***Wow... I have never, ever seen a software product that wasn't working on QA bug reports right up to the minute the gold disc is burned.***
Dead On, Mate. A software project with three quarters of a million lines of code is surely going to have hundreds of open SPRs at the time of its release. If things are going well, most of them are going to be relatively unimportant. Still, You have to wonder if the FBI's CIO knows much about real software projects and how they work.
I prowled around the Internet trying to find any sort of bio on Azmi. Here's what I came up with http://www.federalnewsradio.com/?nid=5&sid=760419. His background isn't awful, but neither is it especially reassuring. There's nothing there that says to me that this isn't a guy with an expensive haircut and no proven technical or administrative ability. But he could also be fighting the good fight against heavy odds. No way to tell I think. At least he doesn't seem to be a political appointee.
Nothing against Mr Azmi, but apparently if has been common knowledge that VCF was a fiasco since early 2005. See http://www.smh.com.au/news/Breaking/New-FBI-softwa re-may-be-unusable/2005/01/14/1105582686258.html I'm sure that the problems are not Azmi's fault. Given the timeline, I don't see how they could be. But he's had two and a half years to recognize that there is a problem and get the project back on track. I have to wonder if he is the right guy to try to fix this mess.
Seems to me like this is largely a list of influential software, not a list of Great software.
There was, for example, nothing much great about OS/360 which was, derivative, far buggier than Windows has ever thought about being, and had the most wrong headed, obtuse and god awful scripting language ever designed. It did however provide a barely adequate vehicle for selling IBM's basically OK hardware. Great? Hell no. Influential? Yes, Very.
So, OS360, Excel, etc are influential rather than great. What would be the list of actually great software? In no particular order:
Surely, the first higher order language? That would be Fortran, Cobol, Algol or something along that line?
The first interpreted language? BASIC is probably the most influential, was it the first?
The Xerox PARC Alto software which was the starting point for the Xerox Star, Lisa and eventually the Mac and Windows. I'd probably give the Macintosh the award for most influential (and I'm not a Mac guy).
Visi-Calc. Lotus and Excel were more influential, but Visi-Calc was the breakthru.
Unix
The first simple OS for small computers, I'm not sure what that'd be. Something at DEC (RSX-11?) begat CPM begat QDOS begat MSDOS. But I'm sure the chain goes back further.
The first file system. I have no idea where the idea of file systems originated, but modern users have no idea what a pain reading and writing a disk was using manual storage space allocation. There have been times as recently as the mid 1990s when I missed punched cards and some other outmoded technologies, but I've never missed manual disk space allocation.
The World Wide Web and browsers therefore. Mosaic? Something else? Whatever..
Fidonet and Compuserve -- the programs that allowed folks sitting at home with dial up modems to talk to more or less randomly selected users around the world via stored text messages.
SAGE -- An air defense system designed on the 1950s that ran on vacuum tube computers and actually worked. It supported several dozen consoles, recieved data from a dozen or more radar sites continuously, vectored manned and unmanned interceptors, talked to adjacent control centers. talked to anti-aircraft batteries, and did some other things. All using a computer roughly comperable in capacity to an IBM PC with 256KILObyes of memory. (Many subsequent attempts to replicate many of its capabilities using more modern hardware and software failed rather dismally). Great? I think so. Influential? Not very -- perhaps unfortunately.
...
There are plenty of others, but typing this isn't getting the lawn mowed.
***Theres no bloody way you can tell me that you need more than a 10 fold markup to cover the research!***
The drug companies need to recover enough money to support all their research -- including trials of the many compounds that just don't work out. And they also need to self-insure against liability should one of their products kill or maim a bunch of folks even after all the testing. Of course, the companies also make obscene profits and it wouldn't bother me or a lot other people much, if that particular brand of obscenity were substantially moderated.
But the system -- bad though it is -- works after a fashion and this is a case (and the only one I can think of) where simply closing all the world's patent offices, voiding all existing patents, and retraining all the patent lawyers as garbage collectors will not make the world a better place. Yes, we pay way too much for drugs until the patents expire. But in this one case, the patent-mongers are probably right. No patents, no new non-recreational drugs.
***Well, kinda, yeah. Can you think of any applications for which 10 digits of precision isn't enough? ***
Basically, I agree with you, but as Hamming pointed out in the 1950s you can get yourself into trouble with some thing like:
A = small number, e.g. size of smallest feature in a Celeron-M CPU in microns
B = big number, e.g. distance to Andromeda galaxy in microns
C = (A+B)... (some set of clever operations)... - B
The math is fine, but the implementation won't yield the correct answer. because the value of interest got scaled off. That isn't rocket science, but it is suprisingly hard to catch.
That said, Floating Point works suprisingly well for most things. Back in the 1960s, things like 2+2=3.99999 were a fairly common phenomena. Hardly ever happens today. I spent about three decades in the defense industry writing and testing systems that got things from point X to point Y. I think I saw a reasonable selection of the possible errors that folks can make. Other than subtracting big numbers from other big numbers and expecting the result to be precise, I can't recall a single problem with floating point.
If we're looking for something that is basically broken to write an article about, forget floating point. Try event driven software architectures. Or the notion that it is possible to write unambiguous specifications that are also comprehensible.
***Google ajaxified their home page a little, moving Groups to an web 2.0-ey submenu that takes me 1 extra click to get to, and replacing it with the ridiculous Video web-2.0 ey thing.***
I'm not sure of the cosmic significance, but if your browser doesn't support Ajax, Google presents the submenu on the main page where you (and I) want it. Perhaps this is a subtle warning from God that she disapproves of AJAX and will eventually smite those who embrace it mightily.
Jeez, the popular kids are having this big patch party, and we Windows 9 and Linux users aren't invited... again.
Anybody want to join me at home for some popcorn and maybe some orange soda while the NT crowd spends the night fixin stuff? I know it's dull, but the most of the computers around here are just too slow to run XP, so I'll just have to pass on the excitement.
Actually, we do have a couple of XP machines, but they are such a pain, that I really don't much care if some sociopathic teenager in Bucharest 'owns' them. Owning a Windows machine is exactly what hackers deserve -- at least until it's technically feasible to send them a case of jock itch or crab lice.
I must say that depending on what the worm does, it really might be 'that' serious. Trouble is that from what I read networks of real computers doing real work for real users are likely to be in considerable difficulty the first time someone plugs an RJ45 into an infected laptop and hits the power switch. I'm glad that I'm retired.
***Then again I could cite my old favorite, the 6809.***
The 6809 was not only easy and fun to program, 6809 programs tended to benchmark out significantly faster than programs for comperable CPUs like the Z80, 6800 and 8080. If the industry ever decides to scrap the x86 mess -- which they won't -- going back to the 6809 for a starting point might not be a bad idea at all. I once did a plot of measured times for a benchmark where timings were available for a bunch of CPUs (Sieve of Eratosthenes). When you plotted out clockspeed vs word width, all the CPUs from the 8080 to the Cray something or other fell out into an untidy straight line, except for the 6809. There were, as I recall, three different results published for SOE on the 6809 and all three were an order of magnitude faster than they had any reasonable expectation of being based on the hardware's apparent capabilities.
***I wonder why kids have internet access at school. Do someone really want them to have ADHD since childhood? Aren't they supposed to learn something while they sit in waiting to be online back home?***
The theory is that the Internet is supposed to make all manner of information available to students that would otherwise be difficult for them to access. Every student has access to the national press, encyclopedias, the latest scientific papers. And they do. The New York Times may have its problems, but it has better and broader news coverage than the Jimsomweed Junction, TX Weekly Courier or the truly awful reporting on the US TV networks. Wikipedia may not be authorative (whatever the hell that is supposed to mean), but it has generally adequate articles on a broad range of subjects -- broader than Brittanica I'm pretty sure. Live in East Bywater, Vermont and want to learn Italian or Mongolian? The Internet really can help you. In the US, school intenet access is heavily subsidized exactly because it makes information so readily available.
The Internet makes information available. However, it turns out that a fair percentage of the information that kids are genuinely interested in isn't exactly what educators, parents, and community leaders had in mind. Moreover it is amazingly difficult to tell which information is which. I suspect that most educators would (and do) look the other way with regard to a bit of pornography, and the odd illegal music download. Problem is that there is so much unfortunate stuff out there. For example, kids are perfectly capable of posting a little amateur photography done in the girl's locker room with a cell phone. That's not really all that bad, but something does have to be done about it. What about the anarchist cookbook? I think it's likely some of the recipies there are capable of distributing pieces of a classroom and the experimenter over maybe half an acre. What about white supremicist sites? Or black supremicist sites? What if some kids put up a "Why we hate Charlie Smith site"? What if Charlie sees the site and decides to hang himself? Things like that could happen. A thirteen year old in our town made a suicide pact over the Internet and carried it through. Fortunately from the school system's point of view, the kid used his home PC, not the school's. But still... aside from the guilt thing, that's a potential multimillion dollar lawsuit.
Also remember, that in a rural area, the school is likely to have the only high speed data line in town.
Consider the costs of blocking malware downloads -- which as we all know is easier said than done. And the cost of removing the stuff that gets downloaded anyway. Some of it can take hours or days to exorcise.
So what's the answer? What most schools use are rather expensive filters whose most obvious characteristic is that they don't work all that well. Personally, I suspect that the eventual answer will be whitelist access to approved sites for most students most of the time with a few machines given full access for staff, administration, and supervised student access. But first, it seems that school administators have to recognize that they have a problem that won't go away, and school IT folks have to admit that technology can't fix their problems at reasonable cost. That'll take a while.
No, I'm not kidding. Exactly how many eMails have you sent Leahy explaining what he is doing wrong? I'm suggesting that you might try it instead of wasting simply writing the guy off as hopeless -- which he possibly is not. You may not get positive results, but the results are unlikely to be worse than those from your present strategy which can't possibly -- as far as I can see -- produce positive results.
Do me a favor. Think about it for a while.
I'll probably send Leahy an eMail myself but I'm a little at a loss what to say. My personal opinion is that the entirety of Intellectual Property law is rotten to the core and should be pitched out -- with the exception of trademark and maybe a 15 year copyright. But I think that mainstream society is two or three decades of stupidity by "Intellectual Property owners" away from being able to deal with a view that extreme. So what should a reasonable patent law for 2006AD look like?
***.. and that's why the halls of IBM Essex are empty (practically), IBM jobs are flowing out of the state under one guise or another. (thank you circ. highway!).***
The BLS says total employment in Vermont is up 12.2% over the past 10 years. Population is only up 5.7% 1995-2005. So actually, jobs would seem to be moving to Vermont. They aren't necessarily great jobs of course
IBM Essex Junction is hiring. Has been for quite some time.
***Voter turnout in Vermont is no better than any other state...***
Don't know where you got that idea. Voter turnout in Vermont is consistently one of the highest percentages in the country at 60+% in the general elections. Turnout was 70.8% in November 2004. The Burlington Free Press had an article a few years ago pointing out the state turnout percentage hasn't changed much since the 1920s, and unlike most states Vermont voter turnout didn't fall off significantly in the 70s, 80s and 90s. It is true that Town Meeting turnout has been falling off -- presumably because more and more people are working 8-5 jobs instead of farming.
***That being said...It's no better here in Nebr.. (7th highest tax burden for one of the smaller populations).***
That's curious. I've driven across the plains a few times in recent years and I was struck by how prosperous the small towns in Nebraska looked compared to Kansas and Oklahoma. Pretty superficial, but maybe you're doing better than you think.
Orrin Hatch is a national embarassment, and Ted Stevens is clearly somewhat dumber than dirt, but Pat Leahy is a pretty decent guy. Anyone who can piss Dick Cheney off can't be all bad.
Leahy's major constituent who has a dog in this fight is IBM. (IBM is the largest employer in Vermont -- by a lot). I'm not sure what IBM's concerns are, but as long as they are taken care of, I doubt Leahy much cares about anything other than doing the right thing. Leahy probably doesn't need large contributions to get reelected assuming that he runs again (He'll be 70 in 2010). To get a feel for Vermont politics, google 'Leahy Fred Tuttle vermont' for the story of the 1998 senate campaign that ended up with the Republican candidate asking people to vote for Leahy because Tuttle's wife didn't want to move to Washington DC. (Incidentally, by the end of the campaign, a lot of Vermonters thought Tuttle might make a pretty good senator)
Vermont politicians listen better than most American politicians. They have to. If they don't make sure that their mail is read, calls are returned and don't spend a lot of their time ouside of Washington in Grange halls and school auditoriums in tiny towns, they don't get re-elected.
So, there is a chance that well written, well reasoned, thoughtful letters, might actually cause Leahy to make changes or even withdraw his sponsorship of the bill. OK, folks what do you want to tell Leahy?
I have to agree with you that the clowns who systematically destroyed the Compuserve Forums deserve to be pounding the bricks. However it sounds to me like they are going to be firing the AOL workers, not the AOL management team. Maybe hold off on the Champagne until you read an announcement that the 3000 top people at AOL were refused entry to the facilities this morning, and were given 30 minutes to clear out their desks. That'll be the time for celebrating.
Don't drink much anymore, but I'll make an exception for that.
Good Question: JAXA appears to be an English Acronym (for Japan Aviation eXploration Agency. I doubt that Japanese attempt to pronounce it as Japanese doesn't have an X character. I'd guess they might do something like ja-ka-tsu-a if they really have to pronounce it.
There's probably a Japanese name for the Agency as well, but Verizon has our DSL running at sub dial up speed today, and I don't have the patience to dig the Japanese name out.
Anyway, I reckon you can pronounce JAXA any way you want to.
Based on many years of watching government and private security in action.
A very important issue is that secrecy is expensive. It is tempting to try to protect a great deal of your data. If you do that, you will spend a great deal of money. And you will impede the work of your people who too often will not have ready access to stuff they actually need to know. And -- pardoxically -- you may reduce the protection given to data that actually needs to be secured because people will develop ways to bypass security so that they can get their job done.
IMO, many companies shouldn't even consider securing any of their data. Those that do, possibly shouldn't put their critical data on computers -- expecially not on networked computers. If all your important data can fit on 200 index cards, put it on 200 index cards and lock them up. What, for example, would be the point in putting the secret formula for Cudweiser beer (12 parts cow urine, 37 parts water, 1 part used motor oil,...etc) on a computer where anyone can steal it?
Companies that deal in personal data e.g. schools, hospitals, etc. Here we are getting beyond my experience. I suspect that mixing sensitive data with accounts payable, staff eMail, memos about the Christmas Party, etc is not a good idea. But beyond that, I just don't know.
I don't see javascript vulnerabilites as being limited to Windows or IE. Why would they be? In fact, attacks on routers and such if they happen are probably going to be attacks on specific unix configurations.
I think the real concern here is facility networks more than home users. If the facility users are allowed to access the Internet, and javascript is permitted by the browsers, there's an open pipeline right past whatever front end filtering is in place. I doubt it will be very practical to scan every block of javascript coming into the building for malicious code.
And how about Javascript embedded in that mother of all bad ideas -- HTML eMail?
***but failed to find my wireless router (through which it had to pass in order to see the rest of my network), or my print server. It also identified as "exists" several IP addresses on which no machine or device exists.***
Doesn't the second part of that make you a little nervous? One possibility is that it is finding your router and print server, but not where they are supposed to be. Could be an error in the program, but it could be some 'feature' of your network environment that you'd like to know about.
***Grossman says 'The users really are at the mercy of the Web sites they visit. Users could turn off JavaScript, which really isn't a solution because so many Web sites rely on it.'" ***
Look folks, this isn't rocket science. Given the current state of Computer Technology, downloading obscure programs from a remote source outside your control and running them can't possibly be a good idea. It may occasionally be necessary, but it's something that should be done as rarely as possible. If you don't even know the programs are there because they are buried in web pages, that just exacerbates the problem.
The answer: Turn off Javascript, and let the web site designers find some other way to entertain themselves. Forcing web sites to use HTML and server side scripting may limit their style and the coolness of your user experience, but if you want your computer(s) and network to be somewhat secure then you better forget Javascript. Personally, I have only one Javascript enabled browser (Firefox) and I try to use it only with sites like Google and a handful of others that I consider to be trustworthy.
FCRC = Federal Contract Research Center. These are a bunch of not for profit (or at least they used to be) outfits that do research for the military, civilian agencies and a lot of the three letter agencies that quietly spend a lot of your tax money. You've probably heard of some FCRCs -- RAND Corporation, FermiLab, etc. They aren't paradise and you may someday face a choice of working on something you disapprove of or quitting. But they are not -- for the most part -- run by the spectacularly incompetent right wing flakes who are currently doing their best to destroy American science and competetiveness while claiming to be doing the opposite.
Actually, there aren't that many schools that provide laptops -- especially not in lower grades. There are a bunch of problems that no one outside the schools considers -- computer left at home, computer stolen, other kids broke the computer, and especially... computer lost (if kids manage to lose their jackets and gloves when it's -20F -- and they do -- you can imagine what'll happen to laptops). It's not that the problems can't be overcome, but it's a lot of work. It's not clear what the payoff is.
As for studies of computers. Lot's of them. Mixed results, but generally mildly negative. Costs of developing and maintaining curricula are high as are support costs for the hardware. And don't get me started on Windows(especially XP), Office, internet filtering, viruses, malware, etc, etc, etc. Google "River Oaks Ontario" for an experiment in depth in computing in schools with mixed results. Some educational positives, but support costs were reported to be unsustainably high once Apple stopped underwriting them.
Computers have some unquestioned utility in school administration and class materials preparation. Some other uses in schools in the US are probably beneficial. Computers will surely someday be a major element in education. But not in K-8 in the US today. I have no idea how things change in an undeveloped country. Maybe computers would be a blessing. Or maybe the wisest thing that India, Paraguay and Chad could do would be to confiscate anything that looks like a computer at the school's front door.
***yes but your car doesn't decide to drive through 3 red lights while your passed-out at the reins either; the vast majority of browsers are reasonable consistant, then the the horse, IE, which decided to wonder off on its own; which means the vast majority of sites should look fine if designed for standards and then later for IE***
I think that many web design issues are self inflicted by trying to force HTML into specific presentation layouts. It really wasn't designed to do that. If IE trashes simple markup, that's one thing. If it just looks a bit different than Opera/Netscape, maybe the problem is unreasonable expectations. I'm not sure how much complaint one is entitled to if they stagger out of the pub, climb into someone else's wagon, and the horse takes them to the wrong house.
I understand that everyone seems to want a language that allows presentation layout to be specified. That's a perfectly reasonable desire. Probably such a language should be written. Maybe the authors of IE should have respected the desire to control presentation and tried to minimize differences in results from pre-existing browsers like Netscape. Hell, maybe they did. It's really hard to anticipate or identify all the peculiarities of browsers. For example, one popular browser -- I've forgotten which, but it wasn't IE, seemed to think that (33+33+33)>100, and therefore displayed three images of 33% width on two lines. It put them in one line if they were 32% width. (For all I know there is a good reason for that). Expecting the folks who wrote IE to know hundreds or thousands of non-obvious behaviors and emulate them in their browser may not be realistic.
Maybe what all this demonstrates is that Dvorak has a point for once. This stuff really doesn't work all that well.
***HTML WAS DESIGNED to MARKUP a DOCUMENT to SPECIFY a LAYOUT!***
That's true only if you think a chocolate mousse must, of necessity, be a large herbivore.
Yes, HTML is largely specific to a document layout -- a sort of ornamented outline without paragraph numbering. It is not specific to a presentation layout. Confusing the two kinds of layout will generally cause anyone who does it substantial grief. What I See Is very likely Not What You Will Get (WISINWYG?) -- by intent.
***With all do [sic] respect, I shouldn't have to be a "decent" web designer to be able to put up a personal homepage that looks the same in all browsers....***
At the risk of pointing out the obvious, HTML was intended to be a MARKUP language, not a LAYOUT language. Why would anyone who understands the meaning of those two words be suprised that dictating LAYOUT in HTML is next to impossible? It's sort of like being constantly suprised that your car, unlike your horse, can not find its way home from the village pub after you have drunk yourself into a stupor.
Since most of the world seems to want a layout language for the web, I reckon someone ought to design one. But my guess is that HTLL (for want of a better name) it is going to be a bundle of grief as people attempt to cope rationally with differing screen sizes, color depths, font availabilities etc.
By all means, create your HTLL. But if it is not too much to ask, how about leaving HTML as a markup language where the browsers have reasonable latitude in presentation? Really, almost always, all I want to do is see the damn content, and all this garbage intended to dictate layout seems to trash the content at least as often as it does in forcing the layout that the page author intended.
***C might have been a reaction to Pascal, which in turn was a reaction to ALGOL.***
According to the introduction to K&R, C is descended from B and BCPL. I strongly suspect that B was influenced mostly by the problems of getting any computer language more abstract than assembler to fit onto and actually do useful work on a PDP-7.
I was thinking in terms of a scenario where in say June of 2009, a routine looking Windows Update changes NTFS to write directories in a new format -- for valid security or performance reasons of course. Good Heavens, Microsoft wouldn't deliberately break Linux. Oh heavens no. How could anyone think such a thing? Old format directories are still accepted of course, but they are changed to the new format when rewritten. Windows users don't see any difference. But Linux users can't access any file in any directory where a file has been updated.
The Linux folks labor mightily and by early August, there is a new NTFS driver that handles the June directory format.... except that the July 20th Windows updates make another change to the directory format.... You can work out the scenario from there.
Slashdot Mirror
Dead On, Mate. A software project with three quarters of a million lines of code is surely going to have hundreds of open SPRs at the time of its release. If things are going well, most of them are going to be relatively unimportant. Still, You have to wonder if the FBI's CIO knows much about real software projects and how they work.
I prowled around the Internet trying to find any sort of bio on Azmi. Here's what I came up with http://www.federalnewsradio.com/?nid=5&sid=760419. His background isn't awful, but neither is it especially reassuring. There's nothing there that says to me that this isn't a guy with an expensive haircut and no proven technical or administrative ability. But he could also be fighting the good fight against heavy odds. No way to tell I think. At least he doesn't seem to be a political appointee.
Nothing against Mr Azmi, but apparently if has been common knowledge that VCF was a fiasco since early 2005. See http://www.smh.com.au/news/Breaking/New-FBI-softwa re-may-be-unusable/2005/01/14/1105582686258.html I'm sure that the problems are not Azmi's fault. Given the timeline, I don't see how they could be. But he's had two and a half years to recognize that there is a problem and get the project back on track. I have to wonder if he is the right guy to try to fix this mess.
There was, for example, nothing much great about OS/360 which was, derivative, far buggier than Windows has ever thought about being, and had the most wrong headed, obtuse and god awful scripting language ever designed. It did however provide a barely adequate vehicle for selling IBM's basically OK hardware. Great? Hell no. Influential? Yes, Very.
So, OS360, Excel, etc are influential rather than great. What would be the list of actually great software? In no particular order:
- Surely, the first higher order language? That would be Fortran, Cobol, Algol or something along that line?
- The first interpreted language? BASIC is probably the most influential, was it the first?
- The Xerox PARC Alto software which was the starting point for the Xerox Star, Lisa and eventually the Mac and Windows. I'd probably give the Macintosh the award for most influential (and I'm not a Mac guy).
- Visi-Calc. Lotus and Excel were more influential, but Visi-Calc was the breakthru.
- Unix
- The first simple OS for small computers, I'm not sure what that'd be. Something at DEC (RSX-11?) begat CPM begat QDOS begat MSDOS. But I'm sure the chain goes back further.
- The first file system. I have no idea where the idea of file systems originated, but modern users have no idea what a pain reading and writing a disk was using manual storage space allocation. There have been times as recently as the mid 1990s when I missed punched cards and some other outmoded technologies, but I've never missed manual disk space allocation.
- The World Wide Web and browsers therefore. Mosaic? Something else? Whatever
..
- Fidonet and Compuserve -- the programs that allowed folks sitting at home with dial up modems to talk to more or less randomly selected users around the world via stored text messages.
- SAGE -- An air defense system designed on the 1950s that ran on vacuum tube computers and actually worked. It supported several dozen consoles, recieved data from a dozen or more radar sites continuously, vectored manned and unmanned interceptors, talked to adjacent control centers. talked to anti-aircraft batteries, and did some other things. All using a computer roughly comperable in capacity to an IBM PC with 256KILObyes of memory. (Many subsequent attempts to replicate many of its capabilities using more modern hardware and software failed rather dismally). Great? I think so. Influential? Not very -- perhaps unfortunately.
- ...
There are plenty of others, but typing this isn't getting the lawn mowed.The drug companies need to recover enough money to support all their research -- including trials of the many compounds that just don't work out. And they also need to self-insure against liability should one of their products kill or maim a bunch of folks even after all the testing. Of course, the companies also make obscene profits and it wouldn't bother me or a lot other people much, if that particular brand of obscenity were substantially moderated.
But the system -- bad though it is -- works after a fashion and this is a case (and the only one I can think of) where simply closing all the world's patent offices, voiding all existing patents, and retraining all the patent lawyers as garbage collectors will not make the world a better place. Yes, we pay way too much for drugs until the patents expire. But in this one case, the patent-mongers are probably right. No patents, no new non-recreational drugs.
Basically, I agree with you, but as Hamming pointed out in the 1950s you can get yourself into trouble with some thing like:
A = small number, e.g. size of smallest feature in a Celeron-M CPU in microns
B = big number, e.g. distance to Andromeda galaxy in microns
C = (A+B) ... (some set of clever operations) ... - B
The math is fine, but the implementation won't yield the correct answer. because the value of interest got scaled off. That isn't rocket science, but it is suprisingly hard to catch.
That said, Floating Point works suprisingly well for most things. Back in the 1960s, things like 2+2=3.99999 were a fairly common phenomena. Hardly ever happens today. I spent about three decades in the defense industry writing and testing systems that got things from point X to point Y. I think I saw a reasonable selection of the possible errors that folks can make. Other than subtracting big numbers from other big numbers and expecting the result to be precise, I can't recall a single problem with floating point.
If we're looking for something that is basically broken to write an article about, forget floating point. Try event driven software architectures. Or the notion that it is possible to write unambiguous specifications that are also comprehensible.
I'm not sure of the cosmic significance, but if your browser doesn't support Ajax, Google presents the submenu on the main page where you (and I) want it. Perhaps this is a subtle warning from God that she disapproves of AJAX and will eventually smite those who embrace it mightily.
Anybody want to join me at home for some popcorn and maybe some orange soda while the NT crowd spends the night fixin stuff? I know it's dull, but the most of the computers around here are just too slow to run XP, so I'll just have to pass on the excitement.
Actually, we do have a couple of XP machines, but they are such a pain, that I really don't much care if some sociopathic teenager in Bucharest 'owns' them. Owning a Windows machine is exactly what hackers deserve -- at least until it's technically feasible to send them a case of jock itch or crab lice.
I must say that depending on what the worm does, it really might be 'that' serious. Trouble is that from what I read networks of real computers doing real work for real users are likely to be in considerable difficulty the first time someone plugs an RJ45 into an infected laptop and hits the power switch. I'm glad that I'm retired.
The 6809 was not only easy and fun to program, 6809 programs tended to benchmark out significantly faster than programs for comperable CPUs like the Z80, 6800 and 8080. If the industry ever decides to scrap the x86 mess -- which they won't -- going back to the 6809 for a starting point might not be a bad idea at all. I once did a plot of measured times for a benchmark where timings were available for a bunch of CPUs (Sieve of Eratosthenes). When you plotted out clockspeed vs word width, all the CPUs from the 8080 to the Cray something or other fell out into an untidy straight line, except for the 6809. There were, as I recall, three different results published for SOE on the 6809 and all three were an order of magnitude faster than they had any reasonable expectation of being based on the hardware's apparent capabilities.
The theory is that the Internet is supposed to make all manner of information available to students that would otherwise be difficult for them to access. Every student has access to the national press, encyclopedias, the latest scientific papers. And they do. The New York Times may have its problems, but it has better and broader news coverage than the Jimsomweed Junction, TX Weekly Courier or the truly awful reporting on the US TV networks. Wikipedia may not be authorative (whatever the hell that is supposed to mean), but it has generally adequate articles on a broad range of subjects -- broader than Brittanica I'm pretty sure. Live in East Bywater, Vermont and want to learn Italian or Mongolian? The Internet really can help you. In the US, school intenet access is heavily subsidized exactly because it makes information so readily available.
The Internet makes information available. However, it turns out that a fair percentage of the information that kids are genuinely interested in isn't exactly what educators, parents, and community leaders had in mind. Moreover it is amazingly difficult to tell which information is which. I suspect that most educators would (and do) look the other way with regard to a bit of pornography, and the odd illegal music download. Problem is that there is so much unfortunate stuff out there. For example, kids are perfectly capable of posting a little amateur photography done in the girl's locker room with a cell phone. That's not really all that bad, but something does have to be done about it. What about the anarchist cookbook? I think it's likely some of the recipies there are capable of distributing pieces of a classroom and the experimenter over maybe half an acre. What about white supremicist sites? Or black supremicist sites? What if some kids put up a "Why we hate Charlie Smith site"? What if Charlie sees the site and decides to hang himself? Things like that could happen. A thirteen year old in our town made a suicide pact over the Internet and carried it through. Fortunately from the school system's point of view, the kid used his home PC, not the school's. But still ... aside from the guilt thing, that's a potential multimillion dollar lawsuit.
Also remember, that in a rural area, the school is likely to have the only high speed data line in town.
Consider the costs of blocking malware downloads -- which as we all know is easier said than done. And the cost of removing the stuff that gets downloaded anyway. Some of it can take hours or days to exorcise.
So what's the answer? What most schools use are rather expensive filters whose most obvious characteristic is that they don't work all that well. Personally, I suspect that the eventual answer will be whitelist access to approved sites for most students most of the time with a few machines given full access for staff, administration, and supervised student access. But first, it seems that school administators have to recognize that they have a problem that won't go away, and school IT folks have to admit that technology can't fix their problems at reasonable cost. That'll take a while.
Do me a favor. Think about it for a while.
I'll probably send Leahy an eMail myself but I'm a little at a loss what to say. My personal opinion is that the entirety of Intellectual Property law is rotten to the core and should be pitched out -- with the exception of trademark and maybe a 15 year copyright. But I think that mainstream society is two or three decades of stupidity by "Intellectual Property owners" away from being able to deal with a view that extreme. So what should a reasonable patent law for 2006AD look like?
The BLS says total employment in Vermont is up 12.2% over the past 10 years. Population is only up 5.7% 1995-2005. So actually, jobs would seem to be moving to Vermont. They aren't necessarily great jobs of course
IBM Essex Junction is hiring. Has been for quite some time.
***Voter turnout in Vermont is no better than any other state ...***
Don't know where you got that idea. Voter turnout in Vermont is consistently one of the highest percentages in the country at 60+% in the general elections. Turnout was 70.8% in November 2004. The Burlington Free Press had an article a few years ago pointing out the state turnout percentage hasn't changed much since the 1920s, and unlike most states Vermont voter turnout didn't fall off significantly in the 70s, 80s and 90s. It is true that Town Meeting turnout has been falling off -- presumably because more and more people are working 8-5 jobs instead of farming.
***That being said...It's no better here in Nebr.. (7th highest tax burden for one of the smaller populations).*** That's curious. I've driven across the plains a few times in recent years and I was struck by how prosperous the small towns in Nebraska looked compared to Kansas and Oklahoma. Pretty superficial, but maybe you're doing better than you think.
Leahy's major constituent who has a dog in this fight is IBM. (IBM is the largest employer in Vermont -- by a lot). I'm not sure what IBM's concerns are, but as long as they are taken care of, I doubt Leahy much cares about anything other than doing the right thing. Leahy probably doesn't need large contributions to get reelected assuming that he runs again (He'll be 70 in 2010). To get a feel for Vermont politics, google 'Leahy Fred Tuttle vermont' for the story of the 1998 senate campaign that ended up with the Republican candidate asking people to vote for Leahy because Tuttle's wife didn't want to move to Washington DC. (Incidentally, by the end of the campaign, a lot of Vermonters thought Tuttle might make a pretty good senator)
Vermont politicians listen better than most American politicians. They have to. If they don't make sure that their mail is read, calls are returned and don't spend a lot of their time ouside of Washington in Grange halls and school auditoriums in tiny towns, they don't get re-elected.
So, there is a chance that well written, well reasoned, thoughtful letters, might actually cause Leahy to make changes or even withdraw his sponsorship of the bill. OK, folks what do you want to tell Leahy?
Don't drink much anymore, but I'll make an exception for that.
There's probably a Japanese name for the Agency as well, but Verizon has our DSL running at sub dial up speed today, and I don't have the patience to dig the Japanese name out.
Anyway, I reckon you can pronounce JAXA any way you want to.
A very important issue is that secrecy is expensive. It is tempting to try to protect a great deal of your data. If you do that, you will spend a great deal of money. And you will impede the work of your people who too often will not have ready access to stuff they actually need to know. And -- pardoxically -- you may reduce the protection given to data that actually needs to be secured because people will develop ways to bypass security so that they can get their job done.
IMO, many companies shouldn't even consider securing any of their data. Those that do, possibly shouldn't put their critical data on computers -- expecially not on networked computers. If all your important data can fit on 200 index cards, put it on 200 index cards and lock them up. What, for example, would be the point in putting the secret formula for Cudweiser beer (12 parts cow urine, 37 parts water, 1 part used motor oil, ...etc) on a computer where anyone can steal it?
Companies that deal in personal data e.g. schools, hospitals, etc. Here we are getting beyond my experience. I suspect that mixing sensitive data with accounts payable, staff eMail, memos about the Christmas Party, etc is not a good idea. But beyond that, I just don't know.
I think the real concern here is facility networks more than home users. If the facility users are allowed to access the Internet, and javascript is permitted by the browsers, there's an open pipeline right past whatever front end filtering is in place. I doubt it will be very practical to scan every block of javascript coming into the building for malicious code.
And how about Javascript embedded in that mother of all bad ideas -- HTML eMail?
Doesn't the second part of that make you a little nervous? One possibility is that it is finding your router and print server, but not where they are supposed to be. Could be an error in the program, but it could be some 'feature' of your network environment that you'd like to know about.
Look folks, this isn't rocket science. Given the current state of Computer Technology, downloading obscure programs from a remote source outside your control and running them can't possibly be a good idea. It may occasionally be necessary, but it's something that should be done as rarely as possible. If you don't even know the programs are there because they are buried in web pages, that just exacerbates the problem.
The answer: Turn off Javascript, and let the web site designers find some other way to entertain themselves. Forcing web sites to use HTML and server side scripting may limit their style and the coolness of your user experience, but if you want your computer(s) and network to be somewhat secure then you better forget Javascript. Personally, I have only one Javascript enabled browser (Firefox) and I try to use it only with sites like Google and a handful of others that I consider to be trustworthy.
That's a true statement. But when you think about it, it is one hell of an indictment of the Personal Computer and the industry it has spawned.
Anyway, here's a list of FCRCs. Maybe you'll find a home with one of them. Personally, I think you'd do better to rethink your position on academia. http://www.nsf.gov/statistics/fedfunds/pubs/ffrdc/ ffrdc.txt
As for studies of computers. Lot's of them. Mixed results, but generally mildly negative. Costs of developing and maintaining curricula are high as are support costs for the hardware. And don't get me started on Windows(especially XP), Office, internet filtering, viruses, malware, etc, etc, etc. Google "River Oaks Ontario" for an experiment in depth in computing in schools with mixed results. Some educational positives, but support costs were reported to be unsustainably high once Apple stopped underwriting them.
Computers have some unquestioned utility in school administration and class materials preparation. Some other uses in schools in the US are probably beneficial. Computers will surely someday be a major element in education. But not in K-8 in the US today. I have no idea how things change in an undeveloped country. Maybe computers would be a blessing. Or maybe the wisest thing that India, Paraguay and Chad could do would be to confiscate anything that looks like a computer at the school's front door.
I think that many web design issues are self inflicted by trying to force HTML into specific presentation layouts. It really wasn't designed to do that. If IE trashes simple markup, that's one thing. If it just looks a bit different than Opera/Netscape, maybe the problem is unreasonable expectations. I'm not sure how much complaint one is entitled to if they stagger out of the pub, climb into someone else's wagon, and the horse takes them to the wrong house.
I understand that everyone seems to want a language that allows presentation layout to be specified. That's a perfectly reasonable desire. Probably such a language should be written. Maybe the authors of IE should have respected the desire to control presentation and tried to minimize differences in results from pre-existing browsers like Netscape. Hell, maybe they did. It's really hard to anticipate or identify all the peculiarities of browsers. For example, one popular browser -- I've forgotten which, but it wasn't IE, seemed to think that (33+33+33)>100, and therefore displayed three images of 33% width on two lines. It put them in one line if they were 32% width. (For all I know there is a good reason for that). Expecting the folks who wrote IE to know hundreds or thousands of non-obvious behaviors and emulate them in their browser may not be realistic.
Maybe what all this demonstrates is that Dvorak has a point for once. This stuff really doesn't work all that well.
That's true only if you think a chocolate mousse must, of necessity, be a large herbivore.
Yes, HTML is largely specific to a document layout -- a sort of ornamented outline without paragraph numbering. It is not specific to a presentation layout. Confusing the two kinds of layout will generally cause anyone who does it substantial grief. What I See Is very likely Not What You Will Get (WISINWYG?) -- by intent.
At the risk of pointing out the obvious, HTML was intended to be a MARKUP language, not a LAYOUT language. Why would anyone who understands the meaning of those two words be suprised that dictating LAYOUT in HTML is next to impossible? It's sort of like being constantly suprised that your car, unlike your horse, can not find its way home from the village pub after you have drunk yourself into a stupor.
Since most of the world seems to want a layout language for the web, I reckon someone ought to design one. But my guess is that HTLL (for want of a better name) it is going to be a bundle of grief as people attempt to cope rationally with differing screen sizes, color depths, font availabilities etc.
By all means, create your HTLL. But if it is not too much to ask, how about leaving HTML as a markup language where the browsers have reasonable latitude in presentation? Really, almost always, all I want to do is see the damn content, and all this garbage intended to dictate layout seems to trash the content at least as often as it does in forcing the layout that the page author intended.
According to the introduction to K&R, C is descended from B and BCPL. I strongly suspect that B was influenced mostly by the problems of getting any computer language more abstract than assembler to fit onto and actually do useful work on a PDP-7.
The Linux folks labor mightily and by early August, there is a new NTFS driver that handles the June directory format. ... except that the July 20th Windows updates make another change to the directory format. ... You can work out the scenario from there.