I'd tentatively agree, this might imply that you couldn't really have multihomed hosts connected directly to external networks, but you could still have that host connected to a multihomed network, in multiple ways so you still avoid a single point of failure. Right?
OK, I'm provisionally accepting the premise of the question-- that something Congress might do could help Internet security-- and trying to figure out what I'd suggest.
It'd help if IP packets couldn't be spoofed (or if such spoofing capabilities were dramatically reduced).
Then any hack attempts could be tracked much, much more easily back to their origins.
In a perfect world, one might upgrade all our networks to employ IPv6 or IPsec to ensure greater packet integrity, but this is prohibitively expensive and leaves the problem largely intact on "legacy" networks.
A simpler solution, which would be greatly accelerated with a Congressional (or Executive?) national security legal mandate, would be a law requiring network owners (ISPs) to install filters on the boundaries of their networks that prevent packets from leaving their networks that didn't originate with IP source addresses owned by their networks. Egress filtering.
While this wouldn't eliminate IP spoofing (someone can still pretend to be another computer on the same network), it would eliminate someone on network A pretending like they came from network B in most cases. At that point, the NOC of the appropriate network can be contacted and the hack can be run to ground.
(Someone more network-savvy than I could articulate the boundaries of which networks should be included under the above statute. Obviously traffic being routed between networks (as opposed to traffic originating from a network) cannot be covered by such a requirement.)
Nobody likes mandates, but I think this one would significantly improve end-to-end network security. Making it a legal requirement would enable the practice to be sufficiently end-to-end to be useful. And it's inexpensive enough that ISPs have debated doing it on their own just as a measure to reduce DOS problems.
--LP
Disclaimer: I program web and TCP/IP software but am not a network admin.
But many religions are very successful without worrying about being rational... or logical.
How do you explain that!?
Simple. The above statement about religions is false.
To understand why, try substituting the words "relying exclusively upon" for your words, "worrying" and examine the semantic difference. Or similarly remove the words "worrying about" and the difference between your false statement, and a more rational one.
Nevertheless, your post conveyed a point pretty well, without being particularly rational. Perhaps religions work the same way, eh?
Why would a religion attempting to communicate with and to people, whether made by man or by god, adhere to the formalism you seek and find in science?
Dude, think about what you are saying. Do you want to keep maintaining your old crappy code or pass that job onto someone else? Or do you want to go write some new code?
Your perspective assumes your company requires a fixed amount of software. Think more imaginatively.
Better documentation means you can shove maintenance to a more junior programmer with less pushback.
Also, without good documentation, its a b*tch to try to outsource/handoff pieces of the code you don't want to bother writing.
Besides, I don't care how well documented your code is, you should always be able to convince a boss that its more efficient for you to make changes to it (even at higher salary) than some cheaper guy who has never seen the code before.
Heck, even if the page doesn't have external links on it, but the manager clicks on his bookmark list to visit his favorite goofoff site immediately afterward, the referrer has his password.
There's a great Slash-based site with loads of articles examining potential next-gen interfaces. Not a huge amount of traffic yet, but the editor seems to be consistently putting up new articles. Check out Nooface.
The USA killed ~7000 innocent Somalian civilians in -93 while failing to kill one single warlord.
Stop showing your ignorance. You may not have liked the Somalian mission, but lets be honest about what the mission was. Killing Somalian warlords was *not* the mission.
Originally the mission was humanitarian, under Bush the elder- "open supply routes, get food moving, prepare the way for a UN peacekeeping force."
Under Clinton, in part due to the deliberate killing of 24 UN peacekeepers, the mission changed somewhat to capturing (that's right, *capturing*, not killing) one warlord, Gen. Mohamed Farrah Aidid, as well as commanders under him. If you find the deaths of 7,000 civilians deeply troubling, as I do, you might try reading BlackHawk Down to get some perspective on how such things occur.
You might pause to consider how (and if, of course) the USA should use its power when attempting to prevent a million starving people from dying due to the fact that food supplies can't get into a country during yet another civil war. Keep in perspective that while the US did sacrifice 34 of its own lives (and a billion or so in cash) and 7000 Somalis died, we were trying to prevent the starvation deaths which had already killed 300,000 Somalis, with the International Red Cross warning at the time of a potential 1.5 million deaths without greater food distribution. (I don't hear you trying to hold any warlords responsible for those 300,000 deaths now, do I? Why didn't the person who handed you that one-liner set of facts bother to mention them?)
Being concerned about the safety of food distribution (having watched rival Somali clans attempt to use food as a weapon by stealing, hoarding, and denying it to particular people), the UN first sent 50 unarmed monitors, then 500 security guards, then 5000, then ultimately 25000 US troops to insure that food aid could get through without being intercepted by warring local warlords. Yeah, USA- those bastards!
After it was clear to the US that its presence wasn't being effective (and the conflict was getting personal), it left, arranged for 25,000 UN troops from scattered countries to replace it, and after 8 more years, the UN has finally helped install Somalia's first government in a decade, the Transitional National Government (interview here). Meanwhile US food aid continues to stream into the country. Man, the USA really sucks, doesn't it!
Hmm, looks like Appian's latest "AppianX" uses some custom chip they might have developed, based on 3Dlabs VHDL? At least, reading between the lines of this press release where Osman Kent mentions licensing VHDL cores but the current products mentioned there use off-the-shelf 3Dlabs parts, and this press release announcing AppianX but not specifying who made the chipset and thus implying that Appian did, presumbably not totally from scratch given the complexity required, right? Ah, speculation.
Still pretty removed from whose (3D) graphics chipset goes with this display, which is what I really want to know. Kinda a relevant question for the CAD market which could afford these things, no?
So what graphics card does it use? I noticed that the specs implied that you needed dual-DVI connections to supply all the bandwidth between graphic card and LCD display. (And you can't use 2 AGP cards to get 2 DVI connections since AGP is designed to be limited to a single point-to-point bus; one AGP slot per system.) But the only card I know of with those is from Appian Graphics but even there, I don't know which 3D chipset they're using these days. (Still 3Dlabs?) Anybody know?
I did notice that while it apparently came with a PC graphics card, the "Prerequisites" line in the specs implied that it required an Intellistation E/M/Z computer from IBM. So yeah, it probably does come with an IBM computer but not quite in the way you mean-- you have to pay extra on top!
Applications could be written to use such anti-aliasing features of course, but I haven't seen them incorporated under-the-covers in 3D graphics drivers and thus used automatically by all applications. If they did so, they'd have to be careful to market around the 4x drop in fill rate performance in that mode.
If you did such resampling in the video display device itself, you'd have a bandwidth problem communicating all that info over the DVI graphicscard-to-LCD bus.
Frankly, I'd prefer to see people working on tiling together LCD panels to make economical but larger displays, somewhat like Mass Multiples. But even there you still have the problem with DVI bandwidth. Even IBM's display required dual DVI cables between graphics card and monitor which is why they tossed in the card.
The US fusion budget is nowhere close to $1 B a year.
Sorry, gross from-memory rounding on my part. The real US fusion budget amount is $250 million a year. Still surprising to me, and I suppose a shrewd long-term move. Especially when one considers how much the US economy and security depends on foreign oil.
Heck, when looked at from that angle, the US could probably re-allocate even $5 billion of its annual spending from middle-east defense to accelerating fusion feasibilty, breaking even long-term by reducing the amount of money that the US spends defending its interests there.
I dunno, most fusion scientists I've heard are always saying "we're only 50 years away!" I marvel at the fact that the U.S. quietly and (hopefully) wisely tucks away a billion a year or pursuing this stuff, with so little short-term payoff.
At the heart of the PR mission is communicating the value of what is being publicized. Good communication skills are pretty important to good science, and especially important if you want anything useful to come out of whatever work you do. There's lots of PR bullshit out there, but the core job of PR is both necessary and useful. PR is not just a shenanigan.
Not everyone by default cares and can fully appreciate good science projects without PR "education", just as not everyone can appreciate good homeless shelters and reading programs without a little PR. Sure, people always seek to know about the stuff they're interested in, but the vast masses have to be convinced that things they care less about are worth sparing a few (taxpayer) dollars for. PR widens the circle of aware people. And it's bottoms-up education rather than top-down. You might think about why such a distributed system has advantages.
--LP, who never thought he'd be defending PR people, on Slashdot of all places
About 3-6 months ago I went looking for notebooks with 1600x1200 screens (and built-in CD-RW drives) and after noticing that Dell had em, I checked to find out who else did. HP, Compaq, Gateway, and Sony didn't, but IBM (and Dell) did.
I think it is good for inventors to take some moral responsibility for their inventions, although that would wisely be tempered with a recognition that all sufficiently useful devices based on publicly understood knowledge are likely to be developed eventually by somebody.
Like most things, there is a necessary balance between the need for transparency in an efficient democratic society, and the need for protection from unreasonable search and seizure (e.g. the fourth ammendment). Phil helped tip the technological balance in one direction, but he didn't upend the scales.
My understanding was that the 65,000 Microsoft "bugs" was basically "potential bugs" generated by a lint-like checker that searched for buffer overflows, certain pointer manipulations and other common 'maybe-a-bug' issues.
My perception is that Bush didn't really tell the DOJ what to do, it was more subtle than that. The Republicans, having taken office and putting their own people in high positions as they have the perogative to do, got rid of many of the anti-trust experts and litigators they hired for the case, put some junior people on the job who don't know much about anti-trust, and those are the people now making these (IMHO poor) decisions.
I think I'd tend to agree with the original poster. I might accept a claim that the same people are running the CIA now as 15-20 years ago, but not 30, for reasons that will be clear if you read the following chronicle of the unveiling of CIA misdeeds.
If there's evidence beyond your blanket assertion that "the same people are running the show now as 30 years ago," I'm all ears.
One aspect of your comments I'd take issue with. Based on the little I've read on the subject, it does not appear to me that Saudi Arabia is in debt due to stupid or corrupt moves by King Fahd. It's mostly due to the fact that Saudi Arabia created huge entitlement programs for its citizens due to its oil wealth, and since then the birthrate and population have exploded (of course), and thus Saudi Arabia can no longer adequately pay for them with current oil revenues.
Saudi Arabia then has the choice of cutting benefits, a hard task as US citizens can appreciate, or shipping more oil, which is tricky due to Saudi agreements with OPEC and desire to keep the price of oil from going lower. Basically, the Saudis are going to have to develop some non-oil industry in their country and that's not an easy process.
Slashdot Mirror
Not to mention that they highlighted Doom and totally ignored Castle Wolfenstein 3D, id's 3D FPS Doom predecessor, as most clueless histories do.
--LP
Egress filtering would be orders of magnitude cheaper to implement.
I'd tentatively agree, this might imply that you couldn't really have multihomed hosts connected directly to external networks, but you could still have that host connected to a multihomed network, in multiple ways so you still avoid a single point of failure. Right?
OK, I'm provisionally accepting the premise of the question-- that something Congress might do could help Internet security-- and trying to figure out what I'd suggest.
It'd help if IP packets couldn't be spoofed (or if such spoofing capabilities were dramatically reduced).
Then any hack attempts could be tracked much, much more easily back to their origins.
In a perfect world, one might upgrade all our networks to employ IPv6 or IPsec to ensure greater packet integrity, but this is prohibitively expensive and leaves the problem largely intact on "legacy" networks.
A simpler solution, which would be greatly accelerated with a Congressional (or Executive?) national security legal mandate, would be a law requiring network owners (ISPs) to install filters on the boundaries of their networks that prevent packets from leaving their networks that didn't originate with IP source addresses owned by their networks. Egress filtering.
While this wouldn't eliminate IP spoofing (someone can still pretend to be another computer on the same network), it would eliminate someone on network A pretending like they came from network B in most cases. At that point, the NOC of the appropriate network can be contacted and the hack can be run to ground.
(Someone more network-savvy than I could articulate the boundaries of which networks should be included under the above statute. Obviously traffic being routed between networks (as opposed to traffic originating from a network) cannot be covered by such a requirement.)
Nobody likes mandates, but I think this one would significantly improve end-to-end network security. Making it a legal requirement would enable the practice to be sufficiently end-to-end to be useful. And it's inexpensive enough that ISPs have debated doing it on their own just as a measure to reduce DOS problems.
--LP
Disclaimer: I program web and TCP/IP software but am not a network admin.
But many religions are very successful without worrying about being rational... or logical.
How do you explain that!?
Simple. The above statement about religions is false.
To understand why, try substituting the words "relying exclusively upon" for your words, "worrying" and examine the semantic difference. Or similarly remove the words "worrying about" and the difference between your false statement, and a more rational one.
Nevertheless, your post conveyed a point pretty well, without being particularly rational. Perhaps religions work the same way, eh?
Why would a religion attempting to communicate with and to people, whether made by man or by god, adhere to the formalism you seek and find in science?
Dude, think about what you are saying. Do you want to keep maintaining your old crappy code or pass that job onto someone else? Or do you want to go write some new code?
Your perspective assumes your company requires a fixed amount of software. Think more imaginatively.
Better documentation means you can shove maintenance to a more junior programmer with less pushback.
Also, without good documentation, its a b*tch to try to outsource/handoff pieces of the code you don't want to bother writing.
Besides, I don't care how well documented your code is, you should always be able to convince a boss that its more efficient for you to make changes to it (even at higher salary) than some cheaper guy who has never seen the code before.
--LP
Heck, even if the page doesn't have external links on it, but the manager clicks on his bookmark list to visit his favorite goofoff site immediately afterward, the referrer has his password.
There's a great Slash-based site with loads of articles examining potential next-gen interfaces. Not a huge amount of traffic yet, but the editor seems to be consistently putting up new articles. Check out Nooface.
--LP
I agree the code name is a tad witty.
So how often would such an eclipse occur? Once in a blue moon?
--LP
Why didn't they use mod_perl's Apache::Registry mode? Was there a particular programming practice that prevented switching easily?
--LP
The USA killed ~7000 innocent Somalian civilians in -93 while failing to kill one single warlord.
Stop showing your ignorance. You may not have liked the Somalian mission, but lets be honest about what the mission was. Killing Somalian warlords was *not* the mission.
Originally the mission was humanitarian, under Bush the elder- "open supply routes, get food moving, prepare the way for a UN peacekeeping force."
Under Clinton, in part due to the deliberate killing of 24 UN peacekeepers, the mission changed somewhat to capturing (that's right, *capturing*, not killing) one warlord, Gen. Mohamed Farrah Aidid, as well as commanders under him. If you find the deaths of 7,000 civilians deeply troubling, as I do, you might try reading BlackHawk Down to get some perspective on how such things occur.
You might pause to consider how (and if, of course) the USA should use its power when attempting to prevent a million starving people from dying due to the fact that food supplies can't get into a country during yet another civil war. Keep in perspective that while the US did sacrifice 34 of its own lives (and a billion or so in cash) and 7000 Somalis died, we were trying to prevent the starvation deaths which had already killed 300,000 Somalis, with the International Red Cross warning at the time of a potential 1.5 million deaths without greater food distribution. (I don't hear you trying to hold any warlords responsible for those 300,000 deaths now, do I? Why didn't the person who handed you that one-liner set of facts bother to mention them?)
Being concerned about the safety of food distribution (having watched rival Somali clans attempt to use food as a weapon by stealing, hoarding, and denying it to particular people), the UN first sent 50 unarmed monitors, then 500 security guards, then 5000, then ultimately 25000 US troops to insure that food aid could get through without being intercepted by warring local warlords. Yeah, USA- those bastards!
After it was clear to the US that its presence wasn't being effective (and the conflict was getting personal), it left, arranged for 25,000 UN troops from scattered countries to replace it, and after 8 more years, the UN has finally helped install Somalia's first government in a decade, the Transitional National Government (interview here). Meanwhile US food aid continues to stream into the country. Man, the USA really sucks, doesn't it!
--LP
Hmm, looks like Appian's latest "AppianX" uses some custom chip they might have developed, based on 3Dlabs VHDL? At least, reading between the lines of this press release where Osman Kent mentions licensing VHDL cores but the current products mentioned there use off-the-shelf 3Dlabs parts, and this press release announcing AppianX but not specifying who made the chipset and thus implying that Appian did, presumbably not totally from scratch given the complexity required, right? Ah, speculation.
Still pretty removed from whose (3D) graphics chipset goes with this display, which is what I really want to know. Kinda a relevant question for the CAD market which could afford these things, no?
--LP
So what graphics card does it use? I noticed that the specs implied that you needed dual-DVI connections to supply all the bandwidth between graphic card and LCD display. (And you can't use 2 AGP cards to get 2 DVI connections since AGP is designed to be limited to a single point-to-point bus; one AGP slot per system.) But the only card I know of with those is from Appian Graphics but even there, I don't know which 3D chipset they're using these days. (Still 3Dlabs?) Anybody know?
--LP
I did notice that while it apparently came with a PC graphics card, the "Prerequisites" line in the specs implied that it required an Intellistation E/M/Z computer from IBM. So yeah, it probably does come with an IBM computer but not quite in the way you mean-- you have to pay extra on top!
Applications could be written to use such anti-aliasing features of course, but I haven't seen them incorporated under-the-covers in 3D graphics drivers and thus used automatically by all applications. If they did so, they'd have to be careful to market around the 4x drop in fill rate performance in that mode.
If you did such resampling in the video display device itself, you'd have a bandwidth problem communicating all that info over the DVI graphicscard-to-LCD bus.
Frankly, I'd prefer to see people working on tiling together LCD panels to make economical but larger displays, somewhat like Mass Multiples. But even there you still have the problem with DVI bandwidth. Even IBM's display required dual DVI cables between graphics card and monitor which is why they tossed in the card.
--LP
The US fusion budget is nowhere close to $1 B a year.
Sorry, gross from-memory rounding on my part. The real US fusion budget amount is $250 million a year. Still surprising to me, and I suppose a shrewd long-term move. Especially when one considers how much the US economy and security depends on foreign oil.
Heck, when looked at from that angle, the US could probably re-allocate even $5 billion of its annual spending from middle-east defense to accelerating fusion feasibilty, breaking even long-term by reducing the amount of money that the US spends defending its interests there.
--LP
I dunno, most fusion scientists I've heard are always saying "we're only 50 years away!" I marvel at the fact that the U.S. quietly and (hopefully) wisely tucks away a billion a year or pursuing this stuff, with so little short-term payoff.
--LP
At the heart of the PR mission is communicating the value of what is being publicized. Good communication skills are pretty important to good science, and especially important if you want anything useful to come out of whatever work you do. There's lots of PR bullshit out there, but the core job of PR is both necessary and useful. PR is not just a shenanigan.
Not everyone by default cares and can fully appreciate good science projects without PR "education", just as not everyone can appreciate good homeless shelters and reading programs without a little PR. Sure, people always seek to know about the stuff they're interested in, but the vast masses have to be convinced that things they care less about are worth sparing a few (taxpayer) dollars for. PR widens the circle of aware people. And it's bottoms-up education rather than top-down. You might think about why such a distributed system has advantages.
--LP, who never thought he'd be defending PR people, on Slashdot of all places
About 3-6 months ago I went looking for notebooks with 1600x1200 screens (and built-in CD-RW drives) and after noticing that Dell had em, I checked to find out who else did. HP, Compaq, Gateway, and Sony didn't, but IBM (and Dell) did.
Check out the Dell Inspiron 8100 (also sold with Linux on it through Emporer Linux.) And IBM has a somewhat more expensive ThinkPad A Series A22p.
While the font size is small, it is configurable and I appreciate the greater screen real estate.
--LP
that was a while ago. this is "news"?
I think it is good for inventors to take some moral responsibility for their inventions, although that would wisely be tempered with a recognition that all sufficiently useful devices based on publicly understood knowledge are likely to be developed eventually by somebody.
Like most things, there is a necessary balance between the need for transparency in an efficient democratic society, and the need for protection from unreasonable search and seizure (e.g. the fourth ammendment). Phil helped tip the technological balance in one direction, but he didn't upend the scales.
--LP
My understanding was that the 65,000 Microsoft "bugs" was basically "potential bugs" generated by a lint-like checker that searched for buffer overflows, certain pointer manipulations and other common 'maybe-a-bug' issues.
--LP
My perception is that Bush didn't really tell the DOJ what to do, it was more subtle than that. The Republicans, having taken office and putting their own people in high positions as they have the perogative to do, got rid of many of the anti-trust experts and litigators they hired for the case, put some junior people on the job who don't know much about anti-trust, and those are the people now making these (IMHO poor) decisions.
--LP
I think I'd tend to agree with the original poster. I might accept a claim that the same people are running the CIA now as 15-20 years ago, but not 30, for reasons that will be clear if you read the following chronicle of the unveiling of CIA misdeeds.
If there's evidence beyond your blanket assertion that "the same people are running the show now as 30 years ago," I'm all ears.
--LP
One aspect of your comments I'd take issue with. Based on the little I've read on the subject, it does not appear to me that Saudi Arabia is in debt due to stupid or corrupt moves by King Fahd. It's mostly due to the fact that Saudi Arabia created huge entitlement programs for its citizens due to its oil wealth, and since then the birthrate and population have exploded (of course), and thus Saudi Arabia can no longer adequately pay for them with current oil revenues.
Saudi Arabia then has the choice of cutting benefits, a hard task as US citizens can appreciate, or shipping more oil, which is tricky due to Saudi agreements with OPEC and desire to keep the price of oil from going lower. Basically, the Saudis are going to have to develop some non-oil industry in their country and that's not an easy process.
--LP