Load the needed environment for the compiler. Load the source Build the source Boot the source
in 15 seconds, when it takes much longer than that for my already booted system to build a kernel? A P4 isn't THAT much faster than an AMD3200! (And I have done the old "drop to RL1 and build" trick, so it is not an issue of other tasks running).
I want to know a) What kernel options are enabled b) From when are they starting the clock (are they counting the time to load the bootloader and initrd?) and c) is this TRULY a fully functional kernel, or "just enough to get a prompt"?
THUNK! THUNK! THUNK! OH, HAIL!
on
Solar Shingles
·
· Score: 1
'Y'know why it's called "hail"?
Because when you start hearing that "tick. TICK! thunk. THUNK! THUNKTHUNKTHUNK!" you say "OH, HAIL! There goes the roof!"
That's one thing I did not see in the article - how does this stuff fare when you start getting one inch hailstones pelting it.
Then add the 70+ mile an hour winds trying to peel it off.
And if you live in California (excuse me, Kah-lee-FOR-nia), you have the ozone trying to "make friends" with the molecules of the roof. Again, how long will this last?
OK, if you want to put it like that, yes, there is no zero *crossing*.
But the signal still goes to zero(ish), and that is what you use as your locking point, so we were saying the same thing.
Normally, you'd run the signal through a squaring function, then look for signal minima, then feed those into a filtering function to determine the slice point.
Or, if you are looking for a slightly simpler decoder, you have a bit timer that retriggers every bit time. You reset it at any zero crossing (after appropriately narrow band filtering the signal, of course), and sample at 1/2 bit time from when the timer triggers. For noise immunity, I'd sample at 1/4, 1/2, and 3/4 bit time and use a weighted average to determine the bit value. You can deal with a hell of a lot of bit time shift with a decoder like that, and unless you have a run of phase reversals you get enough zero crossings to sync (and you can design your data encoder to provide for those crossings, either with HLDC type encoding, a P/N spreader, or other approaches).
In any case, it't DSP 101 to make a tracking loop for that kind of signal that can deal with enormous bit time errors.
But then again, I design test equipment, so I am always designing my decoders to deal with severely broken encoders.
The whole point of an analemma is to do it on 1 piece of film - no fair taking multiple shots and compositing.
Sure, taking the picture itself can easily be automated.
But fixing a camera to a location so that it will not move DAMMIT (relative to the earth, that is), so that it won't get covered in snow/leaves/pigeondoo/..., so that the film won't be ruined by being out in the elements for a year, being in a location where you can reasonably count on having clear skies enough of the time to get the shots (a month of clouds will really screw you up), being able to judge the exposure needed for the sun shots without overexposing the film, then getting the final exposure (to get the background) right....
That takes a lot of skill that you are not going to be easily able to compress into a Perl script.
First of all, you obviously do not understand how to apply the Nyquist criterion, as the Nyquist limit is set by the BANDWIDTH of the signal of interest. You can take a 100kHz IF and sample it at 16ksample/sec if the signal of interest is only 8kHz wide - indeed this is a common way to take an bandwidth limited IF at a high frequency and both sample it and downshift it at the same time.
Secondly, what I was talking about was quite literally taking the analog IF, shunting it to the main transmitter, and sending it. No sampling required - this is much the same trick that analog AMSATS used to redirect the signal.
AND just to reiterate - I am not saying you do that as the ONLY means of communication with the probe. You have that as an emergency, "AW SHIT" backup mode.
No, in a BPSK receiver you sample between the zero crossings, because you switch phase at the zero crossings. If you switch phase elsewhere, you cause ENORMOUS spectral growth as the phase switch looks like an step function (a.k.a. the first integral of a Dirac impulse).
So, you trigger on the zero crossings, wait a portion of the bit time, then sample the signal.
Yes, I *was* doing SDR in 1997. I was in college in 1987, but this sort of demodulation has been done since the 1960's.
The fact that the slicer couldn't be reflashed in flight is just plain stupid - and that is my point.
And as for sending the raw signals - I did not say "make that the only option" - I said "why is that NOT an option". The whole thing about designing spacecraft is to have as much flexibility as possible so that WHEN the unexpected happens, you have a shot at a work-around.
What I find hard to beleive is that the data slicer for the radio was not a chunk of code running on a processor, rather than a hardwired circuit.
I do SDR (Software Defined Radio) for a living - doing a data slicer like this isn't very hard at all. Why they couldn't just reprogram the slicer to take into account the bit timing shift - or better still, why weren't they resyncing on the zero crossings of the signal so they could deal with bit timing errors automatically?
Hell, for that matter why don't they have an option to route the recovered signal verbatim to the main transmitter and send that to earth - and do the signal processing here? NASA *used* to have the philosophy of "all the bits to earth" - the wouldn't even use lossless data compression lest the signal be corrupted and unrecoverable.
I'd like to see them require a ride-along for license renewal, with three possible outcomes:
FAIL - no renewal for you today, try again later. PASS - OK, you get a new license. See you in a couple of years. CONDITIONAL - IF you PASSed your last exam, renew for one year. If you had a CONDITIONAL or FAIL last time, then FAIL.
Make people have to make an appointment (to cut down on waiting) (AND STAFF THE DAMN DMV TO MEET THOSE SCHEDULES) and make them reschedule not sooner than 2 working days after a failed test (e.g. wait 2 days at least to retest).
So if you are not 100% sure of your skills, you'd better try to renew a few weeks before your license expires.
As I read it, this is a Linux session running in a virtual machine under the host operating system - the idea being that any "sensative" data resides in the virtual session, so the host has no visibility to it.
Except that the host is providing all the screen and keyboard access, so if the host is comprimised and is running VNC the attacker can see where you are going, and what your password is.
True, *IF* the password is only the SSH keyphrase for a private key that is only accessible to the virtual machine, then *maybe* it does him no good.
But since the virtual machine needs to access the media through the (comprimised) host OS, the attacker can copy that data as well.
It sounds to me like this is just giving you a false sense of security.
Wish they would list the Linux titles seperately
on
New Atari Games Revealed
·
· Score: 2, Insightful
I wish they would list those titles that are available under Linux seperately, rather than having a small footnote in the PC-CDROM section.
Robert Heinlein had it right - you can tell a society is on the skids when basic politeness is viewed as weakness, and rudeness is viewed as strength./me looks around pointedly at the/trolls
It never ceases to dismay me how people can scream about how *their* right of "free speach!" is being infringed, and then turn right around and infringe upon the free speech of others.
Supporting the speech of those with whom you agree is NOT supporting free speech. Supporting the right of speech of those with whom you vehemetly disagree IS supporting free speech.
You may feel that Candidate Epsilon-1 is perfection incarnate, and that Candidate Epsilon-2 is distilled evil - if you go around taking down signs for Epsilon-2 you are NOT supporting democracy.
The idea of an autotrain is that you drive your normal, every day car to the train, and put it on the train. You then go forward, sit in a nice comfy seat with 110V available, and ride to your destination. You then get off the train, reclaim your car, and go on your merry way.
The problem is that right now, the Autotrains are only on the east coast, and are running over shared rails (thus they cannot get moving much faster than 100MPH).
The point I am making is that rather than trying to get everybody to upgrade their cars, AND building a dedicated lane for the upgraded cars, AND trying to keep morons out of the dedicated lane with their not-yet-updated cars, you build a rail bed and a smaller number of trains.
And building a dedicated railbed and running an AutoTrain would be even easier, far more efficent, and MUCH faster.
This is the thing that ticks me off about all of this "Intelligent Hiway" crap - we KNOW how to build trains. We KNOW how to build railbeds capable of supporting 300MPH trains. We KNOW how to build rail cars that will hold automobiles. R&D? We need no "R" - the research is done, we just need the development.
However, since we DON'T need any research, nobody wants to look at this technology - it isn't "sexy". So everybody talks about building more intelligence into the car - but of course we will need a huge quantity of money to fund research for those pesky problems like actually dealing with the one driver who's car is NOT on full automatic drive who INSISTS upon getting into that lane.
Just as in maths transforming the problem to a new domain might make it easier to solve, perhaps there is an easier way for you to solve this problem:
Put your MySQL data onto a Logical Volume Manager volume. Use the LVM snapshot mechanism to insure a consistent view of the data during backup. Here's how:
Create an LVM physical group with enough space to contain all your database, plus enough overcapacity to store any changes during the backup proceedure. So, let us say your database contains 10G of data, will take 10 hours to back up, and will, during that time, undergo revision to about 1G of that data - you create a PV of 11G or more.
Next, create a logical volume big enough to store your data - in the case of the numbers above, create a 10G logical volume. In that volume create your file system, and set up your database.
Now, when you wish to do a backup, lock out access to the database, and do a snapshot of the logical volume it is on. Then restore access to the database. This won't take very long at all.
Mount the snapshot read-only and back it up. If you need to back up through the MySQL, bind a server onto the read-only data.
Then release the snapshot.
During the time the snapshot exists (in this example, the 10 hours to do the backup), the real, live read/write file system may be updated as desired, as long as the total differences between it and the snapshot do not exceed the reserve capacity of the physical group - in the case of the example numbers above that would be 1G of total differences (NOT 1G of writes - change the same 1k record a billion times and it is still a 1K change).
That's one of the reasons the Big Boys (like Sun) pooh-poohed Linux - it did not have LVM. Now it does.
But they could achive the same results WITHOUT rendering their pages unusable without Javascript, by the simple expedient of making the text sections visible by default, then hiding them as needed from Javascript.
a) run an open mail relay and see how long until you get shut down b) Use Lynx to do any non-trivial interaction c) Find the latest news using Gopher d) Find the latest documentation on any program, again using Gopher
Basically, this guy is saying that the Internet in its current form won't be around in five years.
I have a saying: "It doesn't matter until it affects the common man - then it will get fixed." It does not matter what "it" is - as long as "it" only affects a small number of folks "it" won't get fixed.
Look back at the old DOS days - when the 640K memory limit only affected high-end users, it didn't matter. When Joe Average started to bump his head, the problem was fixed (largely by the introduction of Windows enhanced mode). Look at spam - now that it affects just about everyone, moves are being made to fix it.
Yes, in five years we the Internet as we know it today won't exist - open SMTP proxies won't be allowed to exist, users will have up-to-date virus protection and firewalls, etc.
Guess what - the Internet as it existed five years ago doesn't exist, either!
Slashdot Mirror
How can I, when your game does not run under Linux?
Just use the CTRL+Mousewheel zoom - click up one, click down one.
The resulting forced redraw will correct the page.
I want to make some bumper stickers for the ricers - after all, we should help these people feel "special"
THE BIGGER THE PIPE, THE SMALLER THE PENIS
HIGH WING, LOW IQ
TYPE-Retarded!
Oh, the fun I could have in your average mall parking lot....
Yes, talk radio for geeks. That's never been done before.
Have to give it a listen.
How can this thing:
Load the needed environment for the compiler.
Load the source
Build the source
Boot the source
in 15 seconds, when it takes much longer than that for my already booted system to build a kernel? A P4 isn't THAT much faster than an AMD3200! (And I have done the old "drop to RL1 and build" trick, so it is not an issue of other tasks running).
I want to know a) What kernel options are enabled b) From when are they starting the clock (are they counting the time to load the bootloader and initrd?) and c) is this TRULY a fully functional kernel, or "just enough to get a prompt"?
'Y'know why it's called "hail"?
Because when you start hearing that "tick. TICK! thunk. THUNK! THUNKTHUNKTHUNK!" you say "OH, HAIL! There goes the roof!"
That's one thing I did not see in the article - how does this stuff fare when you start getting one inch hailstones pelting it.
Then add the 70+ mile an hour winds trying to peel it off.
And if you live in California (excuse me, Kah-lee-FOR-nia), you have the ozone trying to "make friends" with the molecules of the roof. Again, how long will this last?
OK, if you want to put it like that, yes, there is no zero *crossing*.
But the signal still goes to zero(ish), and that is what you use as your locking point, so we were saying the same thing.
Normally, you'd run the signal through a squaring function, then look for signal minima, then feed those into a filtering function to determine the slice point.
Or, if you are looking for a slightly simpler decoder, you have a bit timer that retriggers every bit time. You reset it at any zero crossing (after appropriately narrow band filtering the signal, of course), and sample at 1/2 bit time from when the timer triggers. For noise immunity, I'd sample at 1/4, 1/2, and 3/4 bit time and use a weighted average to determine the bit value. You can deal with a hell of a lot of bit time shift with a decoder like that, and unless you have a run of phase reversals you get enough zero crossings to sync (and you can design your data encoder to provide for those crossings, either with HLDC type encoding, a P/N spreader, or other approaches).
In any case, it't DSP 101 to make a tracking loop for that kind of signal that can deal with enormous bit time errors.
But then again, I design test equipment, so I am always designing my decoders to deal with severely broken encoders.
The whole point of an analemma is to do it on 1 piece of film - no fair taking multiple shots and compositing.
Sure, taking the picture itself can easily be automated.
But fixing a camera to a location so that it will not move DAMMIT (relative to the earth, that is), so that it won't get covered in snow/leaves/pigeondoo/..., so that the film won't be ruined by being out in the elements for a year, being in a location where you can reasonably count on having clear skies enough of the time to get the shots (a month of clouds will really screw you up), being able to judge the exposure needed for the sun shots without overexposing the film, then getting the final exposure (to get the background) right....
That takes a lot of skill that you are not going to be easily able to compress into a Perl script.
The phase transitions happen at the zero crossings - otherwise you get HUGE spectral regrowth.
No, I did not.
First of all, you obviously do not understand how to apply the Nyquist criterion, as the Nyquist limit is set by the BANDWIDTH of the signal of interest. You can take a 100kHz IF and sample it at 16ksample/sec if the signal of interest is only 8kHz wide - indeed this is a common way to take an bandwidth limited IF at a high frequency and both sample it and downshift it at the same time.
Secondly, what I was talking about was quite literally taking the analog IF, shunting it to the main transmitter, and sending it. No sampling required - this is much the same trick that analog AMSATS used to redirect the signal.
AND just to reiterate - I am not saying you do that as the ONLY means of communication with the probe. You have that as an emergency, "AW SHIT" backup mode.
No, in a BPSK receiver you sample between the zero crossings, because you switch phase at the zero crossings. If you switch phase elsewhere, you cause ENORMOUS spectral growth as the phase switch looks like an step function (a.k.a. the first integral of a Dirac impulse).
So, you trigger on the zero crossings, wait a portion of the bit time, then sample the signal.
Yes, I *was* doing SDR in 1997. I was in college in 1987, but this sort of demodulation has been done since the 1960's.
The fact that the slicer couldn't be reflashed in flight is just plain stupid - and that is my point.
And as for sending the raw signals - I did not say "make that the only option" - I said "why is that NOT an option". The whole thing about designing spacecraft is to have as much flexibility as possible so that WHEN the unexpected happens, you have a shot at a work-around.
And that is my point - why did they make it so that they could not reflash it in flight?
And again, why did they not design the slicer to resync on zero crossings, which would have prevented the problem in the first place?
What I find hard to beleive is that the data slicer for the radio was not a chunk of code running on a processor, rather than a hardwired circuit.
I do SDR (Software Defined Radio) for a living - doing a data slicer like this isn't very hard at all. Why they couldn't just reprogram the slicer to take into account the bit timing shift - or better still, why weren't they resyncing on the zero crossings of the signal so they could deal with bit timing errors automatically?
Hell, for that matter why don't they have an option to route the recovered signal verbatim to the main transmitter and send that to earth - and do the signal processing here? NASA *used* to have the philosophy of "all the bits to earth" - the wouldn't even use lossless data compression lest the signal be corrupted and unrecoverable.
AMEN BROTHER!
I'd like to see them require a ride-along for license renewal, with three possible outcomes:
FAIL - no renewal for you today, try again later.
PASS - OK, you get a new license. See you in a couple of years.
CONDITIONAL - IF you PASSed your last exam, renew for one year. If you had a CONDITIONAL or FAIL last time, then FAIL.
Make people have to make an appointment (to cut down on waiting) (AND STAFF THE DAMN DMV TO MEET THOSE SCHEDULES) and make them reschedule not sooner than 2 working days after a failed test (e.g. wait 2 days at least to retest).
So if you are not 100% sure of your skills, you'd better try to renew a few weeks before your license expires.
OK, let's think this through:
As I read it, this is a Linux session running in a virtual machine under the host operating system - the idea being that any "sensative" data resides in the virtual session, so the host has no visibility to it.
Except that the host is providing all the screen and keyboard access, so if the host is comprimised and is running VNC the attacker can see where you are going, and what your password is.
True, *IF* the password is only the SSH keyphrase for a private key that is only accessible to the virtual machine, then *maybe* it does him no good.
But since the virtual machine needs to access the media through the (comprimised) host OS, the attacker can copy that data as well.
It sounds to me like this is just giving you a false sense of security.
I wish they would list those titles that are available under Linux seperately, rather than having a small footnote in the PC-CDROM section.
Robert Heinlein had it right - you can tell a society is on the skids when basic politeness is viewed as weakness, and rudeness is viewed as strength. /me looks around pointedly at the /trolls
It never ceases to dismay me how people can scream about how *their* right of "free speach!" is being infringed, and then turn right around and infringe upon the free speech of others.
Supporting the speech of those with whom you agree is NOT supporting free speech. Supporting the right of speech of those with whom you vehemetly disagree IS supporting free speech.
You may feel that Candidate Epsilon-1 is perfection incarnate, and that Candidate Epsilon-2 is distilled evil - if you go around taking down signs for Epsilon-2 you are NOT supporting democracy.
If you are falling asleep on short distances, you have other problems that need addressing.
No, look here
The idea of an autotrain is that you drive your normal, every day car to the train, and put it on the train. You then go forward, sit in a nice comfy seat with 110V available, and ride to your destination. You then get off the train, reclaim your car, and go on your merry way.
The problem is that right now, the Autotrains are only on the east coast, and are running over shared rails (thus they cannot get moving much faster than 100MPH).
The point I am making is that rather than trying to get everybody to upgrade their cars, AND building a dedicated lane for the upgraded cars, AND trying to keep morons out of the dedicated lane with their not-yet-updated cars, you build a rail bed and a smaller number of trains.
And building a dedicated railbed and running an AutoTrain would be even easier, far more efficent, and MUCH faster.
This is the thing that ticks me off about all of this "Intelligent Hiway" crap - we KNOW how to build trains. We KNOW how to build railbeds capable of supporting 300MPH trains. We KNOW how to build rail cars that will hold automobiles. R&D? We need no "R" - the research is done, we just need the development.
However, since we DON'T need any research, nobody wants to look at this technology - it isn't "sexy". So everybody talks about building more intelligence into the car - but of course we will need a huge quantity of money to fund research for those pesky problems like actually dealing with the one driver who's car is NOT on full automatic drive who INSISTS upon getting into that lane.
Just as in maths transforming the problem to a new domain might make it easier to solve, perhaps there is an easier way for you to solve this problem:
Put your MySQL data onto a Logical Volume Manager volume. Use the LVM snapshot mechanism to insure a consistent view of the data during backup. Here's how:
Create an LVM physical group with enough space to contain all your database, plus enough overcapacity to store any changes during the backup proceedure. So, let us say your database contains 10G of data, will take 10 hours to back up, and will, during that time, undergo revision to about 1G of that data - you create a PV of 11G or more.
Next, create a logical volume big enough to store your data - in the case of the numbers above, create a 10G logical volume. In that volume create your file system, and set up your database.
Now, when you wish to do a backup, lock out access to the database, and do a snapshot of the logical volume it is on. Then restore access to the database. This won't take very long at all.
Mount the snapshot read-only and back it up. If you need to back up through the MySQL, bind a server onto the read-only data.
Then release the snapshot.
During the time the snapshot exists (in this example, the 10 hours to do the backup), the real, live read/write file system may be updated as desired, as long as the total differences between it and the snapshot do not exceed the reserve capacity of the physical group - in the case of the example numbers above that would be 1G of total differences (NOT 1G of writes - change the same 1k record a billion times and it is still a 1K change).
That's one of the reasons the Big Boys (like Sun) pooh-poohed Linux - it did not have LVM. Now it does.
But they could achive the same results WITHOUT rendering their pages unusable without Javascript, by the simple expedient of making the text sections visible by default, then hiding them as needed from Javascript.
OK, so why don't you:
a) run an open mail relay and see how long until you get shut down
b) Use Lynx to do any non-trivial interaction
c) Find the latest news using Gopher
d) Find the latest documentation on any program, again using Gopher
Basically, this guy is saying that the Internet in its current form won't be around in five years.
I have a saying: "It doesn't matter until it affects the common man - then it will get fixed." It does not matter what "it" is - as long as "it" only affects a small number of folks "it" won't get fixed.
Look back at the old DOS days - when the 640K memory limit only affected high-end users, it didn't matter. When Joe Average started to bump his head, the problem was fixed (largely by the introduction of Windows enhanced mode). Look at spam - now that it affects just about everyone, moves are being made to fix it.
Yes, in five years we the Internet as we know it today won't exist - open SMTP proxies won't be allowed to exist, users will have up-to-date virus protection and firewalls, etc.
Guess what - the Internet as it existed five years ago doesn't exist, either!