If I'm being insulted that is fine but um... I just need to know it:)
I can't repond more then that cause I really have no idea what you are trying to get at.
I can say that the hit rate during DS was 1) over-hyped and 2) still better then we had hoped and better than any other time in history.
Think about it for a moment. During WW2, Korea and Vietnam we were just happy to hit the target after dropping tens of bombs in vietnam and hundreds of (not or) THOUSANDS of bombs during WW2.
Now we are screaming bloody murder (literally) when a single bomb out of hundreds misses the target. I realize we should always endevour to minimize non-combatant casualties but war is NOT pretty and war does kill people.
PS where do you get the initial position? it is call compass rose. You put the plane, car, boat in a known starting spot like at the end of a run way and you press RESET. there now you are done. It isn't like we don't know where certain points of reference are on the earth.. DUH.
Actually I'm not trolling and I know exactly what I'm talking about. Now when I say this next part you are going to say, "oh yeah right ah huh SUUUuuurreee."
But the reason I didn't give details is because it is classified. But trust me I know what DoD inertial navigation systems can do.
20 min trips? try 24 hour trips with enough accuracy to blindly put your nose gear in the same spot you took off from.
GPR or Loran is preferable. Um.. wrong. 9 times out of 10 the error, even in military GPS (which IS better than civilian grade) in GPS causes a system to default into inertial navigation mode, and often at the worst possible moment.
When you want to guide a plane, car, or boat to a location GPS, LORAN, TACAN, etc are just fine. when you want to put a bomb through a window after a 12 hour flight from 32,000 ft you better use INS.
HAHAHAHA trust me inertial guidence / positioning can be made VERY accurate with nearly NO drift.
When I say nearly I mean... well... lets just say it is more accurate than military grade GPS. But I'm also talking about military grade inertial guidence which as another poster mentioned, is rather classified and expensive.
Right but what is it going to send? not your PGP passphrase cause that was done via the sandbox and that isn't hooked to the network. Hell don't even install email client, just a text editor and the crypto.
so let them put a virus on your "public box" cause there is nothing on there / nothing typed on there that can help them decode your msg.
Couldn't you avoid this by running your encryption software (aka PGP) on a non-networked computer? Then xfer the cyphertext via floppy. And if you don't physically secure a box then you are just asking to be compromised.
No matter what they do they can't get at a non-networked box unless they physicaly break in and hack it and then again to retrieve the data (or transmit via radio waves). As for the networked box it never sees anything but cyphertext, no passphrases are used, and anything it puts on the floppy doesn't matter cause even if it gets on the sandbox it can't get anywhere.
Oh sure they could get tricky, do things with floppy boot sector virii that will run in the sandbox, log and save to the floppy, then re-run once it detects a network connection, but to this non-programmer that seems 1) problematic and 2) pretty easy to avoid. maybe even use CD-R or CD-RW.
Two points. One is legitimate and the other is weak but an unfortunate reality that I hate and which we should never accept while continuing to fight it.
The latter first: all politicians lie to get votes. It is a fact. Deplorable and reprehansible yet a fact. So until we fix it, we must then decide who lies the least and who lies about insignifigant drival because the media and sensationalist populace requires them to; ie not focusing on real issues rather than what the candidates favorite book is.
The idea that because someone lies about one thing means they will lie about another is a horrible slippery slope argument however no one would ever consider supporting a known liar. No one has that good a public opinion to handle the fall out. But the fact remains that just because a political candidate lies to avoid pointless media drama doesn't mean they can't get the job done. Hell i think the ability to lie... er... bluff is a GOOD trait in a political leader that must deal in the international areana. But that is a different conversation entirely.
The second point is really the important one. If I understand the crux of your statement you want to know why we can't divulge the information, that it looks fishy if we don't, and that there is no reason you can think of as to why not to divulge the information.
NEWS FLASH: Telling people what you know can/usually tells them HOW you know. Right now the last thing we need is to compromise the prescious few informational sources we have on Usama (not Osama like the news reports) bin Laden and his network. That is why and it is a very good legitimate reason.
Strawman? or Red Herring. Hum not sure but nice to see someone on here who can actually dissect an argument for what it is. Been so long since I took my debate course I can't remember the key differences with strawman and red herring though.
As someone who worked at a DoD Lab I can tell you that a good number of those supercomputers are NOT at the NSA. Now could they be doing NSA work? Sure I guess they could, since I couldn't possible know the task of all the cycsle of all the machines i managed (admin not tech manage). But I knew the researchers and the work they were doing and most of it revolved around seriously basic science like modeling chemical reactions, not cracking codes.
Yes I physically saw the computers and there are 4 of these "High Performance Computing Labs" called Major Shared Resource Centers and then several "lesser" sites. called Distributed Centers. Check out http://www.hpcmo.hpc.mil/ I used to work for these guys.
Well, because I get way more than 5 spams/day on average, more like 20-30, and I'm lucky enough to have a cable connection so I don't have to wait very long for those emails to download. But I don't even want to imagine what a hassle it is for modem folk.
Try to imagine what it would be like to get 20 phone solicitation calls/day. Now imagine it without caller ID. Now imagine your phone is delayed so that you have to wait say 10 seconds after picking up for the caller to start their pitch.
I have had my same email addy now for 6 years and do not feel I should have to hop to another addy just to try and avoid spammers.
That is why spammers need to be dealt with and why [god i can't believe i'm about to say this] there outta be a law... forcing all info distribution to be opt-in not opt-out so I have to explicatly give my permission; and no i never will.
Actually I DO use VBR. I realize you were mostly saying this in a general sense but I wanted to defend my Em-Pee-Three 1337ness:)
Seriously though I use Lame with V2 with a min bitrate of 192 and a max of 320. Only on a rare few of my techno mixes does it ever even touch 320 for a moments time and otherwise spends the rest of the time around 256 with frequent drops to the 192 level. This puts me at about a 8:1 compression level and about a 200-240kbps equivilent bit rate. The quality, when played through my Onkyo is very good and the bass, played through a Sunfire True Subwoofer is very good (for those that say mp3 kills bass). And no, I didn't pay full price for it >;-)
One question I haven't been able to find the answer to though is: What criteria does the encoder use to determine what bitrate to use, and are some more or less conservative than others?
What the article fails to mention, or I'm too blind to find, is what bit rates these tests were run at and a quick discussion of what bit rates the various formats are considered best at. We all know some encoders are optimized for a given bit rate and encoding above or below that can often greatly impact quality; especially things like freq response, transient response, and spaciousness effected by those "small cues" they talk about.
If I had to guess they used a low bit rate for MP3Pro under the assumption that everyone will shoot for the same quality they do with MP3 but with the Pro versions greater compression. Personally I plan on doing all my futere encoding at the same bit rate looking for the (hopefully) improved quality over MP3. The same comment applies for Vorbis as well.
"This somehow ties into the fight over Windows XP. The big companies didn't care as much when microsoft was going after smaller companies like Netscape. Now suddenly it is their lunch on the table, and it becomes important."
this all sounds like a parable I heard once. It ended something like this:
... and when they came for me there was no one left to complain...
Well lets see. We have heard about some gasoline cars that can get as good as 40-50 mpg right? So if biodiesal is no more than 2x the price of gasoline you effectively have no price increase per mile driven.
On the other hand I wonder how optimized that Audio A2 is or if it is more along the lines of just above average, meaning it is the diesal equivelent of the 25mpg gasoline cars we have now. that would mean biodiesel could cost as much as 3x the gasoine price and still retain an equivilent price/mile.
I think a nice continuation to this study would follow like this:
Ok so we know what the ping distribution of users on his server is. And we have a reasonable idea of the level of usage classified in various ways (domain, location etc). What if now he were to artifically increase the ping rate a little bit every week.
I seem to recall that little gadget that attached to a network which would simulate any sort of network condition you could think of (dropped packets, retransmits, latency etc). Put that bad boy on there, or code up the server in some other way to increase ping, say 10 ms, every week and watch how usage drops. Maybe he should also add in some level of increased packet loss. As someone else mentioned high latency is often times accompanied by high packet loss and then it starts to hurt Quake's prediction (ie try to really simulate what it is like to be a HPB).
I think this is what will really let us see if there is a limit where people, who on avereage would hit your server, now start going away ni search of lower pings.
It's called the scientific method. We have our control server that we observed in the wild and got a baseline. Now lets add the test server and see how changing it's behavior effects usage. THAT is how to really do a test.
Slashdot Mirror
inenventing? thin air and flatulaance?
... I just need to know it :)
If I'm being insulted that is fine but um
I can't repond more then that cause I really have no idea what you are trying to get at.
I can say that the hit rate during DS was 1) over-hyped and 2) still better then we had hoped and better than any other time in history.
Think about it for a moment. During WW2, Korea and Vietnam we were just happy to hit the target after dropping tens of bombs in vietnam and hundreds of (not or) THOUSANDS of bombs during WW2.
Now we are screaming bloody murder (literally) when a single bomb out of hundreds misses the target. I realize we should always endevour to minimize non-combatant casualties but war is NOT pretty and war does kill people.
PS where do you get the initial position? it is call compass rose. You put the plane, car, boat in a known starting spot like at the end of a run way and you press RESET. there now you are done. It isn't like we don't know where certain points of reference are on the earth .. DUH.
Actually I'm not trolling and I know exactly what I'm talking about. Now when I say this next part you are going to say, "oh yeah right ah huh SUUUuuurreee."
.. wrong. 9 times out of 10 the error, even in military GPS (which IS better than civilian grade) in GPS causes a system to default into inertial navigation mode, and often at the worst possible moment.
But the reason I didn't give details is because it is classified. But trust me I know what DoD inertial navigation systems can do.
20 min trips? try 24 hour trips with enough accuracy to blindly put your nose gear in the same spot you took off from.
GPR or Loran is preferable. Um
When you want to guide a plane, car, or boat to a location GPS, LORAN, TACAN, etc are just fine. when you want to put a bomb through a window after a 12 hour flight from 32,000 ft you better use INS.
Please don't take this wrong but ... um ...
... well ... lets just say it is more accurate than military grade GPS. But I'm also talking about military grade inertial guidence which as another poster mentioned, is rather classified and expensive.
HAHAHAHA trust me inertial guidence / positioning can be made VERY accurate with nearly NO drift.
When I say nearly I mean
Right but what is it going to send? not your PGP passphrase cause that was done via the sandbox and that isn't hooked to the network. Hell don't even install email client, just a text editor and the crypto.
so let them put a virus on your "public box" cause there is nothing on there / nothing typed on there that can help them decode your msg.
or again am i missing something?
No matter what they do they can't get at a non-networked box unless they physicaly break in and hack it and then again to retrieve the data (or transmit via radio waves). As for the networked box it never sees anything but cyphertext, no passphrases are used, and anything it puts on the floppy doesn't matter cause even if it gets on the sandbox it can't get anywhere.
Oh sure they could get tricky, do things with floppy boot sector virii that will run in the sandbox, log and save to the floppy, then re-run once it detects a network connection, but to this non-programmer that seems 1) problematic and 2) pretty easy to avoid. maybe even use CD-R or CD-RW.
Comments?
The latter first: all politicians lie to get votes. It is a fact. Deplorable and reprehansible yet a fact. So until we fix it, we must then decide who lies the least and who lies about insignifigant drival because the media and sensationalist populace requires them to; ie not focusing on real issues rather than what the candidates favorite book is.
The idea that because someone lies about one thing means they will lie about another is a horrible slippery slope argument however no one would ever consider supporting a known liar. No one has that good a public opinion to handle the fall out. But the fact remains that just because a political candidate lies to avoid pointless media drama doesn't mean they can't get the job done. Hell i think the ability to lie
The second point is really the important one. If I understand the crux of your statement you want to know why we can't divulge the information, that it looks fishy if we don't, and that there is no reason you can think of as to why not to divulge the information.
NEWS FLASH: Telling people what you know can/usually tells them HOW you know. Right now the last thing we need is to compromise the prescious few informational sources we have on Usama (not Osama like the news reports) bin Laden and his network. That is why and it is a very good legitimate reason.
Alex
Strawman? or Red Herring. Hum not sure but nice to see someone on here who can actually dissect an argument for what it is. Been so long since I took my debate course I can't remember the key differences with strawman and red herring though.
Anyway, good post.
Yes I physically saw the computers and there are 4 of these "High Performance Computing Labs" called Major Shared Resource Centers and then several "lesser" sites. called Distributed Centers. Check out http://www.hpcmo.hpc.mil/ I used to work for these guys.
Try to imagine what it would be like to get 20 phone solicitation calls/day. Now imagine it without caller ID. Now imagine your phone is delayed so that you have to wait say 10 seconds after picking up for the caller to start their pitch.
I have had my same email addy now for 6 years and do not feel I should have to hop to another addy just to try and avoid spammers.
That is why spammers need to be dealt with and why [god i can't believe i'm about to say this] there outta be a law
Seriously though I use Lame with V2 with a min bitrate of 192 and a max of 320. Only on a rare few of my techno mixes does it ever even touch 320 for a moments time and otherwise spends the rest of the time around 256 with frequent drops to the 192 level. This puts me at about a 8:1 compression level and about a 200-240kbps equivilent bit rate. The quality, when played through my Onkyo is very good and the bass, played through a Sunfire True Subwoofer is very good (for those that say mp3 kills bass). And no, I didn't pay full price for it >;-)
One question I haven't been able to find the answer to though is: What criteria does the encoder use to determine what bitrate to use, and are some more or less conservative than others?
If I had to guess they used a low bit rate for MP3Pro under the assumption that everyone will shoot for the same quality they do with MP3 but with the Pro versions greater compression. Personally I plan on doing all my futere encoding at the same bit rate looking for the (hopefully) improved quality over MP3. The same comment applies for Vorbis as well.
this all sounds like a parable I heard once. It ended something like this:
On the other hand I wonder how optimized that Audio A2 is or if it is more along the lines of just above average, meaning it is the diesal equivelent of the 25mpg gasoline cars we have now. that would mean biodiesel could cost as much as 3x the gasoine price and still retain an equivilent price/mile.
Ok so we know what the ping distribution of users on his server is. And we have a reasonable idea of the level of usage classified in various ways (domain, location etc). What if now he were to artifically increase the ping rate a little bit every week.
I seem to recall that little gadget that attached to a network which would simulate any sort of network condition you could think of (dropped packets, retransmits, latency etc). Put that bad boy on there, or code up the server in some other way to increase ping, say 10 ms, every week and watch how usage drops. Maybe he should also add in some level of increased packet loss. As someone else mentioned high latency is often times accompanied by high packet loss and then it starts to hurt Quake's prediction (ie try to really simulate what it is like to be a HPB). I think this is what will really let us see if there is a limit where people, who on avereage would hit your server, now start going away ni search of lower pings.
It's called the scientific method. We have our control server that we observed in the wild and got a baseline. Now lets add the test server and see how changing it's behavior effects usage. THAT is how to really do a test.