CORPORATION art group." Such a note would, of course, have to include a link to www.etoy.com.
It seems to me that EToy, if they so desire, might be able to return the favor to Etoys - a suit, filed in the EU where they're based, because Etoys is diluting the recognizability of EToy.
After all, EToy is recognized as a source of cutting-edge art and social commentary, and it provably existed long before Etoys came along and picked up a similar name. Anyone who goes to the Etoys site seeking art is going to be just as confused as someone going to the EToy site looking for toys.
If the argument that the site is hosted in Switzerland and should be in the.ch top-level domain is allowed to stand, then a counter-argument that Etoys being hosted in California should be in the.ca.us domain is just as valid. In addition, the precedent that the case would set should bring massive support to EToy from other non-US companies with.com domains.
The movement from court to court seems remarkably like a case of shopping for a favorable judge, and for an international matter a California state court seems like a particularly inappropriate venue. The requirement that EToy be physically present in order to participate also seems onerous, particularly for a preliminary hearing - it would make sense for the trial itself, but not for the early stages.
As an interesting thought, can EToy go through discovery and depositions to investigate all records and email from the early days of Etoys in an attempt to determine whether the company was aware of the potential conflict? The depositions in particular would be better handled in a case in the EU - if the creators of EToy must travel to the the site of a US trial, it makes sense that the founders of Etoys be required to travel to the site of a Swiss one.
Even better, on thought: if in discovery it is found that Etoys was aware that they had no real legal standing, would that open them up to a barratry (abuse of lawsuits for intimidation purposes) suit?
If nothing else, they should be appealing for emergency permission to put up a page that describes the situation - "This site is down due to a lawsuit between Etoys, Inc. and the EToy artist group. To reach Etoys, the online toy seller, click {here}. To reach EToy, the artist group, click {here}."
Slashdot Mirror
Is in #760 for those interested.
-- fencepost
Similarly, I suspect (because it seems obvious) that the processor keeps track of how often translated code is used so it can discard the less-used translations when it starts to run out of space. It's like virtual memory, but instead of getting paged out to disk the less-needed translations are discarded since they can be redone if they're needed again. It sounds bad to repeat work, but the impact of re-translating is likely to be fairly small, because the code that needs it doesn't get called as often as the code that's kept.
One other thing that might be interesting is what sort of changes might be made to the code morphing software to improve the optimization. Perhaps it doesn't do a good least-recently-used discard of optimized code, but a processor firmware update adds that capability. What kind of impact will that have on system speed? I don't know if updates like that would be free, but would you pay $20 for a 10-15% improvement processor speed on your system?
It's interesting that Transmeta has four areas of potential improvement where Intel and AMD have only three. The first is improving processor speed, the second is improving the cache size and/or speed, the third is architecture improvements, and the third is better code morphing software (which is a form of architecture improvement). Intel has its BIOS update to let it address fixes in the same way that Transmeta could, but that's not going to make a difference in the overall performance of a processor.
I suspect that internally Transmeta is continuing to test processors to find out things like how much cache memory is needed to keep frequently-used translations from being discarded. I'm sure we'll never publicly see those, but eventually it'll probably be clear from the point at which Transmeta stops increasing the amount - when the amount of local cache stops growing it means they've either hit the point where most of what gets discarded wasn't going to be used again anyway or they've hit the point where indexing into the cache becomes a problem.
-- fencepost
I haven't read through it lately so this is from memory, but in addition to all the controls on liability, lawsuits, etc., I believe there's a difference in how licensing is handled for individuals and corporations - individuals have the option of returning the software for a refund if they don't like the click license. Businesses get no such protection.
-- fencepost
It looks like Netpliance has given up on their original plan of selling it as a loss leader and requiring a long-term contract. They're now $299, with the service at $21.95/month, cancellable anytime.
-- fencepost
A court case like this clearly has potential to impact future business, however if the company was prohibited from mentioning it in filings would they end up with a choice of being screwed by the judge for violating the gag order or being screwed by the SEC for misleading filings?
-- fencepost
Since WML is likely to remain restricted to mostly text because of the nature of the display devices, would a PC-side WML viewer that worked with screen readers or Braille devices be practical?
I'd think this would avoid the increase in graphics use (and abuse) that's caused some problems for blind users of the Web.
If they're doing the same for email I'm surprised, but stranger things have happened. I'd expect them to cap message size first, though - how many mail clients out there are set up to automatically split and join multipart messages?
My mental image of the whole thing is a library with controlled and uncontrolled (or minimally-controlled) terminals, where uncontrolled terminals were only available to adults and controlled terminals were available to children. Uncontrolled terminals could access pretty much anything; controlled terminals could access sites already in the database with a ranking (within an category) above a level set by the local staff. Unlisted sites being accessed from the controlled terminals would be bounced to someone else for approval (I'm imaginging this going to someone supervising the children's area of the library). Alternatively, children's terminals could be in one of two modes: private but limited to approved sites or monitored but allowed pretty much uncontrolled access. That way if little Johnny is looking at porn/bomb howtos/drug recipes/whatever, someone shadowing the terminal can see it and either shut down access to the site or go have a talk with Johnny.
On the moderation side, scoring would be a bit more complex than /.'s - moderators could be ranked, and moderators with higher ranks would have more influence on the scoring. Similarly, moderators with low ranks or moderators who abused the system could be given less influence and eventually dropped. By doing this, you could hopefully prevent "poisoning" of the system by people with agendas. Ranking could be handled by something similar to /.'s meta-moderation, but probably more automated. If librarian A puts a site in a category and librarians B, C, and D all disagree with that assesment, A's ranking (overall or within that category) goes down. There could even be a way for the general public to "nominate" sites for moderation with suggestions as to categories.
The biggest headache with the whole thing, I think, would be who does the moderating, particularly early on - it's not like most libraries have people sitting around doing nothing who could just take it on as an additional duty. Unfortunately, it's not like there's a good base list out there that could be used as a seed.
One interesting thing is that if such a system was developed it could have commercial potential as well - perhaps a version of it that did not include the ability to moderate, but which did allow the use of frequently-updated copies of the score files from libraries. By providing it to libraries for free, a company would both help libraries avoid community pushes for censorware with an agenda and gain access to a pool of site raters.
Overall, there are all sorts of potential approaches to the problem of protecting people from disagreeable information, the problems are that there are so many different ideas of what's considered disagreeable and there are so many different ideas of who needs to be protected. On one end, you have people who believe that what's appropriate is no restrictions at all; on the other you have people who believe that all access should be tightly controlled; in between you have the bulk of the populace. In the US the general tendency is probably toward less control, but the tight-control people are noisy. Assuming that somewhere in the middle (limited control) is where things are going to end up, the problem becomes one of ensuring that the controls that go into place are not too restrictive, and that's where I think approaches like this could come in.
Finally, before people jump all over this as advocating censorware, I lean toward the less-control end personally, but I think that there are some things that need to be discussed with children before they're exposed to them unsupervised. Children who think and parents who teach them to do so are the way to go, but in our current society I'm not sure how much we can depend on that.
B: Why do you say this? Is the tolerance of the positioning tighter than 1/300th of an inch? (I don't have the spec handy...)
The tolerances are tight, but it'd be impossible to get them that tight. The reason consumer-printed checks will spit out is because the MICR lines aren't magnetic.
These days that may mean that they just dump into another system where they'll be OCRed, but a few years ago that wouldn't have been an option and it may not be now. While I know there are check reading systems out there that do imaging, the ones I know of are geared toward businesses rather than banks - I have no idea what the central banks use or how fast their technology changes.
This is important because check readers aren't actually doing any kind of OCR - when they started using MICR lines OCR was a dream even on a very primitive level. The numbers (and the 4 special characters) look that way because when you run them past a reader the waveforms / signals they produce are very identifiable. Think of them as being an early version of a magnetic stripe, but non-writable.
Position is also very important - individuals don't see it, but for any business printing its own checks in any significant quantity it's critical that the position of every character in the MICR line be correct. If the positioning isn't correct, the (often poorly adjusted) readers at the central processing banks will spit it out as a damaged check, and it gets special handling. Consumers never really see this, but for businesses their bank can actually be charged for the extra processing, and they're willing to pass that charge along to the business printing unscannable checks.
Any check printed by a consumer on a home PC is going to be spit out like this, but I'm not sure the banks are set up to deal with extra fees for those - they probably have the cost of a certain percentage of checks being "damaged" factored into their fees. This is also the reason that they prefer that you not fold checks - folded checks are both less likely to read and more likely to jam in the readers.
- If they settle, his opinion doesn't really matter.
- If they don't settle and he takes the DOJ recommendations, an appeals court may ship it back for consideration of other options.
- If they don't settle and he breaks MS into three companies, an appeals court may ship it back for consideration of other options - things like breaking it into only two companies.
- If they don't settle and he doesn't break MS up at all, maybe MS doesn't appeal.
Whatever happens, we'll all get to hear about it eventually....I've seen reports that it uses a stripped down version of Linux, however....
This could be something as limited as understanding the months, days of the week, numbers (possibly as high as the millions), years, AM/PM, morning/afternoon/evening/night, and maybe a few commands or other keywords like "with, alarm, minute(s), hour(s), todo, folder" and the like.
I even know how I'd use it - hit the record button, record a quick note or reminder ("Meet with Scott on Project X"), hold the record button, tell it where to store the recorded item: "Calendar May fifteenth at one pm with fifteen minute alarm" or maybe "Third Monday of every month at nine AM". Or maybe record "Listen to album by Artist X", then store it with "Folder music," or maybe "Todo purchase."
An awful lot of functionality could be managed with just a little bit of voice recognition and a small pre-defined dictionary of words. With the hardware available, enhanced dictionaries would be an obvious add-in, and as faster processors came along the number of distinct words that could be recognized would go up.
Overall I'd classify this as a win - if I was looking to purchase a new PDA right now I'd probably get a Palm Vx for its size, but I'd be tempted by some of the lighter CE handhelds for the recording capability. After all, I'm carrying a little digital voice recorder right now for notes when I can't write things down - should I carry two small pieces of electronics that rack up to about 8 ounces or one larger piece that weighs the same?
Figure that ISP XYZ in Sometown has 48 modems, and 200 subscribers, giving them roughly a 4:1 user to modem ratio - fairly good. However, ISP XYZ has had to invest [...]
I wouldn't be surprised if Netpliance hasn't invested much money at all in a dialup network - they're probably using resold service from UUNet, WorldCom, AT&T/IBM Global Network, or some other service. They'd pay a premium for it, but it avoids most of the up-front costs, scalability problems, and maintenance. It wouldn't even be all that difficult to build in remote-updating capability on the boxes so at some point they can start downloading new phone lists to people to switch to their own service.
As for the numbers tossed around in the previous article, there's a big difference between a $9.95/month ISP that's taking a loss and a $22/month ISP that might actually make a few dollars profit on each customer. Even if they're running their own dialup pools, they may be able to get away with a much higher ratio than 4:1, since they're not targeting the users who are likely to stay connected for long periods - why start a big download if you only have a few MB of flash to store it in and can't install new programs anyway? Basically, there are all sorts of ways to squeeze money out of a setup like this, including providing secondary services like shopping - Netpliance's stores would be the first place most of their users would probably go.
In terms of other savings, how much does your typical ISP spend on support? How much would it be if all customers were using only known hardware and software, didn't have significant download and local storage capacity, and didn't even have to dial in to get mail because it downloaded automagically 6 times a day?
In fact, they may have a single 800-number pool running right now for those mail checks; depending on the level of hardware control they have it should be possible to get message counts in a single call lasting less than a minute - don't bother with a data connection! Just call up, send a string of touch tones with some check digits to identify an account, and get a series of tones back indicating a message count. Much faster than negotiating a data connection.
Their market is probably a bit limited since I believe they'll be bringing them down in the Pacific, but for parts of Asia they might get some real interest!
Well, my old Toshiba Portege 300CT rus at 1024x600 on a 10.4" diagonal screen with no problems at all. People are even running Linux and X on it.
There have also been a few other small systems (mostly from Toshiba, I think) that run at unusual resolutions - Librettos are the most common of them. I think one (a small system with a camera built into the edge of the screen) may even use 1024x480.
Resolution has always been my biggest concern with the newer lightweight laptops - for a long time they were only available at 800x600, and I really wanted that extra width.
This was particularly significant because the items being sold were personalized, and there was a policy in place that the personalization could not include certain selected words. If a customer used words from the "banned" list, they got a polite email saying that their order couldn't be completed and giving a reason why. I suspect that a lot of nasty responses were averted because the email came from "Cheryl" or "Sue."
For the indignant, the reason for the banned list was that the company in question didn't want to have products out there that visibly included both their name and even mild profanity. They were concerned about tarnishing their image, and given their wholesome reputation I can understand why (though I thought and still think their list was a bit too restrictive).
It's a problem with a bundled software package that installs by default - how many Linux distros have been put together, then stayed on the shelves after someone found a hole or significant problem in a bundled package? Heck, how many of them have been sent to manufacturing then had something crop up after they started pressing discs and printing manuals?
Worse than that, unless you're only using the copier for "nefarious" purposes you'd probably have to solder in a socket instead so that you could alternate between the original chip (or a copy of it) and your hacked chip. After all, at some point there's a chance that an identifiable document from that machine will be scanned and the numbers correlated; if that happens at most company locations, when they ask "Well, who might know how to do this?" everyone will immediately think of one person. Even worse, some of these copiers may have a connector that techs use during maintenance to pull information - things like the ID, number of copies, error logs, all sorts of interesting* stats.
So, the old chip has to be removed and read, a socket has to be mounted (using surface-mount tools that most people don't have available), new chips have to be burned, and it all has to be done without attracting attention by having the copier down for a significant amount of time so nobody calls in the techs. Oh, and that socket? Best make sure that board isn't something that gets looked at when the service guys are in, they might just notice something odd...
Overall while it's probably possible to remove the serial number or corrupt it into unusability, it's probably not feasible to do so.
* Well, to some people.
Actually, what would probably work better is making a copy, then making a second-generation copy of it on either the same machine or another machine that uses the same ID method. That way there'd be some chance that the two sets of tags would interfere with each other. Of course, this would depend on what method was used to generate them and how they were placed on the output - it's conceivable that the methods used wouldn't be affected by this at all for some manufacturers.
As for the thought of using two different machines to generate the second-generation copy, that would help conceal which machine made the first-generation copy, but would probably still allow identification of the second-generation copy machine, which gives a location. If the second copy machine is going to be identifiable, why bother making a first-generation copy? For most situations there'd be no advantage to doing so - the primary times when it would be needed is if access to the original was limited, so a copy was needed for later distribution.
There are two reasons why the RIAA doesn't have problems with minidiscs:
If I'm remembering it correctly, minidisc copy protection works like this: there's a two-bit field on each disc that indicates what kind of disc it is. When you make a direct digital copy from a CD or another MD, that field is modified so that the copy can't be used as a source for another digital copy.
Of course, none of this keeps you from making analog copies of MDs, but because MD audio is actually compressed every additional generation adds additional "noise" to the copies. What this does is keep people from distributing first-generation copies of music - exactly what happens when someone rips a CD into MP3s and sends them out.
In fact, this is an improvement in some ways - if there's a basis for a countersuit in either Switzerland or one of the EU countries, it might make sense to check the laws in both. One item that might make waves (at least PR, if not legal) would be an EU-based suit that presented eToys as attempting to force a European artistic group out of existence using US laws. Depending on the local feeling about US domination of the Internet, that could turn into a serious impediment for eToys. The delay by the US trademark office probably wouldn't hurt this either, given the relative startup times of the two groups.
Note that I'm assuming here that eToys will manage to at the least stretch this out for a substantial period by virtue of having a carefully selected judge in a court local to them - if they can exert enough influence they may win the case despite the lack of legal merits, simply because it's difficult for etoy to reach the location to provide an active defense.
A countersuit of the sort I'm talking about would certainly have legal merit at least as strong as eToys' case, but the larger issue might be the PR question - if they manage to build a reputation in Europe as a bullying US megacorporation that beats up local artists it could seriously impact any plans they might have to expand their business beyond the US. If the etoy folks can appear to be charmingly eccentric in court this would be even better - "They may be kooky artists, but they're our kooky artists!" Charm from eToys' suited lawyers is unlikely to be an issue.
The PR issues might be just what was needed to get eToys to discuss settling, assuming that etoy is interested in doing so rather than in fighting it out to the bitter end. A settlement would probably end with etoy agreeing to have a relatively inoffensive front page that indicates that it is not associated with eToys (and probably with a link to the eToys site). If that sort of thing does happen, I personally think that etoy should push for it to be reciprocal - a note at the bottom of the eToys front page that says "This site is not affiliated with the etoy.CORPORATION art group." Such a note would, of course, have to include a link to www.etoy.com.
fencepost
just a little off
It seems to me that EToy, if they so desire, might be able to return the favor to Etoys - a suit, filed in the EU where they're based, because Etoys is diluting the recognizability of EToy.
After all, EToy is recognized as a source of cutting-edge art and social commentary, and it provably existed long before Etoys came along and picked up a similar name. Anyone who goes to the Etoys site seeking art is going to be just as confused as someone going to the EToy site looking for toys.
If the argument that the site is hosted in Switzerland and should be in the .ch top-level domain is allowed to stand, then a counter-argument that Etoys being hosted in California should be in the .ca.us domain is just as valid. In addition, the precedent that the case would set should bring massive support to EToy from other non-US companies with .com domains.
The movement from court to court seems remarkably like a case of shopping for a favorable judge, and for an international matter a California state court seems like a particularly inappropriate venue. The requirement that EToy be physically present in order to participate also seems onerous, particularly for a preliminary hearing - it would make sense for the trial itself, but not for the early stages.
As an interesting thought, can EToy go through discovery and depositions to investigate all records and email from the early days of Etoys in an attempt to determine whether the company was aware of the potential conflict? The depositions in particular would be better handled in a case in the EU - if the creators of EToy must travel to the the site of a US trial, it makes sense that the founders of Etoys be required to travel to the site of a Swiss one.
Even better, on thought: if in discovery it is found that Etoys was aware that they had no real legal standing, would that open them up to a barratry (abuse of lawsuits for intimidation purposes) suit?
If nothing else, they should be appealing for emergency permission to put up a page that describes the situation - "This site is down due to a lawsuit between Etoys, Inc. and the EToy artist group. To reach Etoys, the online toy seller, click {here}. To reach EToy, the artist group, click {here}."
Fencepost
just a little off