[in answer to "what if your child ruins my child's life?":] You can repair that.
Can you? What makes you so sure?
I'm one of those who undoubtedly would have been reported to the folks running this program. In my days in high school, I was a full year younger than the next youngest person in my class. I was smaller, lighter, weaker, and slower than nearly all of my peers. I was also the one they turned to in computer math class (in the days when computer math was taught via teletypes talking to a mainframe halfway acros town) - but no other time. The rest of the time, I was bullied, teased, beaten, and ostracized. As it is, it took a year of psychotherapy and antidepressants, 20 years later, to begin to reverse the effects of my high school experience. I'm still getting over it. I'm now 39 years old, about to turn 40, and only in the past couple of years have I been anything approaching happy.
I know what the kids who commit school violence are going through. I was there myself. The first thought that popped into my head when I heard about Columbine was "That could have been me...".
Those kids don't need others turning them in anonymously, with or without reward, so they can be further singled out for ostracism and abuse at the hands of an uncaring society. They need HELP. They need to know that someone cares about them. They need understanding. They need acceptance.
I'm sorry you can't accept them, whether in the name of your kids' safety or otherwise. The simple truth, though, is that this is exactly the wrong approach. Reach out to these kids. Don't turn them in, or away, or out...
I wouldn't wish any parent to have to deal with having their kid murdered. The solution, as someone wrote in reply to Katz' original article, is to have schools be places no kid would dream of blowing up. If we can accomplish that, then no parent need have that fear again. --
The best thing SCO could do right now is to free the trademark "Unix" SCO doesn't own the trademark, the X/Open Consortium does. Novell donated it to them several years ago. X/Open will license it to any system that can pass the Spec 1170 suite of compatibility tests. That's how DEC OSF/1 became Digital Unix. --
For various reasons, I need to get myself registered and travel booked and so on for a Linux show this year. I'd been planning to go to Linux Expo, but with that one out of the picture, what should I go to instead? I missed Linux World in NYC, and Atlanta is too far in the future - if I pick that one, the money for the trip is likely to dry up before I get to go. What's a good Linux show to go to instead? preferably a week-long event with some real meat to it. --
Once again, RMS has missed a truly crucial point: Amazon has a fiduciary duty to its shareholders to protect its corporate assets. Those include patentable ideas. Further, they also have a fiduciary duty to their shareholders to guarantee their corporate value does not get diminished by others.
Put this together, and Amazon did exactly what they are required by law to do: Suing B&N for patent infringement. If they didn't, Bezos and the rest of the Amazon board could and probably would have been sued by their shareholders.
Companies have a legal duty to protect themselves, and the legal system is an excellent example of "the best defense is a good offense".
Of course, in RMS' communist utopia, such things wouldn't exist, but this is the real world...
I think I'll go buy a book from Amazon in RMS' honor. --
No, I haven't. I guess the central question is: How much time do you spend writing code that does the real work of your application, and how much time do you spend writing code that handles the mechanics of the user interface? My experience with Delphi on the payroll project is that nearly 95% of my time is spent on writing the application, not the mechanics. Can Qt achieve that? --
I think the word RAD is a buzzword. Sure, you can place out a button a few seconds faster in VB than QT, but how much time of the overall project-time does that account for? 2% ?
The real beauty of a RAD environment such as Delphi is that it removes the need to explicitly code the standard boilerplate that has to be done every time you add a user interface control to an application, and the message handlers, and control blocks, and on and on and on...Anyone who's written a Windows (or OS/2 PM, or X, or other stuff) app the old-fashioned way will be amazed at how much time you spend working on the problem to be solved, rather than in the tedious low-level stuff, in a RAD environment. --
Huh?! Unless you're seriously advocating that a 300K program-in-development that doesn't do anything approaching what the user wants is somehow better than a 1.5 MB program that does, you missed my point entirely. My point is more about development leadtimes than about code size.
I'm estimating that the same program, coded with vi and make and done the old-fashioned way, would come out somewhere around a megabyte of finished executable. Okkay, the Delphi program is 50% bigger. Big, fat, hairy deal. The version coded the old-fashioned way would take about 10 times as long to develop and debug. In the world of business, where people expect their computers to actually do useful work, that tradeoff is a no-brainer.
Without a RAD tool, Linux will never make it out of the server room and onto people's desktops, for it's completely unsuitable as an applications platform. People following the One True Linux Way as you advocate it will get run over by others who beat them to market with bigger, slower, but running code.
I care as much about Linux as you claim to. However, to me, Linux has the potential to replace Windows in many more roles than just Web and file servers. To compete in those spaces, it must have comparable capabilities, and Delphi brings a very important set of those capabilities to the Linux world. Market acceptance is much, much more important then ideological purity if the goal is to have Linux be accepted as widely as Windows. --
Who would want to code using the Delphi or C++ Builder environments? Even for a Windows application their interface sucks, and the underlying code base isn't much better. It supports a vast, bloated and confused class structure which encourages the creation of slow, windy programs. Why do we need software like this when we have tools like vi and make already part of Linux? Have you ever developed a serious application with Delphi? I have and am, a large, very customized payroll system for a company with extremely nonstandard payroll requirements. (Their chart of accounts is over 70,000 long, and 95% of that is payroll for 450 employees.) It would have taken me 10 times as long to develop this program in a non-RAD environment. Yes, the program is larger than it would be if I'd used more traditional development tools...but I'd still be developing early functionality, instead of getting ready to hand them a feature-complete version. What's better, a program that they can use that has a 1.5 MB load module, or a 300K load module that they can't use to get real work done? --
No one "reversed engineered" the PC BIOS. It was under an open patent. IBM did not hand them the design, but the legal lisence the BIOS was under was open.
Not so. Compaq did a clean-room implementation of the PC BIOS, instead of using the published code. Several cloners who used the published code were sued by IBM (anyone remember Victor?) and faded away, and Compaq withstood its suit and thrived.
(Disclaimer: Yes, I work for Compaq, but that all happened long, long before I got here.) --
With the big computer companies such as Compaq embracing Linux, what is it that differentiates VA Linux Systems? Why should a company buy one of your boxes instead of one of theirs? --
You're right about the debating style. I should take the time to be more detailed about it. I should do lots of other things, too, that I don't...
Even so, I can't help thinking that, had I posted the same style message slamming the NetBSD Foundation as unsuitable to receive this award, it would not have been moderated down three times, and probably not at all. I consider this more evidence that the Most Holy Stallman, Pope of the Only True Free Software Movement (say hallelujah!), is considered untouchable by the vast majority of people on this forum.
But if you're hooked on karma, you have to be careful how you say what you say.
I dunno if I'd consider myself hooked on karma, but having one post zap me below the +1 threshold was a bit discouraging. I like to think of Linux people as more hackish than most, and therefore more likely to be accepting of opinions other than their own; being proven wrong time and again wears on the nerves. --
Of course, doing anything but following the Slashdot sheep in bowing and scraping before the revered RMS has clobbered my karma, as I knew it would... --
(In fact, in the US it's silly to talk about "licensing" forms of communication, as all of them should be covered by the First Amendment, but that's another argument for another day.)
The First Amendment covers content, not manner of speech. You can say anything you want, but the government has a right to regulate how and when and where in order to keep the peace. --
I'm hoping that this change will start to address the age gap in ham radio, and I'll be working on a campaign to get young people into the hobby and on to our HF bands.
Just like you hoped the no-code Tech would, too? Not to mention bringing in a flood of technically competent folks to revolutionize the state of radio communications, all on a shoestring budget in the ham bands, results published in QST and CQ. Right.
The simple truth of the matter is that the no-code license has failed in its stated objective. The average age of hams has risen at one year per year of elapsed time, and that rate is unchanged before and after the advent of the no-code license. As I've contended all along, the real problem is that ham radio is not attractive to kids today. Your campaign is 10 years too late, and ham radio will suffer for it.
One of the best things about this decision is that it ends a very ugly acrimonious situation in ham radio that has persisted since 1990, when the no-code VHF license was introduced as the first foot in the door for modernization of ham radio. A lot of the older hams alienated the younger ones because they felt that no-coders weren't real hams. Now, those younger hams will have the same licenses as the older ones, and will be in their faces on the HF bands.
Oh, so now the FCC can change human nature with a simple Report and Order? Those of us who had to work to earn our privileges are going to be as resentful of those who do not under the new scheme as was the case with the no-code Tech. This change will, if anything, perpetuate the alienation and division, and spread it to the HF bands as well. Many older hams have resented, and IMAO rightfully so, the dumbing down of the entrance requirements in the name of attracting the "right people" into the service, especially given the 10 years of failure to do anything of the sort that we've seen.
The real solution, as I have been arguing ever since the no-code argument started, is to convince people that ham radio is worth the entrance requirements. No matter how low you set the barrier, the real problem is not getting people past it - the widely diverse ham population before 1990 is all the proof anyone can ever need that desire will conquer all - but getting people to want to climb over it. We should have tried better marketing in 1990, instead of assuming that people were too dumb any more to pass the requirements. This change can only continue the slide toward chaos and lawlessness that we've seen - case in point, the 147.435 LA repeater, a blight on the face of the service if there ever was one - ever since the requirements were first lowered.
I'm one of those (and probably the only one who will post on slashdot for attribution) who thinks dumbing down the code requirement is a Bad Idea. I expect to get flamed for it, but this is one issue I've given a lot of thought to over the last 20 years or so, and don't expect to be convinced otherwise.
Ham radio is not only about technical innovation. Yes, that's one purpose, and one that I, too, believe is important. It's not the only purpose, however. The basis and purpose of the service, as expressed in 97.1, includes fostering international goodwill and providing emergency communications when needed. It takes more than a technogeek to accomplish these ends. One must be able to *communicate*. The dual nature of the ham tests serve, IMAO, to point that out: you can answer all the theory questions they can throw at you and still not be an effective operator. The code test provided some way to check that. To get the Extra ticket, you had to be a well-rounded ham, both a skilled operator and at least fairly knowledgeable technically.
The FCC has goofed here, aided by public pressure from well-meaning techies who can't see that building the technology isn't enough: you have to make use of it, too, to provide communications.
Jay Maynard, K5ZC (licensed as Amateur Extra in 1977) --
The previous poster misspoke. On the Alpha, sizeof(void *)=sizeof(long)!=sizeof(int). ints are 32 bits, longs and pointers are 64 bits. Thus, programmers who think all the world's a Pentium and treat pointers and ints as interchangeable make life hell for us Alpha types.
Why the hell anyone would want to sling pointers around as though they were integers escapes me, anyway; that's a recipe for disastrous, hell-to-find bugs. There's ALWAYS a better way to get there. --
It's worth noting that the Pentium Pro/II/III have a 48-bit segmented addressing mode, allowing physical memory beyond 4GB. Nobody uses this yet, but it's in there.
Actually, SCO UnixWare 7.1.0 (at least) does. I have a Compaq ProLiant 8500 with 8 PIII/500s and 5 GB of RAM on my test bench, and UW will make use of all of it.
(Disclaimer: I work for Compaq. I don't speak for them.) --
Is it a recommended choice for someone only running linux? I mean, do I get twice the performance for the same overall price?
Depends on what you're doing. If you're doing number crunching, it's definitely worth it; if you're doing networking, it's not a bad way to go, as the script kiddiez don't have Alpha exploits to try to run on your system (at least not yet); if you're looking for a general user box, it's about neck-and-neck. My Alpha 21164/500 gets over twice the CSC key rate of my PIII/550. For other things, it feels about as snappy. --
Tell the average Joe about the Alpha and he's like "Alpha what?"
A friend, who works as a tech support manager at a major computer software company which shall remain nameless, tells me that one of her people got a call one day, complaining that they had untested software on their web site for download...
...seems he saw the Alpha versions and thought they were alpha test releases, not production code for the Alpha processor - after all, there was no such thing, didn't everyone run Pentiums?
The tech support person convinced him that the stuff labeled Intel was production code, but she never could convince him that the Alpha software was anything but test releases... --
Don't forget OpenVMS, either; it still has a good market share in places where no downtime is acceptable.
As much as I love *NIX, eliminating themselves from most of the market share
NT didn't sell the most Alphas. Linux did. It isn't even close. There just aren't that many apps compiled natively for the Alpha on NT, and running an x86 binary on the Alpha didn't get you all that much. --
Slashdot Mirror
Can you? What makes you so sure?
I'm one of those who undoubtedly would have been reported to the folks running this program. In my days in high school, I was a full year younger than the next youngest person in my class. I was smaller, lighter, weaker, and slower than nearly all of my peers. I was also the one they turned to in computer math class (in the days when computer math was taught via teletypes talking to a mainframe halfway acros town) - but no other time. The rest of the time, I was bullied, teased, beaten, and ostracized. As it is, it took a year of psychotherapy and antidepressants, 20 years later, to begin to reverse the effects of my high school experience. I'm still getting over it. I'm now 39 years old, about to turn 40, and only in the past couple of years have I been anything approaching happy.
I know what the kids who commit school violence are going through. I was there myself. The first thought that popped into my head when I heard about Columbine was "That could have been me...".
Those kids don't need others turning them in anonymously, with or without reward, so they can be further singled out for ostracism and abuse at the hands of an uncaring society. They need HELP . They need to know that someone cares about them. They need understanding. They need acceptance.
I'm sorry you can't accept them, whether in the name of your kids' safety or otherwise. The simple truth, though, is that this is exactly the wrong approach. Reach out to these kids. Don't turn them in, or away, or out...
I wouldn't wish any parent to have to deal with having their kid murdered. The solution, as someone wrote in reply to Katz' original article, is to have schools be places no kid would dream of blowing up. If we can accomplish that, then no parent need have that fear again.
--
Those are trolling words, troll.
Only to one who hews to the Slashdot orthodoxy. The rest of us recognize it as a valid question, and one that needs to be asked in this discussion.
--
The best thing SCO could do right now is to free the trademark "Unix"
SCO doesn't own the trademark, the X/Open Consortium does. Novell donated it to them several years ago. X/Open will license it to any system that can pass the Spec 1170 suite of compatibility tests. That's how DEC OSF/1 became Digital Unix.
--
For various reasons, I need to get myself registered and travel booked and so on for a Linux show this year. I'd been planning to go to Linux Expo, but with that one out of the picture, what should I go to instead? I missed Linux World in NYC, and Atlanta is too far in the future - if I pick that one, the money for the trip is likely to dry up before I get to go. What's a good Linux show to go to instead? preferably a week-long event with some real meat to it.
--
Amazon has a fiduciary duty to its shareholders to protect its corporate assets. Those include patentable ideas.
Further, they also have a fiduciary duty to their shareholders to guarantee their corporate value does not get diminished by others.
Put this together, and Amazon did exactly what they are required by law to do: Suing B&N for patent infringement. If they didn't, Bezos and the rest of the Amazon board could and probably would have been sued by their shareholders.
Companies have a legal duty to protect themselves, and the legal system is an excellent example of "the best defense is a good offense".
Of course, in RMS' communist utopia, such things wouldn't exist, but this is the real world...
I think I'll go buy a book from Amazon in RMS' honor.
--
Have you ever tried QT yourself?
No, I haven't. I guess the central question is: How much time do you spend writing code that does the real work of your application, and how much time do you spend writing code that handles the mechanics of the user interface? My experience with Delphi on the payroll project is that nearly 95% of my time is spent on writing the application, not the mechanics. Can Qt achieve that?
--
I think the word RAD is a buzzword. Sure, you can place out a button a few seconds faster in VB than QT, but how much time of the overall project-time does that account for? 2% ?
The real beauty of a RAD environment such as Delphi is that it removes the need to explicitly code the standard boilerplate that has to be done every time you add a user interface control to an application, and the message handlers, and control blocks, and on and on and on...Anyone who's written a Windows (or OS/2 PM, or X, or other stuff) app the old-fashioned way will be amazed at how much time you spend working on the problem to be solved, rather than in the tedious low-level stuff, in a RAD environment.
--
Huh?! Unless you're seriously advocating that a 300K program-in-development that doesn't do anything approaching what the user wants is somehow better than a 1.5 MB program that does, you missed my point entirely. My point is more about development leadtimes than about code size.
I'm estimating that the same program, coded with vi and make and done the old-fashioned way, would come out somewhere around a megabyte of finished executable. Okkay, the Delphi program is 50% bigger. Big, fat, hairy deal. The version coded the old-fashioned way would take about 10 times as long to develop and debug. In the world of business, where people expect their computers to actually do useful work, that tradeoff is a no-brainer.
Without a RAD tool, Linux will never make it out of the server room and onto people's desktops, for it's completely unsuitable as an applications platform. People following the One True Linux Way as you advocate it will get run over by others who beat them to market with bigger, slower, but running code.
I care as much about Linux as you claim to. However, to me, Linux has the potential to replace Windows in many more roles than just Web and file servers. To compete in those spaces, it must have comparable capabilities, and Delphi brings a very important set of those capabilities to the Linux world. Market acceptance is much, much more important then ideological purity if the goal is to have Linux be accepted as widely as Windows.
--
Who would want to code using the Delphi or C++ Builder environments? Even for a Windows application their interface sucks, and the underlying code base isn't much better. It supports a vast, bloated and confused class structure which encourages the creation of slow, windy programs. Why do we need software like this when we have tools like vi and make already part of Linux?
Have you ever developed a serious application with Delphi? I have and am, a large, very customized payroll system for a company with extremely nonstandard payroll requirements. (Their chart of accounts is over 70,000 long, and 95% of that is payroll for 450 employees.) It would have taken me 10 times as long to develop this program in a non-RAD environment. Yes, the program is larger than it would be if I'd used more traditional development tools...but I'd still be developing early functionality, instead of getting ready to hand them a feature-complete version. What's better, a program that they can use that has a 1.5 MB load module, or a 300K load module that they can't use to get real work done?
--
Not so. Compaq did a clean-room implementation of the PC BIOS, instead of using the published code. Several cloners who used the published code were sued by IBM (anyone remember Victor?) and faded away, and Compaq withstood its suit and thrived.
(Disclaimer: Yes, I work for Compaq, but that all happened long, long before I got here.)
--
I run computers to do neat things on, not necessarily OS-based. What neat stuff can I do with HURD?
--
There's no reason you can't use GPL code in a BSD kernel.
No, there isn't, if you're willing to have the GPV infect the BSD kernel. The BSD developers aren't, because they want BSD to remain truly free.
--
*ZOT*
You owe the Oracle a Tux plush toy.
--
With the big computer companies such as Compaq embracing Linux, what is it that differentiates VA Linux Systems? Why should a company buy one of your boxes instead of one of theirs?
--
Hey, that's "you ignorant splut!"...
You're right about the debating style. I should take the time to be more detailed about it. I should do lots of other things, too, that I don't...
Even so, I can't help thinking that, had I posted the same style message slamming the NetBSD Foundation as unsuitable to receive this award, it would not have been moderated down three times, and probably not at all. I consider this more evidence that the Most Holy Stallman, Pope of the Only True Free Software Movement (say hallelujah!), is considered untouchable by the vast majority of people on this forum.
But if you're hooked on karma, you have to be careful how you say what you say.
I dunno if I'd consider myself hooked on karma, but having one post zap me below the +1 threshold was a bit discouraging. I like to think of Linux people as more hackish than most, and therefore more likely to be accepting of opinions other than their own; being proven wrong time and again wears on the nerves.
--
Of course, doing anything but following the Slashdot sheep in bowing and scraping before the revered RMS has clobbered my karma, as I knew it would...
--
(In fact, in the US it's silly to talk about "licensing" forms of communication, as all of them should be covered by the First Amendment, but that's another argument for another day.)
The First Amendment covers content, not manner of speech. You can say anything you want, but the government has a right to regulate how and when and where in order to keep the peace.
--
Just like you hoped the no-code Tech would, too? Not to mention bringing in a flood of technically competent folks to revolutionize the state of radio communications, all on a shoestring budget in the ham bands, results published in QST and CQ. Right.
The simple truth of the matter is that the no-code license has failed in its stated objective. The average age of hams has risen at one year per year of elapsed time, and that rate is unchanged before and after the advent of the no-code license. As I've contended all along, the real problem is that ham radio is not attractive to kids today. Your campaign is 10 years too late, and ham radio will suffer for it.
One of the best things about this decision is that it ends a very ugly acrimonious situation in ham radio that has persisted since 1990, when the no-code VHF license was introduced as the first foot in the door for modernization of ham radio. A lot of the older hams alienated the younger ones because they felt that no-coders weren't real hams. Now, those younger hams will have the same licenses as the older ones, and will be in their faces on the HF bands.
Oh, so now the FCC can change human nature with a simple Report and Order? Those of us who had to work to earn our privileges are going to be as resentful of those who do not under the new scheme as was the case with the no-code Tech. This change will, if anything, perpetuate the alienation and division, and spread it to the HF bands as well. Many older hams have resented, and IMAO rightfully so, the dumbing down of the entrance requirements in the name of attracting the "right people" into the service, especially given the 10 years of failure to do anything of the sort that we've seen.
The real solution, as I have been arguing ever since the no-code argument started, is to convince people that ham radio is worth the entrance requirements. No matter how low you set the barrier, the real problem is not getting people past it - the widely diverse ham population before 1990 is all the proof anyone can ever need that desire will conquer all - but getting people to want to climb over it. We should have tried better marketing in 1990, instead of assuming that people were too dumb any more to pass the requirements. This change can only continue the slide toward chaos and lawlessness that we've seen - case in point, the 147.435 LA repeater, a blight on the face of the service if there ever was one - ever since the requirements were first lowered.
Jay Maynard, K5ZC
--
I'm one of those (and probably the only one who will post on slashdot for attribution) who thinks dumbing down the code requirement is a Bad Idea. I expect to get flamed for it, but this is one issue I've given a lot of thought to over the last 20 years or so, and don't expect to be convinced otherwise.
Ham radio is not only about technical innovation. Yes, that's one purpose, and one that I, too, believe is important. It's not the only purpose, however. The basis and purpose of the service, as expressed in 97.1, includes fostering international goodwill and providing emergency communications when needed. It takes more than a technogeek to accomplish these ends. One must be able to *communicate*. The dual nature of the ham tests serve, IMAO, to point that out: you can answer all the theory questions they can throw at you and still not be an effective operator. The code test provided some way to check that. To get the Extra ticket, you had to be a well-rounded ham, both a skilled operator and at least fairly knowledgeable technically.
The FCC has goofed here, aided by public pressure from well-meaning techies who can't see that building the technology isn't enough: you have to make use of it, too, to provide communications.
Jay Maynard, K5ZC (licensed as Amateur Extra in 1977)
--
On a somewhat similar topic, does anyone know what happened to Tri-Teal's code? They went out of business mid-'99 and they made a CDE for Linux
Good question, especially since SCO UnixWare also use[sd] it...
--
Why the hell anyone would want to sling pointers around as though they were integers escapes me, anyway; that's a recipe for disastrous, hell-to-find bugs. There's ALWAYS a better way to get there.
--
It's worth noting that the Pentium Pro/II/III have a 48-bit segmented addressing mode, allowing physical memory beyond 4GB. Nobody uses this yet, but it's in there.
Actually, SCO UnixWare 7.1.0 (at least) does. I have a Compaq ProLiant 8500 with 8 PIII/500s and 5 GB of RAM on my test bench, and UW will make use of all of it.
(Disclaimer: I work for Compaq. I don't speak for them.)
--
Is it a recommended choice for someone only running linux? I mean, do I get twice the performance for the same overall price?
Depends on what you're doing. If you're doing number crunching, it's definitely worth it; if you're doing networking, it's not a bad way to go, as the script kiddiez don't have Alpha exploits to try to run on your system (at least not yet); if you're looking for a general user box, it's about neck-and-neck. My Alpha 21164/500 gets over twice the CSC key rate of my PIII/550. For other things, it feels about as snappy.
--
A friend, who works as a tech support manager at a major computer software company which shall remain nameless, tells me that one of her people got a call one day, complaining that they had untested software on their web site for download...
...seems he saw the Alpha versions and thought they were alpha test releases, not production code for the Alpha processor - after all, there was no such thing, didn't everyone run Pentiums?
The tech support person convinced him that the stuff labeled Intel was production code, but she never could convince him that the Alpha software was anything but test releases...
--
Don't forget OpenVMS, either; it still has a good market share in places where no downtime is acceptable.
As much as I love *NIX, eliminating themselves from most of the market share
NT didn't sell the most Alphas. Linux did. It isn't even close. There just aren't that many apps compiled natively for the Alpha on NT, and running an x86 binary on the Alpha didn't get you all that much.
--