Off the top of my head, the english language doesn't have instructions for getting the operating system to output a line of text to the console and then terminate execution.
"Some lazy programs would have exploited the aliases and have broken code"
Which is why we still need to enable the A20 line on x86 processors when using >1MB RAM, as the very high end of the 1MB address space did(/does) alias the bottom end.
Really? I learnt it at a rather young age, and yes I guess it is said that we can learn languages better at earlier ages, but even so... what's really that complicated about it? I managed to hand code a preemptive multitasking protected mode kernel in x86 assembly, all self taught... so I'm at a bit of a loss understanding your post (and the others of that variant).
There's a difference between can-be-ported and has-been-ported. How many people would buy vista to install on a ppc? MS obviously don't believe it's enough to be worth creating a release for it.
Yup, since as early as I remember (3.2 days) anyway. I use dos for my bootloaders, a small fat16 partition which can hold various kernels (for rolling back to if need be) initrd's etc. If installing a new kernel remotely, not being 100% sure something somewhere isn't gonna go wrong, and knowing the people with access to the machine are really not gonna be of any help, can set the bootloader script to rewrite itself before jumping into the kernel, and then assuming it boots okay, you can write it back to load the new kernel everytime. If it doesn't come back up, all I gotta do is ask someone to press reset, and it'll boot with the replaced script that's set to boot the previous kernel, and i get access back. (Sometimes reboots can be free, but if you want them to plug in a keyboard/monitor to navigate a boot menu, they wanna charge ya). Might be other ways of doing it (I wanna look into kexec crash kernels for instance, not had chance yet) but this way is easy, it works, and I know it already, even if it does make a lot of people go "eugh":-)
Yes the wording is pretty bad, but... you are joking right? "The FBI's act of wiretapping is auditing something" would be "Wiretapping audits [thing]"... "wiretapping audit" is clearly an audit of the wiretapping, as a "security audit" would be an audit of the security of something (which these guys didn't do!)... so that's a noun... it's a thing. It might be something that happens, like an eclipse, but it's not a verb, unless it's eclipsing, eclipsed. And secrets... I can't even think off the top of my head how that could be used in verb form. A secret is a thing that you know... or, more likely, don't know... or was that a joke via the 'secretes'? Hard to tell, not much of this is making sense at all:-/
Just GUI, as Linux. Command line, ctrl+c is break and will also abort many running processes (like a file copy, an oh-my-god-I-didn't-realise-there-were-this-many-files-in-here directory listing etc etc) or if the running process is ignoring it, it'll often be picked up between commands in a running batch file, where you'll get a prompt asking if you wish to terminate the script or continue. If you set 'break on' however, then you get more luck with ctrl+break than ctrl+c, as attention is usually paid to that in more places. Ctrl+Z however is the end-of-file marker, and so will stop text copies coming in from console (eg, 'copy con output.txt' will start copying text from console to output.txt, until you send a line with ctrl+z in... believe me, it can have its uses!)
Or working in vi running in a remote screen session from putty... each one having their own clipboard... very rarely I get mixed up, and if I do it's usually an undo key back... unless some reason I have a load of text or something in one clipboard (like accidentally selected something at some point) and I click the mouse button and just see loads of lines about 'command not found' as it tries to run every line that was in the clipboard, no stopping it, until it stops and begins chunking away at the harddrive and I spot the text 'rm -rf *' as the last line on the screen... if the machine's local I can hit the reset switch hopefully before the write behind cache catches up, but if not, I just sit back, relax, and thank the god I don't believe in that this has never actually happened to me (why would I have rm -rf in a text document? Maybe explaining to someone... I'm never going to write that again without at least a # in front of it just in case it does happen tho!)
Err... I can only assume 'cat' was compiled for a different processor, but there must be emulation or something... I dunno... it just said "illegal instruction", but carried on anyway???
"use OSS if some viable candidate exists. If it doesn't, all I'm arguing is that using something like Skype [...] is a sensible stop-gap"
I wouldn't hold your breath, with the average F/OSS developer, programming for fun, stuff they actually want to write, but looking like this, I can't see them putting too much effort into video conferencing software:-p
I believe locks are already fairly fine grained in most of the system. Problems occur when you have multi level locks (eg, you want to change the value in a structure/object, so you get a lock on the structure, and then copy (duplicate) it, which could involve needing a lock on the table of structures before you can insert a new one. The lock on the table could be harder to get; you may have to wait for certain types of locks on the structures within to be released, but you can't release the lock on the object you're holding until you've created the copy, as you want the copy to be of the object in that state, and something else might change it before you get the lock on the table.
You can minimise the effects, but locking something in always implies the chance you are locking something else out. Where possible, lockless algorithyms can be switched to. One example is creating per-cpu versions of things. For example, a counter which goes up everytime something happens. A process on a CPU could get a lock, read the value, perhaps check it against an upper limit, increment and write it back, then release the lock, to make sure other processors don't change the value in between reading it and writing the new value back, which would lose the value the other process had updated it to. Instead, give each cpu its own counter, and when you want to know the value, you just add up all the values of all the counters.
Another method is using copy-on-write and reference counting. If you want to change a structure which something else has a read lock on, the structure is copied, and you change that. Pointers are updated so that anything that reads from it in the future reads from the new copy, only processes with the lock on the old copy read from that, and when there are no more left, the old copy is released and memory freed.
Both of these methods require that you can have more than one of what you'd want to lock, but obviously in some cases this just isn't true, because there is only one of the resource and it can't be copied.
You've gotta know that's complete nonsense, right? You wanna create an abstraction layer between every single bit of code in the kernel and the hardware it's meant to be running on to handle locking dynamically, making decisions at runtime, rather than expecting developers to be able to design a system and make those decisions for the code outside of its runtime and having the best way compiled to run directly on the hardware at full speed?
Managed languages have their place (and in fact all my work now is with managed languages)... but they also have to run on something. You think your java apps will just run without an operating system underneith it with these kinds of features to support the running of the virtual machine? What language do you think your java machine is written in? Do you think that's air you're breathing now? Hmm...
"This is innacurate, not correct and misleading at the same time"
But hey at least you prefixed what you had to say with this so we did know.
I'm pretty sure Windows has something like this, I recall seeing something in howto's about running different versions of IE along side each other, by creating an empty.local file with the same name as the executable (so in this case would've been iexplore.exe.local) and that affects the shared library loading path, so it will load dlls that are in the same directory before using others elsewhere in the system. Not quite the same, but could be used to achieve the same ends.
What part of the word "prefer" to you implies an assumed dichotomy? Surely, prefer is a conscious tendancy to lean towards one of the two oposites, which to me implies lack of a dichotomy not the presence of one, as with a dichotomy, there's no 'leaning towards', because you're either that or you're not.
"I trust mom, but I wouldn't tell her my passwords. I don't "hide" from her"
The "activeness" is the point though, surely. You might not announce to her your passwords, but you don't have to look over your shoulder every time you log in somewhere to make sure she's not trying to sneak a peak at your passwords right? I know I have no interest in finding out my mom's passwords, and she has no interest in finding out mine. She could even see my typing them in, but she'd have no interest in committing to memory each keypress in which order, so it's not something I have to worry about.
"I trust grandma, but I wouldn't necessarily tell her each time I took a taxi to the railway station"
But you don't have to look over your shoulder each time you jump in a cab to make sure she's not watching? Assuming she's able to get out still (okay this might not be the case, but for the sake of demonstrating a point...) it wouldn't be a problem if she did coincidentally see you taking a taxi one time? Unless she was following you around everywhere so saw you taking lots of taxis, but if that was the case, you'd in fact have other problems, and as we have laws against stalking already, I don't think we need new rules or anything to cover that scenario.
So as you I hope can see the point I'm making... privacy can exist in a mutual exclusion sense, rather than it being "something I must protect or even fight for with laws/regulations/cryptography". Things I do prefer to do in private (take obvious more extreme example, going to the toilet) I don't have to worry that people might be around, because I know they have no interest in watching that, and vacate of their own occord. In the same way, I don't need laws protecting my phone records from people I know, because I know they have no interest in looking at them, and people who don't know me... well they don't know me, they're not exactly going to be going through my phone records! Maybe collecting stats, but a computer is more likely to be used for that, and there's no personal link, they're not gonna be sniggering at the fact I called one person lots for a month then stopped, so I don't see how that would be a privacy issue.
Privacy works best through mutual voluntary exclusion, enforced exclusion I don't see as being too much of a requirement.
That's gotta be the most stupid selfish pathetic thing I've ever heard, and trust me, you've had a lot of competition. So basically you think that people phoning you is rude, so you won't allow that to happen, and only turn your phone on to inflict on other people what you wouldn't have them do to you themselves? *lol*
Also, this whole "people just let phones rudely interrupt them" thing that people come out with... what about if it wasn't a phone? What if it was someone knocking at the door? What would you do, ignore that? Or does someone knocking at the door have you well trained? What's so bad about being talked to?
Sure, I don't too much like listening to people shouting, whether they're on their phone or not, but generally I find the main cause of shouting to be stupidity or deafness... if we could ban stupidity that would be great.
"and do not have the compulsive need to be constantly talking to anyone about anything at all times"
No compulsive need here... but I have to say, I do like the fact that my friends know I'm there for them, wherever we are, if something goes wrong, they never have to deal with it alone, or if something good happens, they can always share it.
"To me, all these modern devices have made communication too easy and cheapened it"
I'm sorry that that has been your experience. Mine's been the oposite, and with it, I get the joy of watching a circle of good people strengthen as we all become more supportive of each other. It can be difficult for new people coming in, after being used to living more private/closed lives, but generally once the realisation comes that we're non judgemental (well, most of us), non manipulative people, the sharing inhibitions go, because there's no need for excessive privacy beyond basic decency. The proud don't feel the need to hide away, which more often than not, can give you more chance to have more to be proud about.
I don't understand the complains about people talking on their phones. What if the other person they were talking to was right there? Still talking about the same things, still as much in your ear space... just that your brain's not trying to listen in to figure out the other side of the conversation, which is why people really have a problem with mobile phones.
Lets say I'm just a crazy person who's attracted to bright lights, and the feeling of a sharp knife going though soft flesh... better turn off all your lights too then!
If you're really that desperate for privacy, turn your phone off! For anyone not suffering delusions of grandure (yeah so satalites can read your clothes labels from space, but who thinks they're really that interesting?) and realises that we buy phones because most of us are social creatures rather than paranoid conspiracy nuts, and actually like being able to make contact with other human beings.
"I usually find it best to err on the side of privacy rather than safety"
Well, I have friends, we share our lives with each other, and would much prefer we look out for each other than live our lives in fear and have to hide from each other. Maybe we're just unique like that.
"To me, "packet" implies that you can break it open and there's something smaller inside"
I guess, perhaps the connotations attached to the word for me that come from networking (where you send/received packets of data) have altered my internal meaning... sure you can look inside a network packet (kind of a requirement!) but it can't go smaller than a certain size, it needs to be big enough to carry it's information, such as the address it's going to, and state information (eg to make sure it won't travel forever if it gets caught in a loop). Any smaller and it can't exist (not as a packet on the network at least). A packet of light has some similarities... it can't exist any smaller, not because it's carrying physical objects, but it does also carry information (wavelength is information, but requires energy to exist)... and in fact quantum cryptography is based on being able to read quantum states of individual (well, entangled) packets of light. So no, we may not be able to physically split it up, but there is more in a packet of light than err.... meets the eye?
But I think we're probably entering more into the realms of language than physics now:-)
Um... yes? Is that a weird thing? I have walls, inside my house, that are made out of bricks. I guess a lot of places just use timber and plasterboard for internal walling, but no, my house, internal and exteral walls, built with bricks. Very solid thing. 100+ years old already, built to last.
Slashdot Mirror
Off the top of my head, the english language doesn't have instructions for getting the operating system to output a line of text to the console and then terminate execution.
iow, "I'm gonna go make my own processor! With blackjack, and hookers!"
"Some lazy programs would have exploited the aliases and have broken code"
Which is why we still need to enable the A20 line on x86 processors when using >1MB RAM, as the very high end of the 1MB address space did(/does) alias the bottom end.
Really? I learnt it at a rather young age, and yes I guess it is said that we can learn languages better at earlier ages, but even so... what's really that complicated about it? I managed to hand code a preemptive multitasking protected mode kernel in x86 assembly, all self taught... so I'm at a bit of a loss understanding your post (and the others of that variant).
"Well then, where's Vista for PowerPC?"
There's a difference between can-be-ported and has-been-ported. How many people would buy vista to install on a ppc? MS obviously don't believe it's enough to be worth creating a release for it.
"I mean how would they even open a window to jump in the first place?"
Airplane?
Yup, since as early as I remember (3.2 days) anyway. I use dos for my bootloaders, a small fat16 partition which can hold various kernels (for rolling back to if need be) initrd's etc. If installing a new kernel remotely, not being 100% sure something somewhere isn't gonna go wrong, and knowing the people with access to the machine are really not gonna be of any help, can set the bootloader script to rewrite itself before jumping into the kernel, and then assuming it boots okay, you can write it back to load the new kernel everytime. If it doesn't come back up, all I gotta do is ask someone to press reset, and it'll boot with the replaced script that's set to boot the previous kernel, and i get access back. (Sometimes reboots can be free, but if you want them to plug in a keyboard/monitor to navigate a boot menu, they wanna charge ya). Might be other ways of doing it (I wanna look into kexec crash kernels for instance, not had chance yet) but this way is easy, it works, and I know it already, even if it does make a lot of people go "eugh" :-)
only to a pessimist ;-)
Yes the wording is pretty bad, but... you are joking right? "The FBI's act of wiretapping is auditing something" would be "Wiretapping audits [thing]"... "wiretapping audit" is clearly an audit of the wiretapping, as a "security audit" would be an audit of the security of something (which these guys didn't do!)... so that's a noun... it's a thing. It might be something that happens, like an eclipse, but it's not a verb, unless it's eclipsing, eclipsed. And secrets... I can't even think off the top of my head how that could be used in verb form. A secret is a thing that you know... or, more likely, don't know... or was that a joke via the 'secretes'? Hard to tell, not much of this is making sense at all :-/
Bout time you's lot should write proper!
Just GUI, as Linux. Command line, ctrl+c is break and will also abort many running processes (like a file copy, an oh-my-god-I-didn't-realise-there-were-this-many-files-in-here directory listing etc etc) or if the running process is ignoring it, it'll often be picked up between commands in a running batch file, where you'll get a prompt asking if you wish to terminate the script or continue. If you set 'break on' however, then you get more luck with ctrl+break than ctrl+c, as attention is usually paid to that in more places. Ctrl+Z however is the end-of-file marker, and so will stop text copies coming in from console (eg, 'copy con output.txt' will start copying text from console to output.txt, until you send a line with ctrl+z in... believe me, it can have its uses!)
Or working in vi running in a remote screen session from putty... each one having their own clipboard... very rarely I get mixed up, and if I do it's usually an undo key back... unless some reason I have a load of text or something in one clipboard (like accidentally selected something at some point) and I click the mouse button and just see loads of lines about 'command not found' as it tries to run every line that was in the clipboard, no stopping it, until it stops and begins chunking away at the harddrive and I spot the text 'rm -rf *' as the last line on the screen... if the machine's local I can hit the reset switch hopefully before the write behind cache catches up, but if not, I just sit back, relax, and thank the god I don't believe in that this has never actually happened to me (why would I have rm -rf in a text document? Maybe explaining to someone... I'm never going to write that again without at least a # in front of it just in case it does happen tho!)
Err... I can only assume 'cat' was compiled for a different processor, but there must be emulation or something... I dunno... it just said "illegal instruction", but carried on anyway???
"use OSS if some viable candidate exists. If it doesn't, all I'm arguing is that using something like Skype [...] is a sensible stop-gap"
:-p
I wouldn't hold your breath, with the average F/OSS developer, programming for fun, stuff they actually want to write, but looking like this, I can't see them putting too much effort into video conferencing software
I believe locks are already fairly fine grained in most of the system. Problems occur when you have multi level locks (eg, you want to change the value in a structure/object, so you get a lock on the structure, and then copy (duplicate) it, which could involve needing a lock on the table of structures before you can insert a new one. The lock on the table could be harder to get; you may have to wait for certain types of locks on the structures within to be released, but you can't release the lock on the object you're holding until you've created the copy, as you want the copy to be of the object in that state, and something else might change it before you get the lock on the table.
You can minimise the effects, but locking something in always implies the chance you are locking something else out. Where possible, lockless algorithyms can be switched to. One example is creating per-cpu versions of things. For example, a counter which goes up everytime something happens. A process on a CPU could get a lock, read the value, perhaps check it against an upper limit, increment and write it back, then release the lock, to make sure other processors don't change the value in between reading it and writing the new value back, which would lose the value the other process had updated it to. Instead, give each cpu its own counter, and when you want to know the value, you just add up all the values of all the counters.
Another method is using copy-on-write and reference counting. If you want to change a structure which something else has a read lock on, the structure is copied, and you change that. Pointers are updated so that anything that reads from it in the future reads from the new copy, only processes with the lock on the old copy read from that, and when there are no more left, the old copy is released and memory freed.
Both of these methods require that you can have more than one of what you'd want to lock, but obviously in some cases this just isn't true, because there is only one of the resource and it can't be copied.
You've gotta know that's complete nonsense, right? You wanna create an abstraction layer between every single bit of code in the kernel and the hardware it's meant to be running on to handle locking dynamically, making decisions at runtime, rather than expecting developers to be able to design a system and make those decisions for the code outside of its runtime and having the best way compiled to run directly on the hardware at full speed?
Managed languages have their place (and in fact all my work now is with managed languages)... but they also have to run on something. You think your java apps will just run without an operating system underneith it with these kinds of features to support the running of the virtual machine? What language do you think your java machine is written in? Do you think that's air you're breathing now? Hmm...
"This is innacurate, not correct and misleading at the same time"
But hey at least you prefixed what you had to say with this so we did know.
I'm pretty sure Windows has something like this, I recall seeing something in howto's about running different versions of IE along side each other, by creating an empty .local file with the same name as the executable (so in this case would've been iexplore.exe.local) and that affects the shared library loading path, so it will load dlls that are in the same directory before using others elsewhere in the system. Not quite the same, but could be used to achieve the same ends.
What part of the word "prefer" to you implies an assumed dichotomy? Surely, prefer is a conscious tendancy to lean towards one of the two oposites, which to me implies lack of a dichotomy not the presence of one, as with a dichotomy, there's no 'leaning towards', because you're either that or you're not.
"I trust mom, but I wouldn't tell her my passwords. I don't "hide" from her"
The "activeness" is the point though, surely. You might not announce to her your passwords, but you don't have to look over your shoulder every time you log in somewhere to make sure she's not trying to sneak a peak at your passwords right? I know I have no interest in finding out my mom's passwords, and she has no interest in finding out mine. She could even see my typing them in, but she'd have no interest in committing to memory each keypress in which order, so it's not something I have to worry about.
"I trust grandma, but I wouldn't necessarily tell her each time I took a taxi to the railway station"
But you don't have to look over your shoulder each time you jump in a cab to make sure she's not watching? Assuming she's able to get out still (okay this might not be the case, but for the sake of demonstrating a point...) it wouldn't be a problem if she did coincidentally see you taking a taxi one time? Unless she was following you around everywhere so saw you taking lots of taxis, but if that was the case, you'd in fact have other problems, and as we have laws against stalking already, I don't think we need new rules or anything to cover that scenario.
So as you I hope can see the point I'm making... privacy can exist in a mutual exclusion sense, rather than it being "something I must protect or even fight for with laws/regulations/cryptography". Things I do prefer to do in private (take obvious more extreme example, going to the toilet) I don't have to worry that people might be around, because I know they have no interest in watching that, and vacate of their own occord. In the same way, I don't need laws protecting my phone records from people I know, because I know they have no interest in looking at them, and people who don't know me... well they don't know me, they're not exactly going to be going through my phone records! Maybe collecting stats, but a computer is more likely to be used for that, and there's no personal link, they're not gonna be sniggering at the fact I called one person lots for a month then stopped, so I don't see how that would be a privacy issue.
Privacy works best through mutual voluntary exclusion, enforced exclusion I don't see as being too much of a requirement.
That's gotta be the most stupid selfish pathetic thing I've ever heard, and trust me, you've had a lot of competition. So basically you think that people phoning you is rude, so you won't allow that to happen, and only turn your phone on to inflict on other people what you wouldn't have them do to you themselves? *lol*
Also, this whole "people just let phones rudely interrupt them" thing that people come out with... what about if it wasn't a phone? What if it was someone knocking at the door? What would you do, ignore that? Or does someone knocking at the door have you well trained? What's so bad about being talked to?
Jeez.
Sure, I don't too much like listening to people shouting, whether they're on their phone or not, but generally I find the main cause of shouting to be stupidity or deafness... if we could ban stupidity that would be great.
"and do not have the compulsive need to be constantly talking to anyone about anything at all times"
No compulsive need here... but I have to say, I do like the fact that my friends know I'm there for them, wherever we are, if something goes wrong, they never have to deal with it alone, or if something good happens, they can always share it.
"To me, all these modern devices have made communication too easy and cheapened it"
I'm sorry that that has been your experience. Mine's been the oposite, and with it, I get the joy of watching a circle of good people strengthen as we all become more supportive of each other. It can be difficult for new people coming in, after being used to living more private/closed lives, but generally once the realisation comes that we're non judgemental (well, most of us), non manipulative people, the sharing inhibitions go, because there's no need for excessive privacy beyond basic decency. The proud don't feel the need to hide away, which more often than not, can give you more chance to have more to be proud about.
I don't understand the complains about people talking on their phones. What if the other person they were talking to was right there? Still talking about the same things, still as much in your ear space... just that your brain's not trying to listen in to figure out the other side of the conversation, which is why people really have a problem with mobile phones.
"How do you decide i have been missing against my will?"
By enforcing it mwuhahahaaa... I got the cage all ready to go... now just working out where you're likely to be using an IP locator
Lets say I'm just a crazy person who's attracted to bright lights, and the feeling of a sharp knife going though soft flesh... better turn off all your lights too then!
If you're really that desperate for privacy, turn your phone off! For anyone not suffering delusions of grandure (yeah so satalites can read your clothes labels from space, but who thinks they're really that interesting?) and realises that we buy phones because most of us are social creatures rather than paranoid conspiracy nuts, and actually like being able to make contact with other human beings.
"I usually find it best to err on the side of privacy rather than safety"
Well, I have friends, we share our lives with each other, and would much prefer we look out for each other than live our lives in fear and have to hide from each other. Maybe we're just unique like that.
"To me, "packet" implies that you can break it open and there's something smaller inside"
:-)
I guess, perhaps the connotations attached to the word for me that come from networking (where you send/received packets of data) have altered my internal meaning... sure you can look inside a network packet (kind of a requirement!) but it can't go smaller than a certain size, it needs to be big enough to carry it's information, such as the address it's going to, and state information (eg to make sure it won't travel forever if it gets caught in a loop). Any smaller and it can't exist (not as a packet on the network at least). A packet of light has some similarities... it can't exist any smaller, not because it's carrying physical objects, but it does also carry information (wavelength is information, but requires energy to exist)... and in fact quantum cryptography is based on being able to read quantum states of individual (well, entangled) packets of light. So no, we may not be able to physically split it up, but there is more in a packet of light than err.... meets the eye?
But I think we're probably entering more into the realms of language than physics now
"Do you actually have any bricks in your house?"
Um... yes? Is that a weird thing? I have walls, inside my house, that are made out of bricks. I guess a lot of places just use timber and plasterboard for internal walling, but no, my house, internal and exteral walls, built with bricks. Very solid thing. 100+ years old already, built to last.
"If not, then you're wrong"
But no... am right... sorry!