Oh, no, they've already missed the Christmas ship date. It's not coming out until Jan. 1 next year at the earliest. And with all the features they've had to pull from it, there is no reason to update to it until at least 1 year after it's released, anyway.
If new OS's and their features turn out to be vital in the next year, it's a good time to be selling Linux or other OS solutions.
Unfortunately, many SSH and SSL users have no idea that it's important to protect the private keys, and do an absolutely horrible job of protecting them. Also, I know of at least 3 distinct man-in-the-middle attacks currently in use, all of which involve either stealing a server's private key or getting an unsophisticated user to simply "accept this key" to monitor all SSL traffic. SSH users are even worse about it, because there is no "web of trust" to signed keys.
Of course, since major companies like Microsoft and Thawte and Google have all cooperated with China's or the US government's "legal requests", the web of trust can be considered broken by the weight of the huge fat spider sitting in the middle of it with a sealed, even secret warrant.
No, encrypted communications do *NOT* look like just random bits. There are a few popular forms of encrypted communications on the Internet, such as SSH and HTTPS. Those have very specific formats indeed, and are easily identifiable to even a remotely intelligent traffic monitor. Encrypted email is even more identifiable: it's still email, it's still port 25 from one Mail Transfer Agent to another. Making the encrypted traffic look like something harmless is a while different layer of complexity, and gets into steganography, which is a whole different art form.
Although I have to admit, if people would concentrate on clearing out the poison ivy instead of scratching their personal itch, there'd need to be a lot less scratching. The "poison ivy" is the plethora of badly written tools already in place, with seriously unfortunate user interfaces.
A famous write-up of the failures of user interfaces and configuration tools in open source got slashdotted when written by Eric Raymond, several years ago, at http://www.catb.org/~esr/writings/cups-horror.html. It's even funnier because the CUPS authors responded very graciously to his complaints by saying they'd fix at least some of them, and don't seem to have actually fixed *ANY* of the things Eric griped about in the 2 years since then.
Nagios is fairly CPU intensive, and the client plug-ins to report on system load and other local characteristics are not well integrated, since they basically date back to NetSaint and a lot of legacy oddness that could stand a complete rewrite. A lighter weight monitoring tool would be good, or a a rebuild of Nagios, especially if most of the worst-built Nagios plug-ins were thrown out due to the extremely poor quality of the Perl or shell code involved.
But there is no hint that this particular set of monitoring and system management tools have anything better than Nagios or the dozens of other monitoring tools also available. Nagios, for example, is a bit complex because it has a *lot* of features that you may not want in your first 15 minutes of monitoring but may really need in your next week of operation. MRTG is the same way.
Agreed. The Alpha developers need to turn off those household furnaces and move on to something remotely useful. And what is the point of doing i386 versions of anything when almost no modern Intel or AMD PC's are less than i686?
I'm sorry, but you've fallen prey to the idea of "anathema": subjects so dirty or vile that any contact with them whatsoever contaminates otherwise innocent subjects. Frankly, the child porn problem through Google or other search engines is not only minor, it's nearly nonexistent. As near as I can tell, most child porn out on the Internet is actually being offered by law enforcement to lure its purchasers into illegal behavior. And a lot of it is frankly entrapment. "Entrapment" is luring people to do illegal things they would not have done if you didn't lure them into it, and it's a nasty legal problem.
If you're not aware of this sort of behavior, I suggest you look at the famous "Amateur Action" case. A postal inspector in Tennessee shipped child porn to a BBS in California, which carried only material legal in their home state of California, and got them raided on the basis of that box of material, which they had never opened after several days because they didn't order it and didn't recognize the sender! The BBS was then sued on the basis of their other content, which was ruled illegal in Tennessee and which they hadn't sent there without the client signing a waiver that said they were obeying local laws. In this case, the client was the postal inspector himself! The Postal Inspector tried to conceal his role in sending the porn, until someone compared the signature on the child porn shipping label to the signature on the warrant to search the building.
This sort of stupidity is unfortunately common in the "protect the children" world: people use it to hide their censoring policies behind the flag of "protecting the children", because no one wants to be seen as risking or harming children.
Material that is obviously child portn also gets Usenet cancelled very, very fast indeed: servers that do not support cancel messages for the various alt.binary groups where most Usenet porn winds up are begging to be deluged and overwhelmed by spammers. So the likelihood of such messages being found on a normal Usenet feed has dropped, especially with the advent of websites, anonymizing proxy servers, and popular pirate software resouces such as the old "FSP" servers.
Only a complete idiot will post child porn on Usenet these days, since NNTP messages started generally including the posting host's IP address.
Sir or madam, if your coding style matches your posting style, I suspect that your HTML is awfully bloated. It's just a suggestion, but please focus on your points and leave out the lengthy descriptions.
Second, for people learning HTML right now, I really recommend the Amaya editor from the World-Wide Web Consortium, at http://www.w3.org/Amaya/. It's open source, well-structured, generates clean code without thousands of lines of unnecessary style and inclusions that the Microsoft editors generate, and encourages people to look at what they're doing and the resulting code.
The complexity of HTML now really does make it equivalent to a programming language, well worth learning, even if technically it's only a markup language.
Python is legible. As much as I've debugged some god-awful crap masquerading as fast code in the Python world, it has been nothing like trying to debug the horrid tangled mess of interwoven hand-waving I've seen with LISP.
This is partly due to the legibility of Python code, with its well-designed built-in libraries for I/O and data handling. It's also partly due to the habits taught to early programmers. Clever recursion is an available technique, not the goal of the first 3 months of course work in it.
I agree with you. But if someone wants to learn real programming these days, I'd start with a simpler scripting language like Perl or Python, which exposes people to pointers, functions, etc. and can do meaningful things very easily. And teach it wall: the number of really bad Perl programmers out there is really scary, due to huge plethora of error ignoring ways to do the same task and the tendency to cut and paste someone else's badly written code.
Then source control systems, so they learn how to work on a project together and keep track of their changes so they can revert things as needed when they write more complex scripts.
Then I'd introduce them to simple C programs, such as the Alec Moffett's old "crack" program for checking easy passwords, just to get their attention and introduce them to compiled programs that work under good old gcc. Then build systems, such as "make" or "ant", so they learn how to build the software up.
That kind of process can introduce them to good open source projects as well: there are plenty of projects on sourceforge.net that could use a classroom of typing moneky programmers to add a few lines of Shakespeare-like code to them.
LISP use has the word "master" in it, but not spelled with an "e". That monstrosity teaches the worst hand-waving programming practices I have ever seen by teaching people to completely ignore the other levels of the program. Putting (and (then (a (miracle (occurs ()))))) and getting all the parentheses matched up does not count as programming. As beautiful as it is for theoretical analyses, it is horrid for programming anything that needs to actually handle data reliably due to the lack of constraints on operations, and it's even worse for anything performance bound because of the over emphasis on recursion taught with it.
Excuse me, please: I've had very nasty times helping new programmers unlearn hideous practices learned from LISP homework assignments.
Buddy, your experience is familiar. But could you shorten up your writing a bit? In your code as well as your commentary, leave out things you don't need to hit in this particular message and your work will be learner, faster, and easier to read.
But if someone can get your name, they should be hiring you for the summer. Definitely look for local internship programs or college courses, you set of skills and basic compreshension sound like you'll be a good prize for whatever group your work with.
True, but it's amazing what you can do with a Knoppix CD.
Comparing Word and Excel to programming classes ia an error, though. Compare it to typing and home economics classes, where you learned to balance a checkbook and write a nice job application, and you'll see what those classes are really focused on: they're useful skills in today's world.
The lack of real computer and programming classes is a separate issue, like the failure to do real experiments in chemistry clase and the use of training films rather than real exercise in gym class.
Or it's a point in favor of having massive air support, electronic surveillance, control of the major highways for re-supply, incredible amounts of ammunition available, and easy access to fresh hand grenades along with a highly trained modern army, much of which is pretty highly paid mercenaries rather than guerrilla rablle to keep the M-16 functional.
Re:Managers know jack, geeks know all
on
Viiv Falls Flat
·
· Score: 1
I don't take them apart enough to know the internals, but they do seem to be on a steady diet of modest improvements. And Apple hardware has been consistently good for years: it's the add-on third party peripherals that have sometimes been extremely flakey.
Re:Managers know jack, geeks know all
on
Viiv Falls Flat
·
· Score: 1
I lived through the '80's. Take a look at the Sony Walkman: that wasn't genius engineering, it was a real market niche with a good product *that steadily got better with small refinements from the factory floor*. It wasn't a big splash: they didn't spen all their development budget on marketing, they eased into the market and used sales income to improve the next generation, staying a step ahead of the knock-off competitors and becoming the generic name for such products.
The engineers were brilliant, gracefully and reliably merging all those little tweaks into the design (such as better controls, more reliable motors, lighter cases, better battery covers, etc., etc., etc.) The managers were so good they were unnoticed: they kept their people working towards that, rather than taking up expensive new "Panasonic-killers" that would get them a great line in their resume but eat the company's profits right up. The result was a very comfortable, reliable device that led the marketplace.
The process of small improvements, especially from the factory floor, is called "Kaizen", described at http://en.wikipedia.org/wiki/Kaizen. It's at the root of some of the best technologies we've seen, but it's fallen by the wayside in US manufacturing where the next big product announcement to bump the next quarter's revenues is much more important than improving the basic product.
My god, Neverhood was brilliant! I did buy it and it was well worth every cent.
Another discarded game or line of games is Outcast, which had very good game play, interesting plots, and a priceless sense of humor. I'd love to see that released for new single-player modules.
I played Gunship. They should have given *me* $50 for wasting my time on it.
Re:Media Center software is not commercially viabl
on
Viiv Falls Flat
·
· Score: 1
It's at odds at the moment, but look carefully at the Trusted Computing initiative, formerly called Palladium. It's being implemented in the Intel CPU's directly, and provides DRM and authentication of software and hardware to force the use of specific software to unlock specific hardware features for specific files.
That is exactly what is needed to provide media distributor of control home recording of television shows, cable broadcasts, and media such as CD's and DVD's. It's also why, for now at least, you need Intel CPU's to do this: the separate Palladium chipsets are too expensive.
Sadly, this is often much more effective and reliable than expensive commercial high-availability systems where the management tools and the absurdities they commit on the actual software actually make the system unreliable and liable to intermittent failures. Before spending the money on using such tools it's often worth examining what downtime costs and factoring in the likelihood of the high-availability tools failing or in fact never working at all.
But a monitoring tool that actually and effectively reports that the server is down or offline is necessary in any case, and costs one heck of a lot less.
It looks like whoever actually wrote that note on the MySQL site does not know how GPL licenses work. It's a violation of the GPL to emulate the protocol? Nonsense, unless there are patents on it.
I agree: buying a monitor based on look alone is only for people who have secretaries to read all their email. The durability, the brightness, the contrast, the ease of the controls, the accuracy of the colors, the responsiveness of the screen, the cost, the evenness of the brightness all across the screen matter a lot if you're doing artistic work, viewing DVD's, or playing games.
But/.'ing an article that says a monitor stinks as a monitor but it has a pretty design is like putting up a billboard that says "Starbucks serves huge cups of coffee!" and ignoring the part that says "which tastes like cream filled cans of turpentine". It ignores the actual review.
I have, but it was years ago, and it was very painful. We finally threw out the Oracle server for something much lighter weight that did exactly what we wanted, and was open source, and run on hardware that was not top of the line in every way.
For multiple versionf of Oracle, you use "/opt" and put it in different version-based locations, just like kernels do for/boot. For multiple copies of it, there' is a fascinating command called "--relocate" that allows certain types of packages to be easily installed to alternative locations.
Do read the man page for the rpm command.
Slashdot Mirror
Oh, no, they've already missed the Christmas ship date. It's not coming out until Jan. 1 next year at the earliest. And with all the features they've had to pull from it, there is no reason to update to it until at least 1 year after it's released, anyway.
If new OS's and their features turn out to be vital in the next year, it's a good time to be selling Linux or other OS solutions.
Unfortunately, many SSH and SSL users have no idea that it's important to protect the private keys, and do an absolutely horrible job of protecting them. Also, I know of at least 3 distinct man-in-the-middle attacks currently in use, all of which involve either stealing a server's private key or getting an unsophisticated user to simply "accept this key" to monitor all SSL traffic. SSH users are even worse about it, because there is no "web of trust" to signed keys.
Of course, since major companies like Microsoft and Thawte and Google have all cooperated with China's or the US government's "legal requests", the web of trust can be considered broken by the weight of the huge fat spider sitting in the middle of it with a sealed, even secret warrant.
No, encrypted communications do *NOT* look like just random bits. There are a few popular forms of encrypted communications on the Internet, such as SSH and HTTPS. Those have very specific formats indeed, and are easily identifiable to even a remotely intelligent traffic monitor. Encrypted email is even more identifiable: it's still email, it's still port 25 from one Mail Transfer Agent to another. Making the encrypted traffic look like something harmless is a while different layer of complexity, and gets into steganography, which is a whole different art form.
Although I have to admit, if people would concentrate on clearing out the poison ivy instead of scratching their personal itch, there'd need to be a lot less scratching. The "poison ivy" is the plethora of badly written tools already in place, with seriously unfortunate user interfaces.
l . It's even funnier because the CUPS authors responded very graciously to his complaints by saying they'd fix at least some of them, and don't seem to have actually fixed *ANY* of the things Eric griped about in the 2 years since then.
A famous write-up of the failures of user interfaces and configuration tools in open source got slashdotted when written by Eric Raymond, several years ago, at http://www.catb.org/~esr/writings/cups-horror.htm
Nagios is fairly CPU intensive, and the client plug-ins to report on system load and other local characteristics are not well integrated, since they basically date back to NetSaint and a lot of legacy oddness that could stand a complete rewrite. A lighter weight monitoring tool would be good, or a a rebuild of Nagios, especially if most of the worst-built Nagios plug-ins were thrown out due to the extremely poor quality of the Perl or shell code involved.
But there is no hint that this particular set of monitoring and system management tools have anything better than Nagios or the dozens of other monitoring tools also available. Nagios, for example, is a bit complex because it has a *lot* of features that you may not want in your first 15 minutes of monitoring but may really need in your next week of operation. MRTG is the same way.
Agreed. The Alpha developers need to turn off those household furnaces and move on to something remotely useful. And what is the point of doing i386 versions of anything when almost no modern Intel or AMD PC's are less than i686?
I'm sorry, but you've fallen prey to the idea of "anathema": subjects so dirty or vile that any contact with them whatsoever contaminates otherwise innocent subjects. Frankly, the child porn problem through Google or other search engines is not only minor, it's nearly nonexistent. As near as I can tell, most child porn out on the Internet is actually being offered by law enforcement to lure its purchasers into illegal behavior. And a lot of it is frankly entrapment. "Entrapment" is luring people to do illegal things they would not have done if you didn't lure them into it, and it's a nasty legal problem.
If you're not aware of this sort of behavior, I suggest you look at the famous "Amateur Action" case. A postal inspector in Tennessee shipped child porn to a BBS in California, which carried only material legal in their home state of California, and got them raided on the basis of that box of material, which they had never opened after several days because they didn't order it and didn't recognize the sender! The BBS was then sued on the basis of their other content, which was ruled illegal in Tennessee and which they hadn't sent there without the client signing a waiver that said they were obeying local laws. In this case, the client was the postal inspector himself! The Postal Inspector tried to conceal his role in sending the porn, until someone compared the signature on the child porn shipping label to the signature on the warrant to search the building.
This sort of stupidity is unfortunately common in the "protect the children" world: people use it to hide their censoring policies behind the flag of "protecting the children", because no one wants to be seen as risking or harming children.
Material that is obviously child portn also gets Usenet cancelled very, very fast indeed: servers that do not support cancel messages for the various alt.binary groups where most Usenet porn winds up are begging to be deluged and overwhelmed by spammers. So the likelihood of such messages being found on a normal Usenet feed has dropped, especially with the advent of websites, anonymizing proxy servers, and popular pirate software resouces such as the old "FSP" servers.
Only a complete idiot will post child porn on Usenet these days, since NNTP messages started generally including the posting host's IP address.
Sir or madam, if your coding style matches your posting style, I suspect that your HTML is awfully bloated. It's just a suggestion, but please focus on your points and leave out the lengthy descriptions.
Second, for people learning HTML right now, I really recommend the Amaya editor from the World-Wide Web Consortium, at http://www.w3.org/Amaya/. It's open source, well-structured, generates clean code without thousands of lines of unnecessary style and inclusions that the Microsoft editors generate, and encourages people to look at what they're doing and the resulting code.
The complexity of HTML now really does make it equivalent to a programming language, well worth learning, even if technically it's only a markup language.
Python is legible. As much as I've debugged some god-awful crap masquerading as fast code in the Python world, it has been nothing like trying to debug the horrid tangled mess of interwoven hand-waving I've seen with LISP.
This is partly due to the legibility of Python code, with its well-designed built-in libraries for I/O and data handling. It's also partly due to the habits taught to early programmers. Clever recursion is an available technique, not the goal of the first 3 months of course work in it.
I agree with you. But if someone wants to learn real programming these days, I'd start with a simpler scripting language like Perl or Python, which exposes people to pointers, functions, etc. and can do meaningful things very easily. And teach it wall: the number of really bad Perl programmers out there is really scary, due to huge plethora of error ignoring ways to do the same task and the tendency to cut and paste someone else's badly written code.
Then source control systems, so they learn how to work on a project together and keep track of their changes so they can revert things as needed when they write more complex scripts.
Then I'd introduce them to simple C programs, such as the Alec Moffett's old "crack" program for checking easy passwords, just to get their attention and introduce them to compiled programs that work under good old gcc. Then build systems, such as "make" or "ant", so they learn how to build the software up.
That kind of process can introduce them to good open source projects as well: there are plenty of projects on sourceforge.net that could use a classroom of typing moneky programmers to add a few lines of Shakespeare-like code to them.
LISP use has the word "master" in it, but not spelled with an "e". That monstrosity teaches the worst hand-waving programming practices I have ever seen by teaching people to completely ignore the other levels of the program. Putting (and (then (a (miracle (occurs ()))))) and getting all the parentheses matched up does not count as programming. As beautiful as it is for theoretical analyses, it is horrid for programming anything that needs to actually handle data reliably due to the lack of constraints on operations, and it's even worse for anything performance bound because of the over emphasis on recursion taught with it.
Excuse me, please: I've had very nasty times helping new programmers unlearn hideous practices learned from LISP homework assignments.
Buddy, your experience is familiar. But could you shorten up your writing a bit? In your code as well as your commentary, leave out things you don't need to hit in this particular message and your work will be learner, faster, and easier to read.
But if someone can get your name, they should be hiring you for the summer. Definitely look for local internship programs or college courses, you set of skills and basic compreshension sound like you'll be a good prize for whatever group your work with.
True, but it's amazing what you can do with a Knoppix CD.
Comparing Word and Excel to programming classes ia an error, though. Compare it to typing and home economics classes, where you learned to balance a checkbook and write a nice job application, and you'll see what those classes are really focused on: they're useful skills in today's world.
The lack of real computer and programming classes is a separate issue, like the failure to do real experiments in chemistry clase and the use of training films rather than real exercise in gym class.
Or it's a point in favor of having massive air support, electronic surveillance, control of the major highways for re-supply, incredible amounts of ammunition available, and easy access to fresh hand grenades along with a highly trained modern army, much of which is pretty highly paid mercenaries rather than guerrilla rablle to keep the M-16 functional.
I don't take them apart enough to know the internals, but they do seem to be on a steady diet of modest improvements. And Apple hardware has been consistently good for years: it's the add-on third party peripherals that have sometimes been extremely flakey.
I lived through the '80's. Take a look at the Sony Walkman: that wasn't genius engineering, it was a real market niche with a good product *that steadily got better with small refinements from the factory floor*. It wasn't a big splash: they didn't spen all their development budget on marketing, they eased into the market and used sales income to improve the next generation, staying a step ahead of the knock-off competitors and becoming the generic name for such products.
The engineers were brilliant, gracefully and reliably merging all those little tweaks into the design (such as better controls, more reliable motors, lighter cases, better battery covers, etc., etc., etc.) The managers were so good they were unnoticed: they kept their people working towards that, rather than taking up expensive new "Panasonic-killers" that would get them a great line in their resume but eat the company's profits right up. The result was a very comfortable, reliable device that led the marketplace.
The process of small improvements, especially from the factory floor, is called "Kaizen", described at http://en.wikipedia.org/wiki/Kaizen. It's at the root of some of the best technologies we've seen, but it's fallen by the wayside in US manufacturing where the next big product announcement to bump the next quarter's revenues is much more important than improving the basic product.
My god, Neverhood was brilliant! I did buy it and it was well worth every cent.
Another discarded game or line of games is Outcast, which had very good game play, interesting plots, and a priceless sense of humor. I'd love to see that released for new single-player modules.
I played Gunship. They should have given *me* $50 for wasting my time on it.
It's at odds at the moment, but look carefully at the Trusted Computing initiative, formerly called Palladium. It's being implemented in the Intel CPU's directly, and provides DRM and authentication of software and hardware to force the use of specific software to unlock specific hardware features for specific files.
That is exactly what is needed to provide media distributor of control home recording of television shows, cable broadcasts, and media such as CD's and DVD's. It's also why, for now at least, you need Intel CPU's to do this: the separate Palladium chipsets are too expensive.
Sadly, this is often much more effective and reliable than expensive commercial high-availability systems where the management tools and the absurdities they commit on the actual software actually make the system unreliable and liable to intermittent failures. Before spending the money on using such tools it's often worth examining what downtime costs and factoring in the likelihood of the high-availability tools failing or in fact never working at all.
But a monitoring tool that actually and effectively reports that the server is down or offline is necessary in any case, and costs one heck of a lot less.
It looks like whoever actually wrote that note on the MySQL site does not know how GPL licenses work. It's a violation of the GPL to emulate the protocol? Nonsense, unless there are patents on it.
I agree: buying a monitor based on look alone is only for people who have secretaries to read all their email. The durability, the brightness, the contrast, the ease of the controls, the accuracy of the colors, the responsiveness of the screen, the cost, the evenness of the brightness all across the screen matter a lot if you're doing artistic work, viewing DVD's, or playing games.
/.'ing an article that says a monitor stinks as a monitor but it has a pretty design is like putting up a billboard that says "Starbucks serves huge cups of coffee!" and ignoring the part that says "which tastes like cream filled cans of turpentine". It ignores the actual review.
But
I have, but it was years ago, and it was very painful. We finally threw out the Oracle server for something much lighter weight that did exactly what we wanted, and was open source, and run on hardware that was not top of the line in every way.
For multiple versionf of Oracle, you use "/opt" and put it in different version-based locations, just like kernels do for /boot. For multiple copies of it, there' is a fascinating command called "--relocate" that allows certain types of packages to be easily installed to alternative locations.
Do read the man page for the rpm command.