My wife and I got hooked on B5 when we were already watching all the Star Trek shows, and we agreed that after B5 started, some ass-kicking must have gone on over at Paramount because all of the sudden, the so-so ST:DS9 and the rapidly declining ST:Voyager suddenly became a hell of a lot better. We were getting ready to write off Voyager. In any case, we REALLY liked B5 and were quite impressed with Crusade, thinking that it in many ways was an improvement over B5. B5 had some truly distinctive characters and damn fine drama as well as some nice comic moments too.
It is absolutely shocking that a respected private security company would come up with such a wrongheaded, misanthropic, and insulting idea.
I'm glad they're reconsidering the hats, T-shirts, and cash idea, because let me tell you in no uncertain terms, the FIRST kid who shows up at the mall with one of those caps or T-shirts on is likely to find one or more of the following soon thereafter: 1) dog excrement in bookbag 2) prank calls at all hours 3) phone number on men's room stall wall under the words "FOR BJ CALL" 4) air let out of tires 5) his/herself dateless in perpetuity 6) his/herself getting the crap beat out of him/her if his/her reporting actually turns out to be accurate. WHAT THE HELL ARE YOU TRYING TO DO, PINKERTON??? I know of REAL KIDS who HAVE BEEN KILLED for less than that!!! How could you have come up with that idea in the first place??
The idea that kids have the experience and knowledge to predict and identify which kids are the most likely to commit acts of violence is throughly preposterous. That Pinkerton would even entertain the notion forever damages the reputation of the company, at least in my eyes. You would think that a professional security company would have better minds on the payroll than that.
I don't recommend this, even though theoretically it should work. I found that machines that are supposed to be the same often aren't. There may be a different rev level of chipset and a bunch of other little things. Most probably won't cause a train wreck, but I bet some will, in which case the cause will be far from easy to determine. One time I got a couple of shipments of Dells back-to-back and sure enough, some of the machines in the 2nd shipment had a slight difference in CD-ROM - probably wouldn't have added up to much, but considering that Linux is pickier about hardware than NT and NT is a LOT pickier than Win98/95/etc., I wouldn't rely on the disk-image trick for "cloning" machines.
I have DOD and DOE experience (12 yrs total) and I would be wondering the same thing. I would first ask to see, on paper, as part of a policy statement or regulation that has gone through the customary chain of approvals for such things, where it is that C2 compliance is a requirement for the application that Linux is in consideration for. If no one can produce such a document, then you win (at least morally). You also win (again, morally if in no other way) if the document *IS* produced but you find widespread non-compliance. I know you can make NT C2 compliant but the laundry list of things you have to do doesn't leave you with a terribly useful machine IMHO.
I've noticed that in Turbolinux' white-paper on a theoretical (but real-sounding) high-availability Web site, they make no bones about it - they recommend Oracle and OPS on clustered Suns for the DB backend.
I would be a lot happier if these guys would use their self-evident capability to produce a PC-like "thin client," even if it were diskless. My mental model would be along the lines of the first Monorail PCs. As an IT manager, I'm always looking for ways to compress the per-user costs of deployment, and if Netpliance would make something that would help me do that, I would be seriously tempted to buy it - in quantity!
I'm not sure if "Beowulf extensions" to Apache are appropriate or necessary. Once can achieve Web server scalability in the manner you describe with something like TurboLinux TurboCluster Server, where a system designated as a traffic manager points incoming Web requests to a farm of Web servers, each running Apache. Slashdot does the same thing but using different products. Of course, you have to put some effort into synchronizing the Web content, either by having all the servers mount common NFS volumes or by some other mechanism.
I have the opposite problem - Mandrake is compiled for a Pentium but I'm trying to get it onto a 486. Don't even tell me about cross-compiling on another machine; that would take a little more chisenbop than I've got and I don't have the disk space anyway. What's worse is that the 486 in question has an EISA-bus 53c710 SCSI as its only disk controller, and that's not supported until 2.2.13 IIRC. Couple all that with a proxy server at work that doesn't let me use a real FTP client (HTTP only) and I'm pretty well screwed, glued, and tattooed. -SN
I have about 5 years of NT experience. I started using it to solve real-world business problems before a huge portion of the IT community even knew it existed. At my place of work at the time (a field office of a cabinet-level Government agency) I became the leading proponent of NT deployment for both the server and workstation space. I was interested in using NT on Intel to run mainframes off the site. An NT system architecture I devised for a particular business need became the model for the entire Federal Government.
I offer the above information in hopes of proving that I am not a knee-jerk MS or NT basher.
Having said that, I feel that MS did a remarkable thing in developing NT, but I do NOT believe that all of NT (or even most of it, for that matter) represents an MS innovation. IIRC, NT and OS/2 are both results of work performed my MS and IBM in partnership, and when that partnership dissolved (less than amicably, I understand), the work done up to that point was shared by both partners. OS/2 and NT both can only claim mixed pedigree, and one of the things that you cite as one of the significant innovations NT represents - portability across processors (as embodied in the NT Hardware Abstraction Layer) - would have had to have been given birth to during the most active period of that MS/IBM partnership.
I want to insert here that as exciting as the HAL concept was for the future of NT (I was certainly excited about the prospect of running NT on any machine alive), we can see how that turned out. MS initially released NT to run on Intel x86, DEC Alpha, and MIPS CPUs, and the PowerPC CPU was added later. Well, MIPS and then PPC were dropped and I feel that Alpha might be the next to go. Development of a HAL is FAR from trivial and pretty much cannot get off the ground for any CPU without INTENSE MS cooperation. My belief is that regardless of any original intent otherwise, NT portability across CPUs became used as a competitive weapon: "I don't want your CPU/platform to flourish; I hereby revoke NT availability for you!"
Digital's VMS had pre-emptive multitasking and virtual memory starting in 1978 (meaning no disrepect to any kind of UNIX). The Amiga had pre-emptive multitasking with a GUI as part of the operating system in 1985. And, IIRC, MS Windows arose directly from MS' encounter with the Apple Lisa, which in turn got its impetus from prototype work done by Xerox (which Xerox did not deem worthy of further development). You can't paint IBM as being derivative for boosting OS/2 GUI without getting some of that same paint on MS.
So, while I admire the work MS did with NT post-IBM-partnership, I cannot accept your view of the degree to which NT represents genuine innovation onthe part of MS - and I certainly don't admire how difficult and expensive MS has made NT to work with and actually utilize. That, more than anything else, was what initially drove me to take an interest in Linux.
I did not expect there to be horrendous Y2K technology collapse, so no surprise there.
What I did expect (but thankfully did not get) was some sparse but very significant terrorism within the USA. Knowing that US officials had made arrests in the weeks prior to 1/1/00 of people crossing the border into the USA with explosives and/or explosive components, I felt that there ought to be some number of like-minded individuals NOT caught. The Centennial Olympic Park and Ok City bombings also weighed on my mind.
That there was not mass mayhem at the big flipover led me to recast my perceptions. Is it known that the people that the Feds arrested were actually planning for something on 1/1/00 as opposed to some other time? It could be that bombers are entering the country routinely and that heightened security got them caught leading up to 1/1/00.
The alternatives would appear to be: 1) The small number of people reported by the media as being arrested by the authorities were all the people setting out to cause big trouble 2) Many more were arrested that we don't know about 3) 1/1/00 was simply not a terrorism target in the first place (leaving open the possibility that there are plenty of others)
I can't speak to the relative or absolute likelihood of any of these alternatives.
Slashdot Mirror
My wife and I got hooked on B5 when we were already watching all the Star Trek shows, and we agreed that after B5 started, some ass-kicking must have gone on over at Paramount because all of the sudden, the so-so ST:DS9 and the rapidly declining ST:Voyager suddenly became a hell of a lot better. We were getting ready to write off Voyager. In any case, we REALLY liked B5 and were quite impressed with Crusade, thinking that it in many ways was an improvement over B5. B5 had some truly distinctive characters and damn fine drama as well as some nice comic moments too.
It is absolutely shocking that a respected private security company would come up with such a wrongheaded, misanthropic, and insulting idea.
I'm glad they're reconsidering the hats, T-shirts, and cash idea, because let me tell you in no uncertain terms, the FIRST kid who shows up at the mall with one of those caps or T-shirts on is likely to find one or more of the following soon thereafter: 1) dog excrement in bookbag 2) prank calls at all hours 3) phone number on men's room stall wall under the words "FOR BJ CALL" 4) air let out of tires 5) his/herself dateless in perpetuity 6) his/herself getting the crap beat out of him/her if his/her reporting actually turns out to be accurate. WHAT THE HELL ARE YOU TRYING TO DO, PINKERTON??? I know of REAL KIDS who HAVE BEEN KILLED for less than that!!! How could you have come up with that idea in the first place??
The idea that kids have the experience and knowledge to predict and identify which kids are the most likely to commit acts of violence is throughly preposterous. That Pinkerton would even entertain the notion forever damages the reputation of the company, at least in my eyes. You would think that a professional security company would have better minds on the payroll than that.
I don't recommend this, even though theoretically it should work. I found that machines that are supposed to be the same often aren't. There may be a different rev level of chipset and a bunch of other little things. Most probably won't cause a train wreck, but I bet some will, in which case the cause will be far from easy to determine. One time I got a couple of shipments of Dells back-to-back and sure enough, some of the machines in the 2nd shipment had a slight difference in CD-ROM - probably wouldn't have added up to much, but considering that Linux is pickier about hardware than NT and NT is a LOT pickier than Win98/95/etc., I wouldn't rely on the disk-image trick for "cloning" machines.
I have DOD and DOE experience (12 yrs total) and I would be wondering the same thing. I would first ask to see, on paper, as part of a policy statement or regulation that has gone through the customary chain of approvals for such things, where it is that C2 compliance is a requirement for the application that Linux is in consideration for. If no one can produce such a document, then you win (at least morally). You also win (again, morally if in no other way) if the document *IS* produced but you find widespread non-compliance. I know you can make NT C2 compliant but the laundry list of things you have to do doesn't leave you with a terribly useful machine IMHO.
I've noticed that in Turbolinux' white-paper on a theoretical (but real-sounding) high-availability Web site, they make no bones about it - they recommend Oracle and OPS on clustered Suns for the DB backend.
I would be a lot happier if these guys would use their self-evident capability to produce a PC-like "thin client," even if it were diskless. My mental model would be along the lines of the first Monorail PCs. As an IT manager, I'm always looking for ways to compress the per-user costs of deployment, and if Netpliance would make something that would help me do that, I would be seriously tempted to buy it - in quantity!
I'm not sure if "Beowulf extensions" to Apache are appropriate or necessary. Once can achieve Web server scalability in the manner you describe with something like TurboLinux TurboCluster Server, where a system designated as a traffic manager points incoming Web requests to a farm of Web servers, each running Apache. Slashdot does the same thing but using different products. Of course, you have to put some effort into synchronizing the Web content, either by having all the servers mount common NFS volumes or by some other mechanism.
I have the opposite problem - Mandrake is compiled for a Pentium but I'm trying to get it onto a 486. Don't even tell me about cross-compiling on another machine; that would take a little more chisenbop than I've got and I don't have the disk space anyway. What's worse is that the 486 in question has an EISA-bus 53c710 SCSI as its only disk controller, and that's not supported until 2.2.13 IIRC. Couple all that with a proxy server at work that doesn't let me use a real FTP client (HTTP only) and I'm pretty well screwed, glued, and tattooed. -SN
I have about 5 years of NT experience. I started using it to solve real-world business problems before a huge portion of the IT community even knew it existed. At my place of work at the time (a field office of a cabinet-level Government agency) I became the leading proponent of NT deployment for both the server and workstation space. I was interested in using NT on Intel to run mainframes off the site. An NT system architecture I devised for a particular business need became the model for the entire Federal Government.
I offer the above information in hopes of proving that I am not a knee-jerk MS or NT basher.
Having said that, I feel that MS did a remarkable thing in developing NT, but I do NOT believe that all of NT (or even most of it, for that matter) represents an MS innovation. IIRC, NT and OS/2 are both results of work performed my MS and IBM in partnership, and when that partnership dissolved (less than amicably, I understand), the work done up to that point was shared by both partners. OS/2 and NT both can only claim mixed pedigree, and one of the things that you cite as one of the significant innovations NT represents - portability across processors (as embodied in the NT Hardware Abstraction Layer) - would have had to have been given birth to during the most active period of that MS/IBM partnership.
I want to insert here that as exciting as the HAL concept was for the future of NT (I was certainly excited about the prospect of running NT on any machine alive), we can see how that turned out. MS initially released NT to run on Intel x86, DEC Alpha, and MIPS CPUs, and the PowerPC CPU was added later. Well, MIPS and then PPC were dropped and I feel that Alpha might be the next to go. Development of a HAL is FAR from trivial and pretty much cannot get off the ground for any CPU without INTENSE MS cooperation. My belief is that regardless of any original intent otherwise, NT portability across CPUs became used as a competitive weapon: "I don't want your CPU/platform to flourish; I hereby revoke NT availability for you!"
Digital's VMS had pre-emptive multitasking and virtual memory starting in 1978 (meaning no disrepect to any kind of UNIX). The Amiga had pre-emptive multitasking with a GUI as part of the operating system in 1985. And, IIRC, MS Windows arose directly from MS' encounter with the Apple Lisa, which in turn got its impetus from prototype work done by Xerox (which Xerox did not deem worthy of further development). You can't paint IBM as being derivative for boosting OS/2 GUI without getting some of that same paint on MS.
So, while I admire the work MS did with NT post-IBM-partnership, I cannot accept your view of the degree to which NT represents genuine innovation onthe part of MS - and I certainly don't admire how difficult and expensive MS has made NT to work with and actually utilize. That, more than anything else, was what initially drove me to take an interest in Linux.
I did not expect there to be horrendous Y2K technology collapse, so no surprise there.
What I did expect (but thankfully did not get) was some sparse but very significant terrorism within the USA. Knowing that US officials had made arrests in the weeks prior to 1/1/00 of people crossing the border into the USA with explosives and/or explosive components, I felt that there ought to be some number of like-minded individuals NOT caught. The Centennial Olympic Park and Ok City bombings also weighed on my mind.
That there was not mass mayhem at the big flipover led me to recast my perceptions. Is it known that the people that the Feds arrested were actually planning for something on 1/1/00 as opposed to some other time? It could be that bombers are entering the country routinely and that heightened security got them caught leading up to 1/1/00.
The alternatives would appear to be:
1) The small number of people reported by the media as being arrested by the authorities were all the people setting out to cause big trouble
2) Many more were arrested that we don't know about
3) 1/1/00 was simply not a terrorism target in the first place (leaving open the possibility that there are plenty of others)
I can't speak to the relative or absolute likelihood of any of these alternatives.