since it is lacking a lot of the CPU code for task switching... i.e. there was no memory protection, any program could write to anywhere in memory. (which spawned a million keyboard-hook malwares)
Memory protection is completely separate from multitasking. For example, Microware OS-9 (from the 1980's, not to be confused with
the much-later Mac OS9) ran on plain old 6809's and did true multitasking but without memory protection.
"Fake" multitasking is what early Macs and MSWindows did. It was also known as "cooperative multitasking". With that, when your
process makes a system call, the OS might take the opportunity to switch to another task instead of immediately finishing the call. The fundamental problem with this is that if some code is stuck in a lengthy loop where it doesn't make system calls, no other tasks can run. A trivial infinite loop would effectively lock up your system. Anyway, the end result to the user was very "jerky" interaction with the GUI since the taskswitching was so coarse and unpredictable.
"Real" multitasking is transparent to each process. Every now and then the CPU will just save the state of a given process and switch to running another process. Eventually it'll restore the state of the original process and pick up where it left off. This is normally done off a timed interrupt.
So an infinite loop won't effectively hang the system, plus the scheduler can (for example) allocate fewer timeslices to processes that do not require user interaction.
Exactly. I want the expert to make the choices. Pick the best things that go well together. If I have a strong opinion on something else then it will be up to me to install that on top of my EASY INSTALL.
The editions are really just about eye candy. Think of it as different "windows explorer" interfaces.
The apps will all be pretty well the same.
I suggest you download the live CDs of KDE and Gnome and just boot from CD and see which one you like best before installing.
If you want things simple, you'll probably like Gnome. I notice that that is what they suggest as the "MAIN EDITION" too.
I think you're being a bit unfair complaining about the 32/64 choice. They can't possible
know if you have 32bit or 64 bit hardware, so they provide bot, just like MS does with windows.
Note that the 32 bit version will work fine on a 64 bit CPU but you're not getting the speed advantage
of your fancy CPU (64 bit code typically runs around 20% faster).
Because every STS mission has different launch parameters. Different inclination, different altitude, eccentricities and orbital velocity, not
to mention that since the exact launch time is unpredictable you're never even going to be anywhere near the RA of any previous tank.
You'll end up with a swarm of ETs in all kinds of different orbits, effectively completely un-collectible.
Now, because we're talking rocket scientists here it's not like it never occurred to them to think of ways to use the tanks, at least on a per-mission basis.
See here for a list of ideas.
Mostly it has to do with making use of the unused fuel and using tethering to steal momentum from the tank so as to increase
the effective payload.
My guess is that they decided not to bother because the payload boost payoff would generally be too unpredictable, which
means that if you estimated wrong you've blown pretty well the whole mission. So since you're never realistically going to want
to rendezvous with an old tank ever again (since you're always bringing up a new one), the only sensible thing is to bring them
out of harm's way.
At the same time we have the Fermi Paradox. Where are they? At our current rate of expansion and technological development we will have colonized the galaxy shortly (universe time-scale).
If there is intelligent life on only a handful of the viable planets, given that some of that life necessarily have several millions years of a head start on us, the first thing that Galileo should have seen when looking up was
"Welcome to the Galaxy humans.
This message sponsored by MickyD's, 25 trillion served each galactic standard day".
The simplest answer to the Fermi Paradox is that faster-than-light travel is not possible.
If it's going to take a century to go ten light years, galactic colonization is just not going to happen.
All societies eventually end up in a stagnant state, with countless generations doing the holodeck equivalent
of watching sports. No doubt there will be a few that have been forced into the "Ark" scenario,
but that's just moving house, not expanding.
You might like Neal Stephenson's "Anathem". Well, or you might hate it.
Is there other intelligent life out there ? I don't know.
Well consider this: there are around 10**20 stars in the universe. If you were to consider every star as a grain of sand,
that would make for a cube of sand 5km in each direction, which is considered to be about as much as all the sand on all the beaches
on planet Earth
So, is our star system the only grain of sand with life on it? That strikes me as unlikely no matter how unlikely you consider the development of life to be.
I am reading (and responding) to this on a Nokia E90, and it sure seems pretty usable... the screen is 800x352 and it has a decent keyboard. Actually it has two browsers; the default Nokia one and Opera.
GPS, wifi, bluetooth, decent camera... it's a nice little unit.
Thank you. Nowhere in all this fuss have I seen mentioned that the Larsson books are very, very bad: ill-written, tedious, preposterous, paper-thin characters and highly misogynist despite their feminist pretensions. Books are like money: the bad drives out the good.
If it's so bad, how about you write something equal or better and subsequently become a multimillionaire? Just do it for the money; afterwards you can write all the marvellously artistic work that plebes don't understand all you want.
For IR, it's because we're warm blooded. If our eyes could detect far-IR (like a FLIR camera), it would see nothing but a white fog since the eye is already at that temperature.
Snakes can sense far-IR because they are cold blooded. Even then, they don't do it that well because the temperature difference between their sensors (which are in the nose) and a mammal isn't that much.
As for near-IR, it's just not useful. It makes for pretty photography effects, but it doesn't help you at night or day.
Metropolis is a composite, but it was originally inspired after Toronto
Re:Vim most definitely can't "do everything"
on
Hacking Vim 7.2
·
· Score: 2, Informative
My main peeve with vim is when it goes into "recording mode" or whatever that nonsense is, when I try to ":q" and hit something by mistake and the screen splits. I still don't know what I did wrong. And that vim always wants to go back to the last place in the file you edited, even if you edited the file ten years ago and really want to start at the top. The only way I found around that one is to make the.viminfo file chmod 000 so it can't save status.
Recording mode is your friend. It means you hit "q" followed by some letter. It's a macro recorder. Let's say you want to do some complicated operation on a number of lines. Type "qq" to start recording into "q" and do the commands in a generalized fashion on one line, making sure that at the end you end up with the cursor on the next line. Type "q" to stop recording. Now type "@q" to execute those commands again, and the current line should get fixed. Or type "20@q" to do it 20 times. You get the idea.
Splitting the windows is done with the ^W commands. ^Ws splits horizontally, ^Wv vertically. Do ^Wq to quite a split window. Or do ^Wo to make the current window the "o"nly window.
I'm a bit puzzled about why you don't like vi remembering where you last edited. If you want to go to the start of the file, just type "1G".
Re:"Steep" learning curve
on
Hacking Vim 7.2
·
· Score: 4, Informative
This is one of my pet peeves.
A steep learning curve refers to something that is quickly learned, as the curve that represents knowledge over time would indeed be steep in that case.
Something difficult would have a shallow learning curve, not a steep one.
I think usually when people talk about learning curves, the horizontal axis represents knowledge and the vertical axis is your investment to acquire that knowledge.
Is there some established convention that contradicts this?
The problem is that in mathematics, the independent variable normally goes on the X axis (horizontal). In the case of learning, "effort" is the independent variable since "knowledge acquired" depends on "effort" and not the other way around. By that model a steep learning curve is one where a little bit of effort gains you a lot of knowledge. Unfortunately this doesn't match the classic usage of the phrase.
And that's why I don't use that phrase. Instead I say things like "python has a gentle learning curve" or "emacs has a tough learning curve"
It is true, however, that cosmic ray collisions completely kill the "LHC will destroy the world" bullshit.
Ah, but you forget that cosmics ray hadrons are natural and organically grown, unlike
those nasty synthetic LHC ones which cause obesity, cancer and black holes. Plus they don't taste as good.
No -- if you put a 5.56 round into a.223 chambered weapon, you run the risk of blowing up the weapon. The chamber specs are different.
The bullet is the same diameter (.224"), but the chamber is different, which means it is dangerous to interchange them -- and incorrect to state that they are the same, which is why I called you on it.
I know nothing about firearms so please bear with me.
5.56mm works out to.219 inches, so a 5.56mm round should be slightly smaller than.223 or.224, though only by about a tenth of a mm. So getting stuck in the barrel doesn't seem like an issue.
What is it about chamber design that could risk rupture? As long as the round and associated gas pressure can exit, I can't see what could cause the weapon to blow up.
Of course, I have an E90 (a "flagship" phone) and of course the bastards at Nokia aren't supplying free navigation for that. I just grabbed the latest, Ovi Maps 3.01, and there's no free nav.
That's what I get for being stupid enough to pay $600 for a premium phone: I get nickled and dimed for every little feature
200m is the absolute accuracy (and is a bit pessimistic). The repeatable accuracy is much better.
That is, if you sail into a port's harbour channel and save that as a LORAN-C waypoint you will typically be able to get back to that same spot within 20m or so easily.
Put simply, because the names microcell and picocell were already taken.
The names are not meant in the traditional mathematical sense; they just refer to coverage. A microcell will cover roughly a hotel, a picocell a typical office floor.
You can do that in Java too. Just define a method that takes a HashMap as its single argument and have the method extract "arguments" from that HashMap by name.
That's essentially what the bozo who wrote your code example did, for no apparent reason.
Slashdot Mirror
That's nice, but this particular obfuscation makes the script bigger and slower
...but the "this is a stupid waste of money" buzzer in my head went off.
since it is lacking a lot of the CPU code for task switching... i.e. there was no memory protection, any program could write to anywhere in memory. (which spawned a million keyboard-hook malwares)
Memory protection is completely separate from multitasking. For example, Microware OS-9 (from the 1980's, not to be confused with the much-later Mac OS9) ran on plain old 6809's and did true multitasking but without memory protection.
"Fake" multitasking is what early Macs and MSWindows did. It was also known as "cooperative multitasking". With that, when your process makes a system call, the OS might take the opportunity to switch to another task instead of immediately finishing the call. The fundamental problem with this is that if some code is stuck in a lengthy loop where it doesn't make system calls, no other tasks can run. A trivial infinite loop would effectively lock up your system. Anyway, the end result to the user was very "jerky" interaction with the GUI since the taskswitching was so coarse and unpredictable.
"Real" multitasking is transparent to each process. Every now and then the CPU will just save the state of a given process and switch to running another process. Eventually it'll restore the state of the original process and pick up where it left off. This is normally done off a timed interrupt. So an infinite loop won't effectively hang the system, plus the scheduler can (for example) allocate fewer timeslices to processes that do not require user interaction.
Exactly. I want the expert to make the choices. Pick the best things that go well together. If I have a strong opinion on something else then it will be up to me to install that on top of my EASY INSTALL.
The editions are really just about eye candy. Think of it as different "windows explorer" interfaces. The apps will all be pretty well the same.
I suggest you download the live CDs of KDE and Gnome and just boot from CD and see which one you like best before installing. If you want things simple, you'll probably like Gnome. I notice that that is what they suggest as the "MAIN EDITION" too.
I think you're being a bit unfair complaining about the 32/64 choice. They can't possible know if you have 32bit or 64 bit hardware, so they provide bot, just like MS does with windows. Note that the 32 bit version will work fine on a 64 bit CPU but you're not getting the speed advantage of your fancy CPU (64 bit code typically runs around 20% faster).
Why would they have to be uncontrolled?
Because every STS mission has different launch parameters. Different inclination, different altitude, eccentricities and orbital velocity, not to mention that since the exact launch time is unpredictable you're never even going to be anywhere near the RA of any previous tank. You'll end up with a swarm of ETs in all kinds of different orbits, effectively completely un-collectible.
Now, because we're talking rocket scientists here it's not like it never occurred to them to think of ways to use the tanks, at least on a per-mission basis. See here for a list of ideas. Mostly it has to do with making use of the unused fuel and using tethering to steal momentum from the tank so as to increase the effective payload.
My guess is that they decided not to bother because the payload boost payoff would generally be too unpredictable, which means that if you estimated wrong you've blown pretty well the whole mission. So since you're never realistically going to want to rendezvous with an old tank ever again (since you're always bringing up a new one), the only sensible thing is to bring them out of harm's way.
It's BRAKES, not breaks. Get it right, people!
At the same time we have the Fermi Paradox. Where are they? At our current rate of expansion and technological development we will have colonized the galaxy shortly (universe time-scale).
If there is intelligent life on only a handful of the viable planets, given that some of that life necessarily have several millions years of a head start on us, the first thing that Galileo should have seen when looking up was "Welcome to the Galaxy humans. This message sponsored by MickyD's, 25 trillion served each galactic standard day".
The simplest answer to the Fermi Paradox is that faster-than-light travel is not possible. If it's going to take a century to go ten light years, galactic colonization is just not going to happen.
All societies eventually end up in a stagnant state, with countless generations doing the holodeck equivalent of watching sports. No doubt there will be a few that have been forced into the "Ark" scenario, but that's just moving house, not expanding.
You might like Neal Stephenson's "Anathem". Well, or you might hate it.
Is there other intelligent life out there ? I don't know.
Well consider this: there are around 10**20 stars in the universe. If you were to consider every star as a grain of sand, that would make for a cube of sand 5km in each direction, which is considered to be about as much as all the sand on all the beaches on planet Earth
So, is our star system the only grain of sand with life on it? That strikes me as unlikely no matter how unlikely you consider the development of life to be.
I am reading (and responding) to this on a Nokia E90, and it sure seems pretty usable... the screen is 800x352 and it has a decent keyboard. Actually it has two browsers; the default Nokia one and Opera.
GPS, wifi, bluetooth, decent camera... it's a nice little unit.
Thank you. Nowhere in all this fuss have I seen mentioned that the Larsson books are very, very bad: ill-written, tedious, preposterous, paper-thin characters and highly misogynist despite their feminist pretensions. Books are like money: the bad drives out the good.
If it's so bad, how about you write something equal or better and subsequently become a multimillionaire? Just do it for the money; afterwards you can write all the marvellously artistic work that plebes don't understand all you want.
Put up or shut up.
For IR, it's because we're warm blooded. If our eyes could detect far-IR (like a FLIR camera), it would see nothing but a white fog since the eye is already at that temperature.
Snakes can sense far-IR because they are cold blooded. Even then, they don't do it that well because the temperature difference between their sensors (which are in the nose) and a mammal isn't that much.
As for near-IR, it's just not useful. It makes for pretty photography effects, but it doesn't help you at night or day.
Metropolis is a composite, but it was originally inspired after Toronto
My main peeve with vim is when it goes into "recording mode" or whatever that nonsense is, when I try to ":q" and hit something by mistake and the screen splits. I still don't know what I did wrong. And that vim always wants to go back to the last place in the file you edited, even if you edited the file ten years ago and really want to start at the top. The only way I found around that one is to make the .viminfo file chmod 000 so it can't save status.
Recording mode is your friend. It means you hit "q" followed by some letter. It's a macro recorder. Let's say you want to do some complicated operation on a number of lines. Type "qq" to start recording into "q" and do the commands in a generalized fashion on one line, making sure that at the end you end up with the cursor on the next line. Type "q" to stop recording. Now type "@q" to execute those commands again, and the current line should get fixed. Or type "20@q" to do it 20 times. You get the idea.
Splitting the windows is done with the ^W commands. ^Ws splits horizontally, ^Wv vertically. Do ^Wq to quite a split window. Or do ^Wo to make the current window the "o"nly window.
I'm a bit puzzled about why you don't like vi remembering where you last edited. If you want to go to the start of the file, just type "1G".
This is one of my pet peeves.
A steep learning curve refers to something that is quickly learned, as the curve that represents knowledge over time would indeed be steep in that case.
Something difficult would have a shallow learning curve, not a steep one.
I think usually when people talk about learning curves, the horizontal axis represents knowledge and the vertical axis is your investment to acquire that knowledge.
Is there some established convention that contradicts this?
The problem is that in mathematics, the independent variable normally goes on the X axis (horizontal). In the case of learning, "effort" is the independent variable since "knowledge acquired" depends on "effort" and not the other way around. By that model a steep learning curve is one where a little bit of effort gains you a lot of knowledge. Unfortunately this doesn't match the classic usage of the phrase.
And that's why I don't use that phrase. Instead I say things like "python has a gentle learning curve" or "emacs has a tough learning curve"
You're mixing up representation systems with voting systems. What Canada has is "regional representation", and they use FPTP as the voting system.
You can keep the regional representation MP system and get rid of FPTP. For example, use the single transferable vote system when voting for your MP.
Do they even read what they write?
"OO.o Writer is the fastest and most responsive word processor available for Linux today."
"KWord is fast. It's probably the fastest-loading and maybe the most responsive word processor in the roundup."
Apparently they didn't try AbiWord then.
On my desktop box, it's up an running within a fraction of a second. Yes really.
It is true, however, that cosmic ray collisions completely kill the "LHC will destroy the world" bullshit.
Ah, but you forget that cosmics ray hadrons are natural and organically grown, unlike those nasty synthetic LHC ones which cause obesity, cancer and black holes. Plus they don't taste as good.
It will probably work with the new WIND mobile network though. They use UMTS AWS band IV (1700/2100). My Nokia E90 picks up the 2100 downlink.
Personally I use the Wifi on the E90 quite a bit. VoIP calling is really cheap.
No -- if you put a 5.56 round into a .223 chambered weapon, you run the risk of blowing up the weapon. The chamber specs are different.
The bullet is the same diameter (.224"), but the chamber is different, which means it is dangerous to interchange them -- and incorrect to state that they are the same, which is why I called you on it.
I know nothing about firearms so please bear with me.
5.56mm works out to .219 inches, so a 5.56mm round should be slightly smaller than .223 or .224, though only by about a tenth of a mm. So getting stuck in the barrel doesn't seem like an issue.
What is it about chamber design that could risk rupture? As long as the round and associated gas pressure can exit, I can't see what could cause the weapon to blow up.
It's the free navigation that's news.
Of course, I have an E90 (a "flagship" phone) and of course the bastards at Nokia aren't supplying free navigation for that. I just grabbed the latest, Ovi Maps 3.01, and there's no free nav.
That's what I get for being stupid enough to pay $600 for a premium phone: I get nickled and dimed for every little feature
200m is the absolute accuracy (and is a bit pessimistic). The repeatable accuracy is much better.
That is, if you sail into a port's harbour channel and save that as a LORAN-C waypoint you will typically be able to get back to that same spot within 20m or so easily.
The names are not meant in the traditional mathematical sense; they just refer to coverage. A microcell will cover roughly a hotel, a picocell a typical office floor.
You can do that in Java too. Just define a method that takes a HashMap as its single argument and have the method extract "arguments" from that HashMap by name. That's essentially what the bozo who wrote your code example did, for no apparent reason.
You can do stupid things in any language.
Over 40? Don't you mean over 80?
I work in electronics, and this 65 year old guy is the master of dealing with quasimicroscopic parts. I've found this is the norm, not the exception.
Basically, most young guys are incredibly clumsy. They have the physical potential but usually completely lack control.