I'm picking a few nits here, but there are several chinks in DES's armor.
It has been known for some time that there are many weak keys in DES.
In 1990, Biham and Shamir published a differential cryptanalysis attack on reduced round variants of DES.
In 1993 Matthew Wiener published an initial design for a DES-breaking machine. At the time, it would have cost $2M.
In 1998 Biryukov and Kushilevitz published a ciphertext-only attack on reduced round DES reducing the workload by 2^20.
Near this time, I recall seeing a paper claiming to reduce DES to a 2^48 problem, but I'm unable to find a citation tonight.
In 1998, Wiener's machine was built by the Paul Kocher and EFF for about $250,000, and it breaks DES keys in about three days, on average.
So yes, I agree that DES is the granddaddy of Feistel network ciphers. Few of the cryptanalytic attacks work without ungodly amounts of chosen plaintext or artificially reduced round counts. But code breaks have generally occurred within months or years of implementation, not decades. As Gwido Langer, the chief of Poland's Biuro Szyfrow, said about breaking the German Enigma (when the British were unable to) "You don't have the same motivation as we do." Until after World War II, most code systems were broken during the same wars they were supposed to be protecting, and for that same reason.
The wartime and/or government codebreakers also have one more advantage: they don't typically announce their breaks to their enemies du jour. The recently released Venona papers show how Soviet spies who were given (faulty) one-time pads in 1942-1946 had them broken between 1948 and 1980.
Yes, it's a constant struggle. Yes, DES looks pretty good. But I wouldn't want to trust ALL of the national eggs to any single one of the currently commercially-available baskets.
The modern set is almost everything I wanted. I might still go dig up the old salamanders because they're still more distinctive than the///, but this is great!
I am fully aware that they are ship's wheels. I chose to mock them by calling them wagon wheels. (At the 16 x 16 resolution, they look the same.) It's still a busy, fugly, blobby mess. The salamander was distinct and clean. It was easy to identify.
They could still have the salamander wrapping around an envelope or a newspaper if they wanted separate icons for mail and news. They chose instead to use the Navigoat icons. They are ugly and dysfunctional, and I don't care if they stand for a ship's wheel, wagon wheel or a "no-llama parking" sign -- they don't work.
You forgot to mention that you can add certain sites to your Proxomitron's "white list" which will then be unmodified.
I love the Proxomitron. I can't believe the crap non-users put up with on their browsers -- pop ups, gratuitous flash advertisements, etc. My only complaint about the Proxomitron so far is that I'd like to be able to write a filter that would allow me to "whitelist" or "blacklist" a site via the right-click popup menu, or via a small toolbar at the page bottom. (I also use proxomitron as a gateway, and don't have it running on the computer where I have the browser.)
Ob Mozilla 1.2a comment: Type ahead find is, well, "interesting." It may take me a while to get used to it. But I have to say that I think the old find tool was kind of clunky to use, so anything will be an improvement. Other than that, I have seen no differences. It kept my old settings and plug ins. I can't tell if it's much faster or slower (this is a fairly fast box with plenty of free RAM.) And I don't use the mail and/or newsgroups, so they don't matter to me either.
Now, if they'd just go back to the old "salamander" icons instead of these heinous wagon wheels that are just more rectangular blobs on a screen overly crowded with rectangular blobs. They're too busy for fast, easy recognition, and have no inherent meaning. A salamander has no meaning, either, but it's a single color in an odd shape, making it very easy to spot.
I got my stats from this quote in the original article: "In a misguided effort to reduce expenses, an increasing number of sites are designed to work only in Internet Explorer, and sometimes only on the Windows platform, thus locking out 15% to 25% of their potential visitors and customers."
About obsolete, the article says this: "If Yahoo would simply replace its deprecated, bandwidth-gobbling font tags with bandwidth-friendly CSS, the cost of serving each page would greatly diminish". With this line, the article's author says "ignore those customers who are still using lynx, Mosaic, Netscape 2.0, et al." He then goes on to write the line I previously quoted saying that IE-only sites eliminate 15%-25% of the clients.
So he wants to drop those users who hang on to old technology (non-CSS browsers) AND he wants no one to write HTML that takes advantage of new browser features that only work in browsers he approves of. He provides examples of "bad" web sites, ignoring that Yahoo! is so bad that it's one of the most widely used portals ever and doesn't seem to offend anyone (except his own personal sense of style.)
I read it as a confusing set of conflicting opinions with no definitive point. So it was rather a lot like Slashdot, without the arguments, but with the trolls.
It's not even a review. The "sample chapter" presented features such nice conflicts as: web pages that are HTML 1.0 compliant waste bandwidth vs. web pages that are written for IE only turn away 25% of their viewers.
Near as I can figure out, he's claiming "the web is broken, don't bother."
Since his address is now blackholed anyway, maybe he should just start up a relay service, and charge spammers to use it?
Anyway, I think he should pick up the phone and call the dudes in Denmark. I think that being on an e-mail black hole list means never being ABLE to say you're sorry...
He described rotating the disc four times in order to see the reflections adequately when he scanned it. He then stitched those four images together.
This implies that the image reflected by the groove is not constant, even along one of those four pictures. So he got different "types" of pixels at different points on the groove. We can assume that some of those types would obscure the audio signal.
I would expect to hear a very regular rise/fall of interference over a period of 4 images : 33 1/3 RPM, (33 1/3 RPM == 1.8 seconds per revolution, yielding a period of.45 seconds or about 2Hz.)
There is probably another period at the edges of the stitch transitions (90 degrees out of phase with the "good" scans) where the increasingly bad groove images of one scan would start to be dithered to match with the decreasingly bad groove images of the next. The combined static beats would oscillate both before and after the.45 second dithering hump.
Strangely enough, when I listened to it again after writing the above, I hear the static rise and fall with about a half-second period. And the rising static has its own internal rise/fall beat, which may be those transitions around the stitching.
Either he chose his static quite cleverly, or he accomplished what he set out to accomplish. I would not be quick to judge him a fraud.
Mine are both RF. Heat-N-Glo has them, and I bought them with my fireplaces.
Damn, I just looked at that page. Some of their remotes are way cooler than mine! I want the remote fan speed and flame height! (I suppose I'd have to get new valves if I wanted remote flame height, though...)
Anyway, the better-than-convenience part is that they measure the ambient temperature right there on the remote and use that to turn the burner on/off. So if I set it to 78 degrees and put it on the table in front of the fireplace, it shuts the burners down at 78, and relights them at 75. The response time is quite slow (big mass of air, changing very slowly over time) so hysteresis is very noticable. That +0/-2 degrees easily becomes +5/-5 in the room.
But the biggest drawback is that the RF signal has no feedback, so the remote doesn't know if it was successful or not. And it doesn't keep trying even if it gets warmer. (The internal fireplace thermostat shuts it down if the firebox gets too hot, but I don't like relying on what should be a worst-case safety device.) Also, if you glance at the fireplace it may not be lit, but the remote can still be on (just too hot) so it could relight later as it cools down. So my advice would be twofold: keep fresh batteries in your remote, and CHECK TO MAKE SURE THE REMOTE IS OFF WHEN YOU'RE LEAVING.
The RIAA and MPAA want money. Lots of money. The kind of money they're used to. The P2P sharers want music. Lots of music. For free, just like they're used to.
Everybody keeps ranting "why don't they find a business model that works?" Here's your answer: There isn't one; there won't be one; there can never be one. First, it's an argument of corporations vs. the marketplace. Can you speak for every P2P user? Can anyone even claim to? Of course not, no one can. So it's already a one-sided discussion. The industries have no incentive to "talk" to the marketplace, since their only feedback comes in the form of "no revenue, no sales" in any case.
Jack and Hillary aren't stupid -- they've already figured that much out, so I think they've come up with a simple plan. They've decided to squeeze every last nickel from every last legitimate consumer until the whole production system implodes from lack of revenue. Their business plan is to get to be so rich now that they won't care when it implodes.
Under this plan, Jack and Hillary have no need to talk to anybody except to placate their respective industries. "Studios, crank out those movies. Recording companies, press those discs. We're taking good care of the whole Internet for you. We promise we'll have this piracy thing licked about the same time we reach $1,000,000,000 net worth (each.) So keep your stock prices up, please."
But if I were the RIAA, my legions of henchmen would be voting down the servers that supply "stolen" music, and voting up the servers that supply poison. And they would meta-mod down anyone who disagrees with their votes.
So to be useful, votes would require authentication in order to avoid ballot box stuffing. But authentication goes hand in glove with identification, and that's something the users of the P2P networks seem to be trying to avoid.
Bottom line: voting is subject to the same poisoning that the files are subject to. It adds a layer of complexity that simply delays poisoning, but probably not for long. Hell, with the inevitable bugs (that end up denying users unpoisoned files) and long-term ineffectiveness, voting would probably be smiled upon by the RIAA.
Frell the laptop users. The rest of us have to fly in that aluminum can, too.
The biggest threat to your health in the air is that of being hit by flying debris tossed about by a bit of random turbulence. And a 5 pound laptop is basically a large rock in those circumstances.
Check 'em all, I say. They can read a book or magazine, or even use <GASP>pen and paper</GASP> to get their last minute work done. (Of course, the latest Tom Clancy in hard cover flying about is almost as heavy, but it's probably not as aerodynamic as a laptop.)
Filmstrip media has been replaced by videotapes and DVDs, and LP and cassette tape recordings have been replaced by CD audio and MP3 recordings. Only the old physical media have died. The idea of using audio/visual media presentation of materials as a teaching tool remains. It was never and is still not a fad.
So now the Apple ][s and PC-XTs are in the dustbins with the filmstrips and cassettes. Keep in mind that the Apples replaced Teletypes. And they've since been replaced by Macs and Dells.
Don't make the classic mistake of confusing the medium with the message.
Computers have been in the classrooms since at least 1972 when I was in the 5th grade using them. Classroom computers may have been seen by some as a fad back then, but they were still taken seriously. Computers are not a fad now. Computers are a tool used by modern human society, and as such our society has decided to teach our children how to use them.
In high school, we spent some time at the start of our science classes learning how to use the (provided) calculators. Calculators in the classrooms were just as controversial a topic at that time, and for the same reasons you continue to suggest. Some of the people back then were also unable to recognize the difference between the use of a tool to learn vs. learning how to use that tool.
Schools have changed dramatically since I was a student. They try to keep up with modern society. But modern technology has outpaced our teachers' abilities to simultaneously keep up with every innovation and teach it. Kids don't have the responsibilties that keep them from learning the newest tech. That's the point of the original article. And this has been true since at least 1978, when I was denied my request to take our school's "computer class." The teacher knew me well from the science classes I had taken, and because I spent every day after school in the Teletype room. He acknowledged that I knew far more than he did, and that it would be a waste of my time to take his class. That was quite a shock to an 11th grader who had always been taught that teachers know more than the students.
In no way is any of this an argument against computers in the classroom. It's simply an observation of the problems involved in trying to use computers in school.
(Oh, and driver's ed is a required class in my son's 9th grade curriculum, and it's still a required course even if we choose to send him to a private driver's training school. Apparently, acknowledging the existence of automobiles is no longer a fad, either.)
I disagree, but I still think teachers need to take a different approach.
First, you probably agree that students need to know how to use computers. Claiming "computers in the classroom is an educational fad" is like claiming that telephones, televisions and automobiles are all passing fads. Computers are obviously here to stay. The real world bears this out as well. You almost can't get a job that doesn't involve computers at some level. Auto mechanics use computers extensively. Factory workers use data collection computers to control quality. You can't even ring up a sale at McDonalds without using a computer.
The way the real world works is: Here is your computer training. Now, using this computer like we just showed you, this is the job we want you to perform. How well prepared would an auto mechanic be if he walked into a job never having seen a computer, and the boss said "Great to have you, your work tickets are on that computer over there, the maintenance manuals are on that set of CD-ROMs, and you can order parts on this web page?"
So I see a need for some computer training in the schools. Sure, the kids need to learn their multiplication tables without them, (among a thousand other things,) but there comes a time when the computer becomes a part of most of these activities. Teaching an English class without having all the students use a word processor has become difficult for two reasons. First, there is the "haves vs. have-nots" gap, ensuring only your rich or privileged kids turn in pretty papers. Second, the act of writing using a word processor is fundamentally different than writing on a typewriter or with paper and note-cards. You cut and paste thoughts and shuffle them around dynamically, you don't type rough drafts once, hand correct them and retype them just once for a finished draft. Using the computer has made writing an iterative process. And this process has to be taught, and that requires computers.
So, should teachers use computers, and teach with them? Yes. Should every teacher have to hand out the basics of word processing to every student who walks in the door? No. That's where we get to the change needed. Schools need to make sure that "computer fundamentals" are taught early on so that
kids aren't left behind. (God, I hate to use that Dubyaism.)
The big problem is that computers leave both kids and teachers behind. "State of the art" hardware and software now changes hourly. (Six projects have shown up on Freshmeat in the last 60 minutes.) The schools have had to make do by purchasing a flock of computers, buying the current software of the day, and sticking with that for four years. They invest huge amounts of time and energy developing a curriculum based on that level of software. But the software world doesn't stand still, and the haves keep up at home while the have-nots stagnate on the school provided equipment.
This is a new problem. Change used to occur at a more humanly comprehensible pace. Keeping your math classes up-to-date with advancements in math happens at a fairly slow pace. Keeping up with the new software-of-the-day is a full time job for those of us in the software industry. Now, look at the teachers who have to be instant experts in front of classes full of students. Teachers have not yet learned how to simultaneously keep up with technology and still do their day job while retaining their sanity and their personal lives. That's a tall order for anyone. Nobody has a great answer yet. It's no wonder that students who have far more free time than their teachers are able to outpace them.
The whole point of the spyware stuff is that it's installed on the victim's computer. It reads the contents of the screens after they've been decrypted while they're being presented to the user.
I assume their product works by installing a global hook via SetWindowsHookEx(). They probably register to be notified of window messages pertaining to keyboard and drawing.
Sure enough, a google search of 'eblaster dll' turns up URLMKPL.DLL in the first hit. I'd like to dumpbin this DLL to see exactly what they call.
The point is that https: protects the links. It cannot protect the endpoints.
I agree, that made the radial menu unusable for me.
And uninstalling hasn't been possible. I remove some of the stuff from prefs.js, some of it from installed-chrome.txt, some from chrome.rdf, try deleting directories, and any random combination of these either doesn't remove the radial menus, or it keeps mozilla from functioning. I have not yet found the magic combo that actually just removes the radial menus. (I'm having troubles removing the Googlebar too, just in case you're the sort of person who wants to post instructions for me....:-)
Mozdev has some possibly great features, but I've been bitten by the inability to uninstall things often enough now that I'm gun-shy to try new ones. Mozilla needs a base-level uninstaller that can undo the damage done by any of these random add-ons. Mozilla should be able to keep track of all files added/modified by any of these add-ons, and be able to undo them at the click of a button. Asking the user to edit prefs.js, chrome.rdf, etc., etc., etc. is not end-user acceptable.
There are many people who don't know about popup-blockers. Joe and Jane SixPack, living in Farmtown, Minnesota, simply don't know anything different. "That's just the way it is, isn't it?" 500,000 usernames are subscribed to Slashdot. That leaves only 99,500,000 other internet users.
When Earthlink comes around and says "We promise no more pop-ups" this can actually awaken something within them that says "Hey, what a good idea. I'd pay for that." So they do.
Over 90% of the users have EVERYTHING default on their PCs.
I apologize, I didn't see your point this clearly in your first email.
And I understand your second point about the utility of having an agent learning your prefrerences. But that is of maximum value in the situation you describe: reading news. That's a situation where you don't care if you miss 1% of the valuable stories. That's not true of email; or even if it is for some specific people it probably isn't acceptable performance for someone trying to distribute something called a "spam filter".
I also understand your point about including headers vs. just the content, but I think it was a good choice on his part. I assume you've done some work on spam-fighting software. I have, and I can assure you that (at least a few years ago) some headers are truly spam-only-markers. He'd be passing up a great filtering chance if he didn't look at them.
Perhaps the software needs to go that extra step: rather than have a "this is spam button", it probably already has a "move to this folder" option. Why not create a probability array rather than a single spam probability assigned to each dictionary word? Tie the folder names to probabilities in the array.
Word Spam Inbox ChainLetters angels.2.0.8 opportunity.8.05.15 bertha.01.01.98 FW:.05.40.45
It could then transparently learn to move all my email, sending my Pilotgear mailings to my Pilotgear folder, etc. It would also reduce the "value" of automated senders as being only spam-related.
I still wonder how an agent will be able to discern the difference between a chain letter from Aunt Bertha and a "Hi, this is Aunt Bertha, meet my plane tomorrow please?" Every message from her gets a "Love, Aunt Bertha/get your free Juno account" tagline at the bottom. So, "love" "aunt" "bertha" all become words that are very strongly associated with ChainLetter-taint, when the real fingerprint probably should be the "Fw:" at the head of the subject line (as well as every line beginning with '>') Without at least one good letter from Bertha, her email will always end up in the ChainLetters bin. But with an array of sorting options, at least they wouldn't end up heading straight to/dev/null.
Perhaps other message-cumulative characteristics should be used in conjunction with word counts, such as message length, total count of exclamation points or dollar signs (or even of all individual ASCII characters,) grammar checker score, spell check score, etc. I think the overall concept of using a probability based mechanism rather than a score/threshhold mechanism is sound. I think we both agree that his approach needs more refinement.
Your filters are trained BY YOU to blacklist or whitelist words based on whether or not YOU decide if the message is spam. So if you mark Tcl-type messages as non-spam, then you'll get them.
It sounds about halfway to the "user agent" concept that all the futurists say will be the Next Big Thing. Think about applying the same filtering characteristics to reading articles from a mailing list, or your perusing of Slashdot. If you never read "YRO" articles and always read "spam" articles, this could be the mechanism your agent would use to float the anti-spam articles to the top of your screen, or the head of your inbasket.
Anyway, my point was they're YOUR filters, YOUR reading habits decide what's spam and what's not. And if you don't trust it, don't install it. You can spend all day reading all about organ-lengthening treatments, it's no skin off my nose (no pun intended.)
Spammers could start with the simple "leet" misspellings of their pitch words: 'Ea$iest way to get a j0b.' They could rotate these or generate them on the fly with a suitable mailer. 'Easies+ way to 6et a jo8' and 'Eas!est way to ge+ a j*b' are all variations that would pass the spam filters once (although they're not terribly effective.)
But the serious threat will come next in the form of the abomination that is Unicode. There are an infinite number of combinations of foreign letters that look 'Roman enough' that the casual user would have no trouble reading. The whole pitch would be crafted from randomly generated unicode look-alike letters. These words would then never appear twice in a dictionary.
A related problem bit me this very morning. I was debugging some printed text that someone had cut-n-pasted from a Word document into a field on their maintenance web page. Turns out Word had used the unicode character 0x2019 to represent an apostrophe, but the Microsoft-provided wcstombs() function choked on it, unable to translate it into a recognizable 8-bit printable equivalent.
So there will be ways around these filters. The question is now how long it will take for the spammers to start trying to beat them? I don't think they care about hitting every last hacker's inbox, but I do think they need to avoid ISP-level spam filtering.
I'd kept my address clean for many years, but I got bit because I wrote a letter to the editor of a scientific journal, who reposted it with my email address including a mailto: URL. God that hurt.
I had my first spam before I received my e-mailed copy of the journal. It was "related" to the topic of the journal, and said something like "I sure agree with what you wrote about in the journal. What's your opinion about http://my.url ?" But the To: line was the clue. It included not only me@myrealaddress.com but also that of smart.guy@nospam.address.com (another poster in the journal.) It was very apparent that the author had simply harvested the HTML and dropped it into his address database.
It was only hours before I was getting offers for detoxifying myself, HGH, climax gel and all-free teen pr0n.
Slashdot Mirror
So yes, I agree that DES is the granddaddy of Feistel network ciphers. Few of the cryptanalytic attacks work without ungodly amounts of chosen plaintext or artificially reduced round counts. But code breaks have generally occurred within months or years of implementation, not decades. As Gwido Langer, the chief of Poland's Biuro Szyfrow, said about breaking the German Enigma (when the British were unable to) "You don't have the same motivation as we do." Until after World War II, most code systems were broken during the same wars they were supposed to be protecting, and for that same reason.
The wartime and/or government codebreakers also have one more advantage: they don't typically announce their breaks to their enemies du jour. The recently released Venona papers show how Soviet spies who were given (faulty) one-time pads in 1942-1946 had them broken between 1948 and 1980.
Yes, it's a constant struggle. Yes, DES looks pretty good. But I wouldn't want to trust ALL of the national eggs to any single one of the currently commercially-available baskets.
The modern set is almost everything I wanted. I might still go dig up the old salamanders because they're still more distinctive than the ///, but this is great!
Thank you!
They could still have the salamander wrapping around an envelope or a newspaper if they wanted separate icons for mail and news. They chose instead to use the Navigoat icons. They are ugly and dysfunctional, and I don't care if they stand for a ship's wheel, wagon wheel or a "no-llama parking" sign -- they don't work.
I love the Proxomitron. I can't believe the crap non-users put up with on their browsers -- pop ups, gratuitous flash advertisements, etc. My only complaint about the Proxomitron so far is that I'd like to be able to write a filter that would allow me to "whitelist" or "blacklist" a site via the right-click popup menu, or via a small toolbar at the page bottom. (I also use proxomitron as a gateway, and don't have it running on the computer where I have the browser.)
Ob Mozilla 1.2a comment: Type ahead find is, well, "interesting." It may take me a while to get used to it. But I have to say that I think the old find tool was kind of clunky to use, so anything will be an improvement. Other than that, I have seen no differences. It kept my old settings and plug ins. I can't tell if it's much faster or slower (this is a fairly fast box with plenty of free RAM.) And I don't use the mail and/or newsgroups, so they don't matter to me either.
Now, if they'd just go back to the old "salamander" icons instead of these heinous wagon wheels that are just more rectangular blobs on a screen overly crowded with rectangular blobs. They're too busy for fast, easy recognition, and have no inherent meaning. A salamander has no meaning, either, but it's a single color in an odd shape, making it very easy to spot.
About obsolete, the article says this: "If Yahoo would simply replace its deprecated, bandwidth-gobbling font tags with bandwidth-friendly CSS, the cost of serving each page would greatly diminish". With this line, the article's author says "ignore those customers who are still using lynx, Mosaic, Netscape 2.0, et al." He then goes on to write the line I previously quoted saying that IE-only sites eliminate 15%-25% of the clients.
So he wants to drop those users who hang on to old technology (non-CSS browsers) AND he wants no one to write HTML that takes advantage of new browser features that only work in browsers he approves of. He provides examples of "bad" web sites, ignoring that Yahoo! is so bad that it's one of the most widely used portals ever and doesn't seem to offend anyone (except his own personal sense of style.)
I read it as a confusing set of conflicting opinions with no definitive point. So it was rather a lot like Slashdot, without the arguments, but with the trolls.
Near as I can figure out, he's claiming "the web is broken, don't bother."
The book looks broken. Don't bother.
Anyway, I think he should pick up the phone and call the dudes in Denmark. I think that being on an e-mail black hole list means never being ABLE to say you're sorry...
Plus you don't have to do ANYTHING geeky to put Linux on it. It already comes that way.
But it's shopping at Walmart... that's the part that's like having an AOL address.
Disclaimer: I NEVER shop at Walmart unless I need a gun in a hurry...
This implies that the image reflected by the groove is not constant, even along one of those four pictures. So he got different "types" of pixels at different points on the groove. We can assume that some of those types would obscure the audio signal.
I would expect to hear a very regular rise/fall of interference over a period of 4 images : 33 1/3 RPM, (33 1/3 RPM == 1.8 seconds per revolution, yielding a period of .45 seconds or about 2Hz.)
There is probably another period at the edges of the stitch transitions (90 degrees out of phase with the "good" scans) where the increasingly bad groove images of one scan would start to be dithered to match with the decreasingly bad groove images of the next. The combined static beats would oscillate both before and after the .45 second dithering hump.
Strangely enough, when I listened to it again after writing the above, I hear the static rise and fall with about a half-second period. And the rising static has its own internal rise/fall beat, which may be those transitions around the stitching.
Either he chose his static quite cleverly, or he accomplished what he set out to accomplish. I would not be quick to judge him a fraud.
Damn, I just looked at that page. Some of their remotes are way cooler than mine! I want the remote fan speed and flame height! (I suppose I'd have to get new valves if I wanted remote flame height, though...)
Anyway, the better-than-convenience part is that they measure the ambient temperature right there on the remote and use that to turn the burner on/off. So if I set it to 78 degrees and put it on the table in front of the fireplace, it shuts the burners down at 78, and relights them at 75. The response time is quite slow (big mass of air, changing very slowly over time) so hysteresis is very noticable. That +0/-2 degrees easily becomes +5/-5 in the room.
But the biggest drawback is that the RF signal has no feedback, so the remote doesn't know if it was successful or not. And it doesn't keep trying even if it gets warmer. (The internal fireplace thermostat shuts it down if the firebox gets too hot, but I don't like relying on what should be a worst-case safety device.) Also, if you glance at the fireplace it may not be lit, but the remote can still be on (just too hot) so it could relight later as it cools down. So my advice would be twofold: keep fresh batteries in your remote, and CHECK TO MAKE SURE THE REMOTE IS OFF WHEN YOU'RE LEAVING.
The RIAA and MPAA want money. Lots of money. The kind of money they're used to. The P2P sharers want music. Lots of music. For free, just like they're used to.
Everybody keeps ranting "why don't they find a business model that works?" Here's your answer: There isn't one; there won't be one; there can never be one. First, it's an argument of corporations vs. the marketplace. Can you speak for every P2P user? Can anyone even claim to? Of course not, no one can. So it's already a one-sided discussion. The industries have no incentive to "talk" to the marketplace, since their only feedback comes in the form of "no revenue, no sales" in any case.
Jack and Hillary aren't stupid -- they've already figured that much out, so I think they've come up with a simple plan. They've decided to squeeze every last nickel from every last legitimate consumer until the whole production system implodes from lack of revenue. Their business plan is to get to be so rich now that they won't care when it implodes.
Under this plan, Jack and Hillary have no need to talk to anybody except to placate their respective industries. "Studios, crank out those movies. Recording companies, press those discs. We're taking good care of the whole Internet for you. We promise we'll have this piracy thing licked about the same time we reach $1,000,000,000 net worth (each.) So keep your stock prices up, please."
So to be useful, votes would require authentication in order to avoid ballot box stuffing. But authentication goes hand in glove with identification, and that's something the users of the P2P networks seem to be trying to avoid.
Bottom line: voting is subject to the same poisoning that the files are subject to. It adds a layer of complexity that simply delays poisoning, but probably not for long. Hell, with the inevitable bugs (that end up denying users unpoisoned files) and long-term ineffectiveness, voting would probably be smiled upon by the RIAA.
The biggest threat to your health in the air is that of being hit by flying debris tossed about by a bit of random turbulence. And a 5 pound laptop is basically a large rock in those circumstances.
Check 'em all, I say. They can read a book or magazine, or even use <GASP>pen and paper</GASP> to get their last minute work done. (Of course, the latest Tom Clancy in hard cover flying about is almost as heavy, but it's probably not as aerodynamic as a laptop.)
And yes, I fly for business frequently.
So now the Apple ][s and PC-XTs are in the dustbins with the filmstrips and cassettes. Keep in mind that the Apples replaced Teletypes. And they've since been replaced by Macs and Dells.
Don't make the classic mistake of confusing the medium with the message.
Computers have been in the classrooms since at least 1972 when I was in the 5th grade using them. Classroom computers may have been seen by some as a fad back then, but they were still taken seriously. Computers are not a fad now. Computers are a tool used by modern human society, and as such our society has decided to teach our children how to use them.
In high school, we spent some time at the start of our science classes learning how to use the (provided) calculators. Calculators in the classrooms were just as controversial a topic at that time, and for the same reasons you continue to suggest. Some of the people back then were also unable to recognize the difference between the use of a tool to learn vs. learning how to use that tool.
Schools have changed dramatically since I was a student. They try to keep up with modern society. But modern technology has outpaced our teachers' abilities to simultaneously keep up with every innovation and teach it. Kids don't have the responsibilties that keep them from learning the newest tech. That's the point of the original article. And this has been true since at least 1978, when I was denied my request to take our school's "computer class." The teacher knew me well from the science classes I had taken, and because I spent every day after school in the Teletype room. He acknowledged that I knew far more than he did, and that it would be a waste of my time to take his class. That was quite a shock to an 11th grader who had always been taught that teachers know more than the students.
In no way is any of this an argument against computers in the classroom. It's simply an observation of the problems involved in trying to use computers in school.
(Oh, and driver's ed is a required class in my son's 9th grade curriculum, and it's still a required course even if we choose to send him to a private driver's training school. Apparently, acknowledging the existence of automobiles is no longer a fad, either.)
First, you probably agree that students need to know how to use computers. Claiming "computers in the classroom is an educational fad" is like claiming that telephones, televisions and automobiles are all passing fads. Computers are obviously here to stay. The real world bears this out as well. You almost can't get a job that doesn't involve computers at some level. Auto mechanics use computers extensively. Factory workers use data collection computers to control quality. You can't even ring up a sale at McDonalds without using a computer.
The way the real world works is: Here is your computer training. Now, using this computer like we just showed you, this is the job we want you to perform. How well prepared would an auto mechanic be if he walked into a job never having seen a computer, and the boss said "Great to have you, your work tickets are on that computer over there, the maintenance manuals are on that set of CD-ROMs, and you can order parts on this web page?"
So I see a need for some computer training in the schools. Sure, the kids need to learn their multiplication tables without them, (among a thousand other things,) but there comes a time when the computer becomes a part of most of these activities. Teaching an English class without having all the students use a word processor has become difficult for two reasons. First, there is the "haves vs. have-nots" gap, ensuring only your rich or privileged kids turn in pretty papers. Second, the act of writing using a word processor is fundamentally different than writing on a typewriter or with paper and note-cards. You cut and paste thoughts and shuffle them around dynamically, you don't type rough drafts once, hand correct them and retype them just once for a finished draft. Using the computer has made writing an iterative process. And this process has to be taught, and that requires computers.
So, should teachers use computers, and teach with them? Yes. Should every teacher have to hand out the basics of word processing to every student who walks in the door? No. That's where we get to the change needed. Schools need to make sure that "computer fundamentals" are taught early on so that kids aren't left behind. (God, I hate to use that Dubyaism.)
The big problem is that computers leave both kids and teachers behind. "State of the art" hardware and software now changes hourly. (Six projects have shown up on Freshmeat in the last 60 minutes.) The schools have had to make do by purchasing a flock of computers, buying the current software of the day, and sticking with that for four years. They invest huge amounts of time and energy developing a curriculum based on that level of software. But the software world doesn't stand still, and the haves keep up at home while the have-nots stagnate on the school provided equipment.
This is a new problem. Change used to occur at a more humanly comprehensible pace. Keeping your math classes up-to-date with advancements in math happens at a fairly slow pace. Keeping up with the new software-of-the-day is a full time job for those of us in the software industry. Now, look at the teachers who have to be instant experts in front of classes full of students. Teachers have not yet learned how to simultaneously keep up with technology and still do their day job while retaining their sanity and their personal lives. That's a tall order for anyone. Nobody has a great answer yet. It's no wonder that students who have far more free time than their teachers are able to outpace them.
I assume their product works by installing a global hook via SetWindowsHookEx(). They probably register to be notified of window messages pertaining to keyboard and drawing.
Sure enough, a google search of 'eblaster dll' turns up URLMKPL.DLL in the first hit. I'd like to dumpbin this DLL to see exactly what they call.
The point is that https: protects the links. It cannot protect the endpoints.
Professional DAT recorders (every studio would have one) don't respect the SCMS flag.
And uninstalling hasn't been possible. I remove some of the stuff from prefs.js, some of it from installed-chrome.txt, some from chrome.rdf, try deleting directories, and any random combination of these either doesn't remove the radial menus, or it keeps mozilla from functioning. I have not yet found the magic combo that actually just removes the radial menus. (I'm having troubles removing the Googlebar too, just in case you're the sort of person who wants to post instructions for me.... :-)
Mozdev has some possibly great features, but I've been bitten by the inability to uninstall things often enough now that I'm gun-shy to try new ones. Mozilla needs a base-level uninstaller that can undo the damage done by any of these random add-ons. Mozilla should be able to keep track of all files added/modified by any of these add-ons, and be able to undo them at the click of a button. Asking the user to edit prefs.js, chrome.rdf, etc., etc., etc. is not end-user acceptable.
There are many people who don't know about popup-blockers. Joe and Jane SixPack, living in Farmtown, Minnesota, simply don't know anything different. "That's just the way it is, isn't it?" 500,000 usernames are subscribed to Slashdot. That leaves only 99,500,000 other internet users.
When Earthlink comes around and says "We promise no more pop-ups" this can actually awaken something within them that says "Hey, what a good idea. I'd pay for that." So they do.
Over 90% of the users have EVERYTHING default on their PCs.
And I understand your second point about the utility of having an agent learning your prefrerences. But that is of maximum value in the situation you describe: reading news. That's a situation where you don't care if you miss 1% of the valuable stories. That's not true of email; or even if it is for some specific people it probably isn't acceptable performance for someone trying to distribute something called a "spam filter".
I also understand your point about including headers vs. just the content, but I think it was a good choice on his part. I assume you've done some work on spam-fighting software. I have, and I can assure you that (at least a few years ago) some headers are truly spam-only-markers. He'd be passing up a great filtering chance if he didn't look at them.
Perhaps the software needs to go that extra step: rather than have a "this is spam button", it probably already has a "move to this folder" option. Why not create a probability array rather than a single spam probability assigned to each dictionary word? Tie the folder names to probabilities in the array.
It could then transparently learn to move all my email, sending my Pilotgear mailings to my Pilotgear folder, etc. It would also reduce the "value" of automated senders as being only spam-related.
I still wonder how an agent will be able to discern the difference between a chain letter from Aunt Bertha and a "Hi, this is Aunt Bertha, meet my plane tomorrow please?" Every message from her gets a "Love, Aunt Bertha/get your free Juno account" tagline at the bottom. So, "love" "aunt" "bertha" all become words that are very strongly associated with ChainLetter-taint, when the real fingerprint probably should be the "Fw:" at the head of the subject line (as well as every line beginning with '>') Without at least one good letter from Bertha, her email will always end up in the ChainLetters bin. But with an array of sorting options, at least they wouldn't end up heading straight to /dev/null.
Perhaps other message-cumulative characteristics should be used in conjunction with word counts, such as message length, total count of exclamation points or dollar signs (or even of all individual ASCII characters,) grammar checker score, spell check score, etc. I think the overall concept of using a probability based mechanism rather than a score/threshhold mechanism is sound. I think we both agree that his approach needs more refinement.
Your filters are trained BY YOU to blacklist or whitelist words based on whether or not YOU decide if the message is spam. So if you mark Tcl-type messages as non-spam, then you'll get them.
It sounds about halfway to the "user agent" concept that all the futurists say will be the Next Big Thing. Think about applying the same filtering characteristics to reading articles from a mailing list, or your perusing of Slashdot. If you never read "YRO" articles and always read "spam" articles, this could be the mechanism your agent would use to float the anti-spam articles to the top of your screen, or the head of your inbasket.
Anyway, my point was they're YOUR filters, YOUR reading habits decide what's spam and what's not. And if you don't trust it, don't install it. You can spend all day reading all about organ-lengthening treatments, it's no skin off my nose (no pun intended.)
The only problem is that too many people would want hotmail, aol and msn to be in the "trusted" list. And we all know that can't be.
Still a great idea.
I haven't been hurting for spam samples recently... :-(
Spammers could start with the simple "leet" misspellings of their pitch words: 'Ea$iest way to get a j0b.' They could rotate these or generate them on the fly with a suitable mailer. 'Easies+ way to 6et a jo8' and 'Eas!est way to ge+ a j*b' are all variations that would pass the spam filters once (although they're not terribly effective.)
But the serious threat will come next in the form of the abomination that is Unicode. There are an infinite number of combinations of foreign letters that look 'Roman enough' that the casual user would have no trouble reading. The whole pitch would be crafted from randomly generated unicode look-alike letters. These words would then never appear twice in a dictionary.
A related problem bit me this very morning. I was debugging some printed text that someone had cut-n-pasted from a Word document into a field on their maintenance web page. Turns out Word had used the unicode character 0x2019 to represent an apostrophe, but the Microsoft-provided wcstombs() function choked on it, unable to translate it into a recognizable 8-bit printable equivalent.
So there will be ways around these filters. The question is now how long it will take for the spammers to start trying to beat them? I don't think they care about hitting every last hacker's inbox, but I do think they need to avoid ISP-level spam filtering.
I had my first spam before I received my e-mailed copy of the journal. It was "related" to the topic of the journal, and said something like "I sure agree with what you wrote about in the journal. What's your opinion about http://my.url ?" But the To: line was the clue. It included not only me@myrealaddress.com but also that of smart.guy@nospam.address.com (another poster in the journal.) It was very apparent that the author had simply harvested the HTML and dropped it into his address database.
It was only hours before I was getting offers for detoxifying myself, HGH, climax gel and all-free teen pr0n.