As I see it, this all looks like a lofty annoucement with little actual funding increases to back it up. It only sounds like a lot, because it is spread over a long period of time. It'll just do what always happens... start some promising new projects, then cause them to get canceled due to funding before truely getting off the ground.
If he wants my vote, he'll have to do a lot better than this. Doubling NASA's "annual" budget might actuall get my attention. Honestly, we always talk about NASA being short sighted and having no long term goals, but ever wonder why? It's probably due to the simple fact that they don't have enough money to achieve any of their goals. I seriously think they're in a bit of a catch-22 on issues like the shuttle. Too expensive to operate, so it saps their budget. But also too expensive to replace, so none of those projects seem to get anywhere. (and political meddling in their affairs, etc.)
I seriously hope that someone wins the X-Prize soon, and opens the door for serious commercial space programs. It's probably the only thing that'll really get us moving, and eventually fix the #1 problem... getting stuff from ground to LEO.
I was really annoyed by having that situation myself. That's why I love now having a TiVo so much. I can watch TV when I want to, not when a show is on, so I never have to even think of planning my life around TV.
>Here, however, you seem to be confusing the function of a NAT with the function of a firewall.
I totally agree, and this is something that annoys me quite a bit. For the longest time, way too many people made the assumption that: Firewall == NAT system
Now, due to the prevalence of those cheap Linksys/Netgear/Belkin/D-Link boxes, people now seem to assume: Router == NAT system (though this one is a tat OT)
In any case, the purpose of a firewall is to protect your network by filtering traffic, and there is NOTHING that says you can't do this without a NAT. You can get the same port-blocking effects there by default on NAT with a simple default-deny policy on incoming traffic with a normal packet-filtering firewall. The only thing NAT actually gains you is to obscure the IP of the originating host, and that actually causes more problems that it likely solves.
Actually, the government in the US is already planning IPv6 migration, and there are mandates for the DoD to go to IPv6 by 2008. Sure, that's a few years off, but it means that in the mean time there will be many pilot programs and gradual migrations. It is going to happen, and even if the corporate world lags, the gov't will be pusing it.
Yeah, I've noticed that as well. In my college math classes, calculators were not allowed at all. However, the classes were designed so you really didn't need one anyways.
In other classes where you dealt with numbers, they were allowed. In fact, some classes even allowed a full-blown laptop with programs like Maple and Matlab.
The truth is that it really depends on what you're actually trying to get done in the class. If the goal of the class is learning the math, then no calculators. If the goal of the class is to learn how to apply the math to a bigger problem, then use calculators because it isn't the math itself they're testing you on.
The best solution for this I've found so far, is a program called OpenVPN. The thing is easy to configure, runs on most 'nix platforms, tunnels over UDP (I'm sorry, but IP over TCP is stupid), and works wonderfuly.
Sure, binary compatability is all nice and stuff, but only works if the binary in question, and the machine you're running it on, both use the same processor architecture.
As AIX only runs on RS/6000 (pSeries) machines, which use POWER and PowerPC processors, for any binary compatability to work, you'd need binaries from Linux/PPC.
What's really nice, though, is binary compatability between open source OSes. For example, FreeBSD has excellent Linux binary compatability. It's really easy too, as it just implements a system call translaton layer, and keeps a collection of the Linux libs off in a directory as needed. In fact, it's just as fast as running the same binary under Linux (some might argue faster).
I've also seen kernel options for much wilder binary compatability features to run binaries from commercian 'nixes on OpenBSD/NetBSD, though I've never really looked into them, or used them.
However, I'm wondering, does Linux have much in the way of these capabilities in and of itself? (I remember this "ibcs" thing many years ago that could run SCO binaries, but havn't really seen anything since)
It's not really just an x86 vs Apple thing, but rather an x86 vs "Everyone Else".
Around the 1990'ish (give or take), lots of companies who had been using antiquated CISC architectures decided to break compatability and move up to the world of clean RISC architectures that were easy to make the 32-to-64-bit jump with, and not break software compatability.
For example: Apple: M680XX -> PowerPC Sun: M680XX -> SPARC DEC: VAX -> Alpha (64-bit from the start) HP: M680XX -> PA-RISC
Yup, I definitely pushed the limits... Thankfully they revamped the electrical system in the dorm I lived in before I moved in there. I'm one of the few people who actually took the time to map out the outlet-to-breaker-circuit mapping for the room. Hey, I had to make sure the refrigerator/microwave used a different circuit than my computers. Before I did that, I found that if I had all my monitors on (including one that was normally off), and my roommate microwaved something, the breaker tripped.
Of course, I also push the limits on how much power one uses for computer equipment. The height of my computer power draw was probably my sophomore year. At the end of that year, I borrowed an ammeter and took some measurements. One favorable thing I discovered was that most computers generally use nowhere near as much power as their power supplies are rated for. However, I still managed to max out my load at 14.5A (approx 1740W) continuous draw. Thank goodness I wasn't paying directly for that off the meter;-)
Of course in those days I was running like 8 computers and 3 monitors. Now that I pay for my power, I have cut down a bit on the amount of 24/7 stuff I'm running, though I still manage to use quite a bit of electricity. Then again, now that I live in Florida, I'll bet that a lot of my juice goes too power the A/C.
Hey, on a machine like that you should be running AIX, not Linux. Then again, if you insist on only wanting to run Linux on that fine RS/6000, I'll gladly take it off your hands and load a proper OS for the machine.
Hehe... In college, I was in a club where we managed to pick up and use old computer hardware. One class of machines I played with a lot were some Sun 3/280s. The one I worked with the most had SunOS 4.1, a 68020 processor, 32MB of RAM, a 600MB disk, and an 800MB disk. (these disks took 2 people to lift)
The impressive part was that I actually got gcc to compile, then built ssh and apache. SSH actually worked decently on the machine as well:) (commercial SSH, which has 68k asm optimizations in the source, unlike OpenSSH)
Unfortunately, I think the machines aren't running any more, as their space has been taken over by a DEC-purist, and turned into VAX-country:) (but they're probably still there, being poked at every once in a while) At least these are the "real man's VAXen", complete with 3-phase power.
The oldest computer currently running in my setup would have to be my console server. It's an IBM RS/6000 POWERstation 250. (PowerPC 601 at 66MHz, about 80MB of RAM, and a 400MB disk) Still runs just fine. However, it's getting replaced sometime soon by an embedded console server of sorts, simply because it's a cleaner solution.
I agree. However, as the pundits in the Sun groups on Usenet keep pointing out, Apple still doesn't put ECC memory in these things. If the dual G5 had ECC memory, they'd probably shut up, and might even come down off their high horses and buy 'em.
Actually, from observations, I've noticed that the Intel 80186 is IBM's embedded I/O processor of choice for most of the microchannel cards for RS/6000 systems that I've seen.
Ok, that statement is just plain wrong, unless you're comparing something other than CPUs.
First, that Indigo2 is not "plain-jane IDE" (unless you're using some weird adapter board), but rather "plain-jane SCSI-2" (10MB/s).
Second, one big factor you notice when comparing CPUs, especially when some are "budget models" is that magic thing known as cache. Ever wonder what feature they're cutting to lower cost? I'll bet the R4400 has plenty of cache, while the Duron cuts cache (so does the Celeron, and some of Sun's older and slower microSPARC CPUs)
Third, even with those factors, there's no way in hell that the MIPS R4400 (at 120MHz) CPU could ever come close to touching the performance of an AMD Duron (750MHz). You have to be comparing graphics cards.
Now, one of the features of the Indigo2 that you might be using, is the "Impact" line of graphics cards. The Solid (no texturing) and High Impact (texturing) versions have about 450 MFLOPS of performance on the card itself, and the Max Impact has double that. I will believe that your Indigo2 whoops the crap out of the Duron on graphics, if you're comparing one of those fine GIO64 graphics cards to some POS card you threw in the PC.
But I will NOT believe you're comparing CPU performance.
How do I know this? Well, let's just say I've got an R10000 (195MHz) SGI Indigo2 High Impact sitting next to me.
Yes, I'm using a Sun workstation right now to post this comment. (and yes, my new laptop is a PowerBook) In fact, I recently deployed some cheap Sun Ray thin clients around my apartment, which makes for really nice hot-desking around the place.
It's funny how people keep knocking XFree86 for performance... What ever happened to those days back in 1995, when XF86 (when using the accelerated server on an ATI Mach64) was actually FASTER than Win95? (yes, that was on a friend's P-100, which was brand-spanking new)
Actually, there is a real operating system out there, called "P/OS". It was the DEC "Professional Operating System", and I think it ran on their "Rainbow" series of microcomputers. It's CLI is VMS-style (I guess that means DCL-style). I first saw it running on a "VAX Console" (rebadged PDP-11 based microcomputer cabled to a "VAX 8530").
The only possible form of terrorist attack would have been pre-launch sabotage of the craft. Then again, given all the checks and rechecks, they'd probably have to infiltrate most of the shuttle inspectors. Basically, it's not realistically possible.
I think that's because at RPI it's well understood that the H/SS classes are just "requirements to get out the way" and generally aren't serious and important classes for the vast majority of the students. So, they try to make many of them interesting, but rarely academically challenging.
Speaking as a very recent RPI graduate, I think I only ever got 1 B among all my H/SS coursework.
Eh... The Creator3D is really just one of those "we support 3D, but don't buy this card if you care about 3D" board. Sun's overpriced number for the card is around $715, but they can be had easily on eBay for thr $100-170 range. They are pretty nice 2D cards though, and people like 'em when they want to run at high-res.
Now as far as 3D performance, I wouldn't be the least bit surprised if the High Impact board in my SGI Indigo2 workstation blows the crap out of it. Oh, and the I2 I have is from '96, but the copyright data on the card is '95. (and it's not even the highest-end option for the machine) Oh, and that card also supports stereo viewing.
Ahh, reminds me of my days around the E-Club at dear old RPI... (ok, this was just the past few years) the club website
It all started when the school threw out a VAX 8530... Thankfully, we have 3-phase power available to run the thing, and it's now happily running VMS 5.5-2 in all it's massive glory. Over the years we've also accumulated a VAX 11/780 (dead unfortunately), a PDP-11/45 (which one club member had to rebuild the power supply for), a pair of Sun 3/280s (complete with 12" platter hard drives), along with various other "smaller" machines that might be to new to be considered classic. (some MicroVAX-class machines, a bunch of old model IBM RS/6000s, some HP9000 stuff, etc, etc.)
Slashdot Mirror
As I see it, this all looks like a lofty annoucement with little actual funding increases to back it up. It only sounds like a lot, because it is spread over a long period of time. It'll just do what always happens... start some promising new projects, then cause them to get canceled due to funding before truely getting off the ground.
If he wants my vote, he'll have to do a lot better than this. Doubling NASA's "annual" budget might actuall get my attention. Honestly, we always talk about NASA being short sighted and having no long term goals, but ever wonder why? It's probably due to the simple fact that they don't have enough money to achieve any of their goals. I seriously think they're in a bit of a catch-22 on issues like the shuttle. Too expensive to operate, so it saps their budget. But also too expensive to replace, so none of those projects seem to get anywhere. (and political meddling in their affairs, etc.)
I seriously hope that someone wins the X-Prize soon, and opens the door for serious commercial space programs. It's probably the only thing that'll really get us moving, and eventually fix the #1 problem... getting stuff from ground to LEO.
I was really annoyed by having that situation myself. That's why I love now having a TiVo so much. I can watch TV when I want to, not when a show is on, so I never have to even think of planning my life around TV.
>Here, however, you seem to be confusing the function of a NAT with the function of a firewall.
I totally agree, and this is something that annoys me quite a bit. For the longest time, way too many people made the assumption that:
Firewall == NAT system
Now, due to the prevalence of those cheap Linksys/Netgear/Belkin/D-Link boxes, people now seem to assume:
Router == NAT system
(though this one is a tat OT)
In any case, the purpose of a firewall is to protect your network by filtering traffic, and there is NOTHING that says you can't do this without a NAT. You can get the same port-blocking effects there by default on NAT with a simple default-deny policy on incoming traffic with a normal packet-filtering firewall. The only thing NAT actually gains you is to obscure the IP of the originating host, and that actually causes more problems that it likely solves.
Actually, the government in the US is already planning IPv6 migration, and there are mandates for the DoD to go to IPv6 by 2008. Sure, that's a few years off, but it means that in the mean time there will be many pilot programs and gradual migrations. It is going to happen, and even if the corporate world lags, the gov't will be pusing it.
Yeah, I've noticed that as well. In my college math classes, calculators were not allowed at all. However, the classes were designed so you really didn't need one anyways.
In other classes where you dealt with numbers, they were allowed. In fact, some classes even allowed a full-blown laptop with programs like Maple and Matlab.
The truth is that it really depends on what you're actually trying to get done in the class. If the goal of the class is learning the math, then no calculators. If the goal of the class is to learn how to apply the math to a bigger problem, then use calculators because it isn't the math itself they're testing you on.
The best solution for this I've found so far, is a program called OpenVPN. The thing is easy to configure, runs on most 'nix platforms, tunnels over UDP (I'm sorry, but IP over TCP is stupid), and works wonderfuly.
But this device has one big feature you neglected to mention, that all of the "common low-end gateway boxes" seem to lack almost completely...
Dynamic routing
(BGP, OSPF, etc.)
Seriously, how can you call something a "router" when it doesn't even support any useful "routing protocols"?
Sure, binary compatability is all nice and stuff, but only works if the binary in question, and the machine you're running it on, both use the same processor architecture.
As AIX only runs on RS/6000 (pSeries) machines, which use POWER and PowerPC processors, for any binary compatability to work, you'd need binaries from Linux/PPC.
What's really nice, though, is binary compatability between open source OSes. For example, FreeBSD has excellent Linux binary compatability. It's really easy too, as it just implements a system call translaton layer, and keeps a collection of the Linux libs off in a directory as needed. In fact, it's just as fast as running the same binary under Linux (some might argue faster).
I've also seen kernel options for much wilder binary compatability features to run binaries from commercian 'nixes on OpenBSD/NetBSD, though I've never really looked into them, or used them.
However, I'm wondering, does Linux have much in the way of these capabilities in and of itself? (I remember this "ibcs" thing many years ago that could run SCO binaries, but havn't really seen anything since)
It's not really just an x86 vs Apple thing, but rather an x86 vs "Everyone Else".
Around the 1990'ish (give or take), lots of companies who had been using antiquated CISC architectures decided to break compatability and move up to the world of clean RISC architectures that were easy to make the 32-to-64-bit jump with, and not break software compatability.
For example:
Apple: M680XX -> PowerPC
Sun: M680XX -> SPARC
DEC: VAX -> Alpha (64-bit from the start)
HP: M680XX -> PA-RISC
Yup, I definitely pushed the limits... Thankfully they revamped the electrical system in the dorm I lived in before I moved in there. I'm one of the few people who actually took the time to map out the outlet-to-breaker-circuit mapping for the room. Hey, I had to make sure the refrigerator/microwave used a different circuit than my computers. Before I did that, I found that if I had all my monitors on (including one that was normally off), and my roommate microwaved something, the breaker tripped.
;-)
Of course, I also push the limits on how much power one uses for computer equipment. The height of my computer power draw was probably my sophomore year. At the end of that year, I borrowed an ammeter and took some measurements. One favorable thing I discovered was that most computers generally use nowhere near as much power as their power supplies are rated for. However, I still managed to max out my load at 14.5A (approx 1740W) continuous draw. Thank goodness I wasn't paying directly for that off the meter
Of course in those days I was running like 8 computers and 3 monitors. Now that I pay for my power, I have cut down a bit on the amount of 24/7 stuff I'm running, though I still manage to use quite a bit of electricity. Then again, now that I live in Florida, I'll bet that a lot of my juice goes too power the A/C.
It's amazing how often I randomly and indirectly notice RPI folk posting on /. ;-)
Hey, on a machine like that you should be running AIX, not Linux. Then again, if you insist on only wanting to run Linux on that fine RS/6000, I'll gladly take it off your hands and load a proper OS for the machine.
Hehe... In college, I was in a club where we managed to pick up and use old computer hardware. One class of machines I played with a lot were some Sun 3/280s. The one I worked with the most had SunOS 4.1, a 68020 processor, 32MB of RAM, a 600MB disk, and an 800MB disk. (these disks took 2 people to lift)
:) (commercial SSH, which has 68k asm optimizations in the source, unlike OpenSSH)
:) (but they're probably still there, being poked at every once in a while) At least these are the "real man's VAXen", complete with 3-phase power.
The impressive part was that I actually got gcc to compile, then built ssh and apache. SSH actually worked decently on the machine as well
Unfortunately, I think the machines aren't running any more, as their space has been taken over by a DEC-purist, and turned into VAX-country
The oldest computer currently running in my setup would have to be my console server. It's an IBM RS/6000 POWERstation 250. (PowerPC 601 at 66MHz, about 80MB of RAM, and a 400MB disk) Still runs just fine. However, it's getting replaced sometime soon by an embedded console server of sorts, simply because it's a cleaner solution.
I agree. However, as the pundits in the Sun groups on Usenet keep pointing out, Apple still doesn't put ECC memory in these things. If the dual G5 had ECC memory, they'd probably shut up, and might even come down off their high horses and buy 'em.
Actually, from observations, I've noticed that the Intel 80186 is IBM's embedded I/O processor of choice for most of the microchannel cards for RS/6000 systems that I've seen.
Ok, that statement is just plain wrong, unless you're comparing something other than CPUs.
First, that Indigo2 is not "plain-jane IDE" (unless you're using some weird adapter board), but rather "plain-jane SCSI-2" (10MB/s).
Second, one big factor you notice when comparing CPUs, especially when some are "budget models" is that magic thing known as cache. Ever wonder what feature they're cutting to lower cost? I'll bet the R4400 has plenty of cache, while the Duron cuts cache (so does the Celeron, and some of Sun's older and slower microSPARC CPUs)
Third, even with those factors, there's no way in hell that the MIPS R4400 (at 120MHz) CPU could ever come close to touching the performance of an AMD Duron (750MHz). You have to be comparing graphics cards.
Now, one of the features of the Indigo2 that you might be using, is the "Impact" line of graphics cards. The Solid (no texturing) and High Impact (texturing) versions have about 450 MFLOPS of performance on the card itself, and the Max Impact has double that. I will believe that your Indigo2 whoops the crap out of the Duron on graphics, if you're comparing one of those fine GIO64 graphics cards to some POS card you threw in the PC.
But I will NOT believe you're comparing CPU performance.
How do I know this? Well, let's just say I've got an R10000 (195MHz) SGI Indigo2 High Impact sitting next to me.
Yes, I'm using a Sun workstation right now to post this comment. (and yes, my new laptop is a PowerBook) In fact, I recently deployed some cheap Sun Ray thin clients around my apartment, which makes for really nice hot-desking around the place.
It's funny how people keep knocking XFree86 for performance... What ever happened to those days back in 1995, when XF86 (when using the accelerated server on an ATI Mach64) was actually FASTER than Win95? (yes, that was on a friend's P-100, which was brand-spanking new)
Actually, there is a real operating system out there, called "P/OS". It was the DEC "Professional Operating System", and I think it ran on their "Rainbow" series of microcomputers. It's CLI is VMS-style (I guess that means DCL-style). I first saw it running on a "VAX Console" (rebadged PDP-11 based microcomputer cabled to a "VAX 8530").
Gnome is unusually slow on Solaris/Sparc, for some reason. Try KDE or any other WM and it will perform MUCH better.
The only possible form of terrorist attack would have been pre-launch sabotage of the craft. Then again, given all the checks and rechecks, they'd probably have to infiltrate most of the shuttle inspectors. Basically, it's not realistically possible.
I think that's because at RPI it's well understood that the H/SS classes are just "requirements to get out the way" and generally aren't serious and important classes for the vast majority of the students. So, they try to make many of them interesting, but rarely academically challenging.
Speaking as a very recent RPI graduate, I think I
only ever got 1 B among all my H/SS coursework.
Eh... The Creator3D is really just one of those "we support 3D, but don't buy this card if you care about 3D" board. Sun's overpriced number for the card is around $715, but they can be had easily on eBay for thr $100-170 range. They are pretty nice 2D cards though, and people like 'em when they want to run at high-res.
Now as far as 3D performance, I wouldn't be the least bit surprised if the High Impact board in my SGI Indigo2 workstation blows the crap out of it. Oh, and the I2 I have is from '96, but the copyright data on the card is '95. (and it's not even the highest-end option for the machine) Oh, and that card also supports stereo viewing.
Ahh, reminds me of my days around the E-Club at dear old RPI... (ok, this was just the past few years)
the club website
It all started when the school threw out a VAX 8530... Thankfully, we have 3-phase power available to run the thing, and it's now happily running VMS 5.5-2 in all it's massive glory. Over the years we've also accumulated a VAX 11/780 (dead unfortunately), a PDP-11/45 (which one club member had to rebuild the power supply for), a pair of Sun 3/280s (complete with 12" platter hard drives), along with various other "smaller" machines that might be to new to be considered classic. (some MicroVAX-class machines, a bunch of old model IBM RS/6000s, some HP9000 stuff, etc, etc.)