Oh, come on. The old dot matrix printers were waaay better than these modern inkjetty things. At least they were fairly robust..... especially the 9-pin ones. The Citizen 120D was an unstoppable machine! I recall knackering a few 24-pin printing heads in my time, though -- a worn ribbon is the kiss of death for a 24-pin head. Fortunately, when you've mangled two or three, you usually have enough bits to make one good one out of them.....
I miss dot matrix printers. And the manuals were invariably packed with programming examples showing what control code created what effect..... nowadays manufacturers treat that information like it was a secret or something. I'm always wary of such reserve, it suggests to me that they are hiding something. {"We can't tell you the code to select 2400 DPI, because that might give away important information to our competitors." Translation: there isn't one anyway. This is a 300 DPI printer and it's all done in the [closed-source] driver software. But if we let on then we'd be found out for the scam-merchants we are.}
If it isn't rated for industrial use, don't buy it.
If it's rated for industrial use, but it either doesn't have Linux drivers, or the Linux drivers aren't under an OSI-approved licence, don't buy it -- even if you don't want to use it with Linux today.
If the Linux drivers for your industrially-rated printer were written by a third party, it might be worth buying -- after all, it's a good sign that somebody actually thought it was worth bothering to support.
Bear in mind that you have already managed up to now without an inkjet printer. Investigate all alternatives fully before you buy one. Can you e-mail your files, or host them on some web space somewhere? If you want to show off some photos, try burning them onto a CD-R -- most DVD players will read CD-Rs of.jpg files and display them on the TV. Even if the person hasn't got a DVD player, you should be able to connect your DVD player to your VCR with a simple SCART to SCART cable. Just running out of screen space? Increase your number of virtual desktops.
In reality, when a researcher puts customers at risk by releasing exploit code for a vulnerability before the vendor has had a chance to fix it, it's ridiculous to expect the vendor to say, "Thank you for putting our customers at risk." I've never had a customer ask us for exploit code or exploit details, though they do want enough information to do a risk assessment.
Excuse me?! The security researcher did not put anybody at risk -- you put your customers at risk, when you wrote that code.
I firmly believe that if software vendors wish to keep their code secret from the people whom they expect to use it, then they should be obliged at the very least (1) to offer a performance warranty on the software, and (2) to place a copy of the source code in escrow with an approved agency. In the event of any dispute, then the Courts would have the power to order the source to be unsealed, and appoint experts to examine it in order to settle the dispute.
Open Source software {the only kind I will use; if you are not prepared to show me the code, then I am not prepared to allow it on my system} would not be affected by these requirements. The source code -is- a warranty. Likewise, by supplying the source code with every copy of the software, the escrow requirement is taken care of {since everyone who has the software already has a copy of the source code}. The analogy would be with goods supplied in kit form requiring to be built by the purchaser.
{And by the way, I use MySQL -- and a lot of glue scripting to take care of the "missing features" like stored procedures and subqueries. They only slow down the DB engine when they're not being used, which is most of the time. And even if the unthinkable happens and the server's power fails mid-transaction, someone can always deal with it later. It's not likely enough to be worth worrying about till it happens.}
Surely a bunch of local taxpayers {who now have to pay the firee's dole out of their own hard-earned} could come together and sue the company for firing you on the grounds that it puts them out of pocket?
What you need is a law allowing people unfettered access to any true fact, but at the same time forbidding the use of information not specifically volunteered in any decision-making process. So, for instance, unless someone specifically tells you that they are a 27-year-old bisexual asian female, you must assume in all dealings with them that they might be any age, sex, race and preference.
{Like the instructions "place each foot on the opposite thigh" or "fill in the grid so that the numbers one to nine occur exactly once in each row, column or 3x3 square", that probably is a lot harder than it sounds; since people would effectively be obliged to keep secrets from themselves and feign ignorance all over the place. But it'd work just fine.....}
Why the hell does a command line interface need to incorporate Object Oriented features? This sounds to me like adding features for features' sake.
The more sophisticated you make a system, the more failure modes you introduce -- and the harder it gets to test the edge cases, because there end up being too many edges. You want Obejct Oriented? I'll give you an Object Oriented example. Let's have a "length" type with properties which correspond to its conversion into different measuring units.
var height IsOfType length reset height let height = 1.75 print height.feet # prints 5 print height.feet.inches # prints 8.8975 print height.inches # prints 68.8975 reset height let height.inches = 72 print height.feet # prints 6 print height # prints 1.8288 forget height
It may well be pretty, but outside of any programme dealing with units conversion it's fairly unnecessary. And it contains many programming hazards which would thwart the careless implementor. {BTW, that was a fictitious example; but I'm willing to bet there is at least one programming language out there that actually implements something like it.}
All a command shell really has to do is be able to launch programmes, police the I/O traffic and keep hold of some state information. If it can do all that right, any other functionality you need can be provided by external programmes. That way, everything is kept as simple as it needs to be; you haven't got code cluttering up things that don't need it. If you do build functionality into the shell, there should be a bloody good reason -- usually that reason is that some external programme is getting launched more than its fair share. And in that case you already have the code you need to incorporate and it's been thoroughly tested.
Ah, but first you have to be able to write a cross-compiler. That still presumes a thorough working knowledge of the target architecture, since MOVs, ADDs, CMPs, JMPs and so forth are what the compiler must ultimately translate high-level instructions into. So you do still have to bootstrap the compiler, but you're effectively just doing it at the other end of the chain.
Just because a computer is based around a particular processor, does not necessarily mean it is going to be software- and hardware-compatible with other machines based around the same processor. For instance, the BBC model B, the Commodore PET and the Atari 400/800 were all based around the 6502 processor. The Sinclair Spectrum and the Amstrad PCW8256 were based around the Z80A processor. Both the Acorn Atom and the Dragon 32 used the 6847 video chip. And both the Oric-1 and the Amstrad CPC464 used the Yamaha AY-3-8910 sound/PIO chip. Anyone who was reading computer magazines in the 1980s will tell you that even when it was possible, it was a devil of a job to translate type-in listings meant for one machine so they could be run on another. It wasn't made any easier by the almost total lack of abstraction in many systems -- you would often have to use PEEK and POKE statements {and IN and OUT on Z80 machines with their separate memory and I/O buses} to manipulate memory and peripherals directly.
All Apple would need to do to prevent their OS from being run on generic hardware, would be to use a totally different addressing schema than that used by generic DOS/Windows/Linux PCs. Of course, it will still be possible to compile Linux and BSD for Apple hardware {the absolute minimum you need to port any OS written in a compiled language to a new architecture is an interpreter, written in assembler, and just capable enough to understand the compiler compiling itself}. But OS XI would not run on generic hardware because the memory map would be all wrong and the I/O devices would be in the wrong places. And as long as Apple did not release the source code, nobody would be able to recompile OS XI for generic hardware.
From the moment you swipe out, to the moment you swipe in the next morning, your employer has no power over you. The company rulebook does not apply to you. And your boss is just another civilian like you. Nothing that you do while you are on your own time has any relevance in any matter to do with the company; to all intents and purposes, "you at work" and "you at home" are two different people.
Even if you don't have a "proper" time clock, every place of employment needs some way of knowing who is in the building at any time in case of emergencies -- if there's a fire and someone is not at the designated assembly point, does that mean they're still in the building or they aren't at work today? So there must always a clear record of who is on the company's premises, and therefore bound by the company's rules -- or not. And if there is no such record, your employer could be in trouble with the Fire authority.
What really matters is that your labour is worth more to the company than the company's wages are worth to you. And watch out for the little creep who doesn't bleed when his brother gets cut.
Tracking moderator points and commenting on polls both are options that involve disk writes. So I suspect that a disk or partition probably has got full somewhere. I know that Linux -- and probably other unix-like systems -- will quite happily boot in that state; root even has a little bit of extra private space on each filesystem, which mortals cannot touch, for dealing with such emergencies.
By default, MySQL databases are under/var/lib/mysql/, mail messages are under/var/mail/ and logs go in/var/log/. If/var/ ever gets really full, then your automated e-mail notification system won't work -- and nor will it show up in the system logs.
I don't think you give people sufficient credit. Please try to remember that we would never have survived this long if we had been so stupid all along. Ignorance is curable!
In any competitive sport, if you want to improve your game, you have to play against an opponent who is capable of beating you. Otherwise, you will stagnate, and you may as well retire from competition.
Radio programmes that play only requests from listeners tend to be as boring as hell, because people only request their current favourite songs. On the other hand, DJs who have some input into their show's output -- e.g. the sadly-missed John Peel -- actually introduce listeners to new music they might never otherwise have thought of. And you would never have had a John Peel or Zane Lowe-type character on GWR. The advertisers {perhaps more strictly, the station bosses who suck up to the advertisers} would never sanction the playing of an "alternative" record. Real new music can nowadays only be found on the BBC, who are not beholden to advertisers. {Independent stations typically don't have a large enough audience, or a long enough broadcasting lifetime, to make a difference.}
The point is it's a human-nature thing. Day-to-day survival is so well taken care of in modern society that you cannot commit suicide by default. Contrast that with the situation faced by our cave-man ancestors..... and remember that our bodies and brains still work in more or less exactly the same way: we have all the same instincts, they just manifest themselves a little differently. {Ever felt a 'buzz' because you just bought something cheap? That's your brain releasing a shot of endorphins -- homebrew heroin for want of a better description -- as your chemical reward for helping the tribe survive by being a good hunter-gatherer. Cave-man would have felt the same when his spear struck home, or when he spied the ripe berries.} You can slag off "sheeple" all you like, but that is exactly what everyone is capable of becoming without a bit of a challenge from time to time. Cave-man probably could get comfortable for awhile, if the tribe had a good leader..... but sooner or later, and certainly before the rot had chance to set in, something would upset the stagnating situation.
And if the newspapers today are full of shallow material, and the wireless is full of the same six songs, it's not really the fault of the readership and the listeners. The papers and the radio stations have to take some responsibility for engaging their audience or not. They are the modern-day tribal leaders that we are biologically programmed to follow at some level, and have to take responsibility for the fact.
{Please note: I am not trying to promote authoritarianism over democracy! In a tribe of 20 people, the leadership would have been somewhat flexible according to the activity being performed later at the time. Human beings' specialisation is being able to do anything not quite as well as the best, but better than the worst -- while we cannot climb trees as well as a monkey, we can do so better than a wolf. Within the tribe there would be good hunters, good scouts, good weaponsmiths and so on. In a small group, the cream rises rapidly to the top and leaders are chosen on merit.}
You would think that in this day and age someone would have figured out a way to power a wireless mouse/trackball by harvesting the energy in moving the mouse/ball.
Nice idea, till you remember the first law of thermodynamics. There really isn't enough work done in moving the mouse to generate enough electricity to (1) run the optical sensor {ball mice are less power-hungry, but they still contain two LEDs for the shaft encoders; and given that your hand is over the mouse, use of daylight is hardly an option} and (2) transmit the information to the receiver {it needs a range of the order of a metre}. It can't help power consumption that there is generally a large, electrically conductive object -- your hand -- on top of the mouse. {Not only are you blocking daylight that could have illuminated the shaft encoders, now you're blocking the RF!} My corded optical mouse glides across the desk with almost no effort whatsoever. I would not like to think of how stiff it would be if it had to generate its own power.
A hybrid mouse, with an inertia generator to power the optics and an absorptive transponder {using an RF field from the "receiver", and passing information by drawing more or less power from the source à la RFID} might work; but I wouldn't like to be the engineer on that project, much less a technician. And it probably would require the Tx/Rx coil to be embedded in the mouse mat. Remember those Sparcstation mice which used a special pattern etched into the mouse mat?
All in all, I still think that when it comes to the twin tasks of getting power to the mouse and a signal back to the computer, a piece of wire is still the best way of doing it.
..... Or maybe the problem is the nature of the media companies. The instant you begin introducing advertisements with the news, you begin serving two masters: the general public, who have a right to know every true fact, and the advertisers.
Such a situation is a fertile breeding ground for conflict of interest; because the advertising introduces an obligation simply to sell more papers, and sod old-fashioned journalistic integrity.
{One almost gets the feeling that certain media outlets -- TV in particular, but many print outlets are just as bad -- see advertising as their "main" business and editorial content as just a kind of necessary evil, the sugar coating to persuade the customer to take the advert-laden pill}.
The only sensible solution seems to be to restrict the amount of advertising that newspapers can carry. And this has to be brought about through legislation, not through some sort of voluntary effort on the part of the newspaper industry; otherwise, the papers which continue to carry excessive advertising will continue to dominate, because they will be able to be sold at a lower price.
It's different altogether with science. If an "anonymous source" tells you that a spoon held so the back of the bowl is in the stream of water from a tap will be drawn into the stream and not ejected, or that there is always a bright spot in the centre of the shadow cast by a small sphere illuminated by a point source, then you can confirm or deny their claims by experiment.
And anyway, since you might get something -- ideally a measuring unit, though a university or a national museum would also be acceptable -- named after you, then it is generally in your best interest as a scientist to put your name to your discoveries.
I think duress passwords in general are a good idea. {Even 15 years ago there were burglar alarm systems which implemented such a feature: enter the last digit wrong by one, and the alarm appears to disarm correctly but the silent alert still triggers. Why don't banks do something similar with HITW cards -- deliberately mis-state your PIN to the man with the knife and he gets a message that your withdrawal limit was exceeded, his photograph taken and the account frozen? This would make chip and PIN payment a bit less insecure too.}
I'd really like with a server which gives crackers a busybox environment in a chroot if they try to login as root. And a HDD with two encrypted file systems wouldn't be too different. There might need to be two levels of duress code: one just to mount the innocuous file system {for demonstration}, and one to mount the innocuous file system and start trashing the real one {for real}. There exist crypto systems that are wildly intolerant of errors {muck up one character and the whole of the rest of the message is garbled}; using one such would help to maximise the damage from even a partial overwrite. {For times when the integrity of the data is less important than the need to stop other people from accessing it.}
I think what you really mean is that the OS division of Microsoft should not be making certain information available to the apps division of Microsoft that they do not make available to third parties -- like the "alternative" APIs and the documentation without the deliberate mistakes.
Certain MS apps ride the metal in places, simply because they can -- this is most of the reason why IE is such a useful attack vector. Everyone else has to put up with the "official" {slow and buggy} abstraction layers. MS aren't fixing the abstraction layers because they don't really care -- they don't use them anyway.
If Microsoft were to split into two -- and the job was done properly -- then the applications division would not be able to get this privileged information from the OS division unless every other software company was also allowed access to it. So either MS apps would have to start using the official APIs {which would be sorted sharpish} or third party apps would get to use the unofficial ones.
Meanwhile, the rest of the world is catching up and even overtaking. GNU/Linux on the desktop is entirely feasible unless you insist to have certain crappy appliances, to the point where absolute n00bs without preconceptions actually prefer it over Windows. Anything Microsoft do that breaks backward compatibility will give Open Source a foot in the door -- if the insecurity inherent in the closed-source model does not lead to its undoing first. {Who'd've thought that just by reverse-engineering a stack of patches, you can make certain discoveries about the vulnerability being patched -- and either slip in one last attack in the brief window just before people apply the patch, or work around the patch? Expect worse the closer decompilation comes to readiness.}
You actually can use LILO with Ubuntu -- it's really just using the new Debian installer, so this applies to Debian Sarge as well. When you get prompted whether to install Grub in the MBR or some partition, go back to the main menu, cursor down and select "install LILO". Proceed as normal from there.
It took me awhile to work out how to do this. I now do it on every Debian install. Grub may be new and trendy, but LILO just works.
Well, Ubuntu is basically a snapshot of Debian Unstable but with a slightly prettier front end. Debian Stable only generally gets updated for security-critical updates..... most of which are actually a lot less security-critical when you aren't on the Internet. Unstable gets updated whenever new packages are created {random intervals} and to fix any breakages caused by introducing new packages {soon afterwards if you're lucky}. Of course, broken stuff doesn't make it into Ubuntu -- their devs will fix it for Debian if necessary and possible, otherwise use an older version.
I'd reckon that, if you weren't on a broadband connection, you probably could get by without using every single update that Ubuntu offer you, just upgrading when the next CD set is released.
This is why I have ".inc" on my AddType application/x-httpd-php line. I also have ".asp" and ".cfm", but I don't really expect anybody to be fooled for a moment.
Of course, if you use.inc.php for an extension on your includes {unix doesn't care about double-barrelled extensions; in fact, it doesn't care about extensions at all, a dot is just a legal character in a filename} then you won't have to worry about this sort of thing. And the chances are that the worst thing you can expose is a MySQL username and password..... in all probability, "root" and "". Since MySQL usernames and passwords are not the same as unix usernames and passwords, this isn't nearly as vulnerable as it sounds..... the root MySQL account can only connect from localhost, and still gets all activities logged.
In some parts of the world, broadband internet has a somewhat lower priority than things like clean drinking water and efficient sewers. {Even though you have to admit the logistics are simpler}.
The bandwidth of a Ford Transit packed with CD-Rs should certainly not be underestimated!
I hear your argument, and it is a compelling one. But then, there are a lot of people out there who don't understand the politics. {Like anything left-of-centre, there are competing factions with a corresponding gamut of opinions. For all I know, the right-wingers may have divisions of their own too.} It may well be worth encouraging them to take at least a few baby-steps into the world of Open Source.
And let's face it, the politics {evenm though they tend to become remarkably easy to understand when Gates's Hired Goons start threatening to bust you for piracy} aren't everything. Once they've played a few games of Frozen Bubble, or burned a copy of a DVD+RW {from the TV recorder} onto a DVD+R {for keeps} using K3B, who's going to go back to Windows?
You physically own the disc or whatever other article on which the software is stored. Therefore, you are privy to the software; it is not secret from you {though you may be bound to keep it secret from others}.
Not true. Ignorance of the facts is a defence. If the person watching TV from outside the house can show they did not know that the set was unlicenced, they are not committing a crime.
Slashdot Mirror
Oh, come on. The old dot matrix printers were waaay better than these modern inkjetty things. At least they were fairly robust ..... especially the 9-pin ones. The Citizen 120D was an unstoppable machine! I recall knackering a few 24-pin printing heads in my time, though -- a worn ribbon is the kiss of death for a 24-pin head. Fortunately, when you've mangled two or three, you usually have enough bits to make one good one out of them .....
..... nowadays manufacturers treat that information like it was a secret or something. I'm always wary of such reserve, it suggests to me that they are hiding something. {"We can't tell you the code to select 2400 DPI, because that might give away important information to our competitors." Translation: there isn't one anyway. This is a 300 DPI printer and it's all done in the [closed-source] driver software. But if we let on then we'd be found out for the scam-merchants we are.}
I miss dot matrix printers. And the manuals were invariably packed with programming examples showing what control code created what effect
If it isn't rated for industrial use, don't buy it.
.jpg files and display them on the TV. Even if the person hasn't got a DVD player, you should be able to connect your DVD player to your VCR with a simple SCART to SCART cable. Just running out of screen space? Increase your number of virtual desktops.
If it's rated for industrial use, but it either doesn't have Linux drivers, or the Linux drivers aren't under an OSI-approved licence, don't buy it -- even if you don't want to use it with Linux today.
If the Linux drivers for your industrially-rated printer were written by a third party, it might be worth buying -- after all, it's a good sign that somebody actually thought it was worth bothering to support.
Bear in mind that you have already managed up to now without an inkjet printer. Investigate all alternatives fully before you buy one. Can you e-mail your files, or host them on some web space somewhere? If you want to show off some photos, try burning them onto a CD-R -- most DVD players will read CD-Rs of
I firmly believe that if software vendors wish to keep their code secret from the people whom they expect to use it, then they should be obliged at the very least (1) to offer a performance warranty on the software, and (2) to place a copy of the source code in escrow with an approved agency. In the event of any dispute, then the Courts would have the power to order the source to be unsealed, and appoint experts to examine it in order to settle the dispute.
Open Source software {the only kind I will use; if you are not prepared to show me the code, then I am not prepared to allow it on my system} would not be affected by these requirements. The source code -is- a warranty. Likewise, by supplying the source code with every copy of the software, the escrow requirement is taken care of {since everyone who has the software already has a copy of the source code}. The analogy would be with goods supplied in kit form requiring to be built by the purchaser.
{And by the way, I use MySQL -- and a lot of glue scripting to take care of the "missing features" like stored procedures and subqueries. They only slow down the DB engine when they're not being used, which is most of the time. And even if the unthinkable happens and the server's power fails mid-transaction, someone can always deal with it later. It's not likely enough to be worth worrying about till it happens.}
Surely a bunch of local taxpayers {who now have to pay the firee's dole out of their own hard-earned} could come together and sue the company for firing you on the grounds that it puts them out of pocket?
What you need is a law allowing people unfettered access to any true fact, but at the same time forbidding the use of information not specifically volunteered in any decision-making process. So, for instance, unless someone specifically tells you that they are a 27-year-old bisexual asian female, you must assume in all dealings with them that they might be any age, sex, race and preference.
.....}
{Like the instructions "place each foot on the opposite thigh" or "fill in the grid so that the numbers one to nine occur exactly once in each row, column or 3x3 square", that probably is a lot harder than it sounds; since people would effectively be obliged to keep secrets from themselves and feign ignorance all over the place. But it'd work just fine
The more sophisticated you make a system, the more failure modes you introduce -- and the harder it gets to test the edge cases, because there end up being too many edges. You want Obejct Oriented? I'll give you an Object Oriented example. Let's have a "length" type with properties which correspond to its conversion into different measuring units.It may well be pretty, but outside of any programme dealing with units conversion it's fairly unnecessary. And it contains many programming hazards which would thwart the careless implementor. {BTW, that was a fictitious example; but I'm willing to bet there is at least one programming language out there that actually implements something like it.}
All a command shell really has to do is be able to launch programmes, police the I/O traffic and keep hold of some state information. If it can do all that right, any other functionality you need can be provided by external programmes. That way, everything is kept as simple as it needs to be; you haven't got code cluttering up things that don't need it. If you do build functionality into the shell, there should be a bloody good reason -- usually that reason is that some external programme is getting launched more than its fair share. And in that case you already have the code you need to incorporate and it's been thoroughly tested.
Ah, but first you have to be able to write a cross-compiler. That still presumes a thorough working knowledge of the target architecture, since MOVs, ADDs, CMPs, JMPs and so forth are what the compiler must ultimately translate high-level instructions into. So you do still have to bootstrap the compiler, but you're effectively just doing it at the other end of the chain.
Just because a computer is based around a particular processor, does not necessarily mean it is going to be software- and hardware-compatible with other machines based around the same processor. For instance, the BBC model B, the Commodore PET and the Atari 400/800 were all based around the 6502 processor. The Sinclair Spectrum and the Amstrad PCW8256 were based around the Z80A processor. Both the Acorn Atom and the Dragon 32 used the 6847 video chip. And both the Oric-1 and the Amstrad CPC464 used the Yamaha AY-3-8910 sound/PIO chip. Anyone who was reading computer magazines in the 1980s will tell you that even when it was possible, it was a devil of a job to translate type-in listings meant for one machine so they could be run on another. It wasn't made any easier by the almost total lack of abstraction in many systems -- you would often have to use PEEK and POKE statements {and IN and OUT on Z80 machines with their separate memory and I/O buses} to manipulate memory and peripherals directly.
All Apple would need to do to prevent their OS from being run on generic hardware, would be to use a totally different addressing schema than that used by generic DOS/Windows/Linux PCs. Of course, it will still be possible to compile Linux and BSD for Apple hardware {the absolute minimum you need to port any OS written in a compiled language to a new architecture is an interpreter, written in assembler, and just capable enough to understand the compiler compiling itself}. But OS XI would not run on generic hardware because the memory map would be all wrong and the I/O devices would be in the wrong places. And as long as Apple did not release the source code, nobody would be able to recompile OS XI for generic hardware.
No they can't.
From the moment you swipe out, to the moment you swipe in the next morning, your employer has no power over you. The company rulebook does not apply to you. And your boss is just another civilian like you. Nothing that you do while you are on your own time has any relevance in any matter to do with the company; to all intents and purposes, "you at work" and "you at home" are two different people.
Even if you don't have a "proper" time clock, every place of employment needs some way of knowing who is in the building at any time in case of emergencies -- if there's a fire and someone is not at the designated assembly point, does that mean they're still in the building or they aren't at work today? So there must always a clear record of who is on the company's premises, and therefore bound by the company's rules -- or not. And if there is no such record, your employer could be in trouble with the Fire authority.
What really matters is that your labour is worth more to the company than the company's wages are worth to you. And watch out for the little creep who doesn't bleed when his brother gets cut.
Tracking moderator points and commenting on polls both are options that involve disk writes. So I suspect that a disk or partition probably has got full somewhere. I know that Linux -- and probably other unix-like systems -- will quite happily boot in that state; root even has a little bit of extra private space on each filesystem, which mortals cannot touch, for dealing with such emergencies.
/var/lib/mysql/, mail messages are under /var/mail/ and logs go in /var/log/. If /var/ ever gets really full, then your automated e-mail notification system won't work -- and nor will it show up in the system logs.
By default, MySQL databases are under
I don't think you give people sufficient credit. Please try to remember that we would never have survived this long if we had been so stupid all along. Ignorance is curable!
..... and remember that our bodies and brains still work in more or less exactly the same way: we have all the same instincts, they just manifest themselves a little differently. {Ever felt a 'buzz' because you just bought something cheap? That's your brain releasing a shot of endorphins -- homebrew heroin for want of a better description -- as your chemical reward for helping the tribe survive by being a good hunter-gatherer. Cave-man would have felt the same when his spear struck home, or when he spied the ripe berries.} You can slag off "sheeple" all you like, but that is exactly what everyone is capable of becoming without a bit of a challenge from time to time. Cave-man probably could get comfortable for awhile, if the tribe had a good leader ..... but sooner or later, and certainly before the rot had chance to set in, something would upset the stagnating situation.
In any competitive sport, if you want to improve your game, you have to play against an opponent who is capable of beating you. Otherwise, you will stagnate, and you may as well retire from competition.
Radio programmes that play only requests from listeners tend to be as boring as hell, because people only request their current favourite songs. On the other hand, DJs who have some input into their show's output -- e.g. the sadly-missed John Peel -- actually introduce listeners to new music they might never otherwise have thought of. And you would never have had a John Peel or Zane Lowe-type character on GWR. The advertisers {perhaps more strictly, the station bosses who suck up to the advertisers} would never sanction the playing of an "alternative" record. Real new music can nowadays only be found on the BBC, who are not beholden to advertisers. {Independent stations typically don't have a large enough audience, or a long enough broadcasting lifetime, to make a difference.}
The point is it's a human-nature thing. Day-to-day survival is so well taken care of in modern society that you cannot commit suicide by default. Contrast that with the situation faced by our cave-man ancestors
And if the newspapers today are full of shallow material, and the wireless is full of the same six songs, it's not really the fault of the readership and the listeners. The papers and the radio stations have to take some responsibility for engaging their audience or not. They are the modern-day tribal leaders that we are biologically programmed to follow at some level, and have to take responsibility for the fact.
{Please note: I am not trying to promote authoritarianism over democracy! In a tribe of 20 people, the leadership would have been somewhat flexible according to the activity being performed later at the time. Human beings' specialisation is being able to do anything not quite as well as the best, but better than the worst -- while we cannot climb trees as well as a monkey, we can do so better than a wolf. Within the tribe there would be good hunters, good scouts, good weaponsmiths and so on. In a small group, the cream rises rapidly to the top and leaders are chosen on merit.}
A hybrid mouse, with an inertia generator to power the optics and an absorptive transponder {using an RF field from the "receiver", and passing information by drawing more or less power from the source à la RFID} might work; but I wouldn't like to be the engineer on that project, much less a technician. And it probably would require the Tx/Rx coil to be embedded in the mouse mat. Remember those Sparcstation mice which used a special pattern etched into the mouse mat?
All in all, I still think that when it comes to the twin tasks of getting power to the mouse and a signal back to the computer, a piece of wire is still the best way of doing it.
..... Or maybe the problem is the nature of the media companies. The instant you begin introducing advertisements with the news, you begin serving two masters: the general public, who have a right to know every true fact, and the advertisers.
Such a situation is a fertile breeding ground for conflict of interest; because the advertising introduces an obligation simply to sell more papers, and sod old-fashioned journalistic integrity.
{One almost gets the feeling that certain media outlets -- TV in particular, but many print outlets are just as bad -- see advertising as their "main" business and editorial content as just a kind of necessary evil, the sugar coating to persuade the customer to take the advert-laden pill}.
The only sensible solution seems to be to restrict the amount of advertising that newspapers can carry. And this has to be brought about through legislation, not through some sort of voluntary effort on the part of the newspaper industry; otherwise, the papers which continue to carry excessive advertising will continue to dominate, because they will be able to be sold at a lower price.
It's different altogether with science. If an "anonymous source" tells you that a spoon held so the back of the bowl is in the stream of water from a tap will be drawn into the stream and not ejected, or that there is always a bright spot in the centre of the shadow cast by a small sphere illuminated by a point source, then you can confirm or deny their claims by experiment.
And anyway, since you might get something -- ideally a measuring unit, though a university or a national museum would also be acceptable -- named after you, then it is generally in your best interest as a scientist to put your name to your discoveries.
I think duress passwords in general are a good idea. {Even 15 years ago there were burglar alarm systems which implemented such a feature: enter the last digit wrong by one, and the alarm appears to disarm correctly but the silent alert still triggers. Why don't banks do something similar with HITW cards -- deliberately mis-state your PIN to the man with the knife and he gets a message that your withdrawal limit was exceeded, his photograph taken and the account frozen? This would make chip and PIN payment a bit less insecure too.}
I'd really like with a server which gives crackers a busybox environment in a chroot if they try to login as root. And a HDD with two encrypted file systems wouldn't be too different. There might need to be two levels of duress code: one just to mount the innocuous file system {for demonstration}, and one to mount the innocuous file system and start trashing the real one {for real}. There exist crypto systems that are wildly intolerant of errors {muck up one character and the whole of the rest of the message is garbled}; using one such would help to maximise the damage from even a partial overwrite. {For times when the integrity of the data is less important than the need to stop other people from accessing it.}
I think what you really mean is that the OS division of Microsoft should not be making certain information available to the apps division of Microsoft that they do not make available to third parties -- like the "alternative" APIs and the documentation without the deliberate mistakes.
Certain MS apps ride the metal in places, simply because they can -- this is most of the reason why IE is such a useful attack vector. Everyone else has to put up with the "official" {slow and buggy} abstraction layers. MS aren't fixing the abstraction layers because they don't really care -- they don't use them anyway.
If Microsoft were to split into two -- and the job was done properly -- then the applications division would not be able to get this privileged information from the OS division unless every other software company was also allowed access to it. So either MS apps would have to start using the official APIs {which would be sorted sharpish} or third party apps would get to use the unofficial ones.
Meanwhile, the rest of the world is catching up and even overtaking. GNU/Linux on the desktop is entirely feasible unless you insist to have certain crappy appliances, to the point where absolute n00bs without preconceptions actually prefer it over Windows. Anything Microsoft do that breaks backward compatibility will give Open Source a foot in the door -- if the insecurity inherent in the closed-source model does not lead to its undoing first. {Who'd've thought that just by reverse-engineering a stack of patches, you can make certain discoveries about the vulnerability being patched -- and either slip in one last attack in the brief window just before people apply the patch, or work around the patch? Expect worse the closer decompilation comes to readiness.}
And pray why, if 20th century software does the job just fine, might he want to do that?
You actually can use LILO with Ubuntu -- it's really just using the new Debian installer, so this applies to Debian Sarge as well. When you get prompted whether to install Grub in the MBR or some partition, go back to the main menu, cursor down and select "install LILO". Proceed as normal from there.
It took me awhile to work out how to do this. I now do it on every Debian install. Grub may be new and trendy, but LILO just works.
Well, Ubuntu is basically a snapshot of Debian Unstable but with a slightly prettier front end. Debian Stable only generally gets updated for security-critical updates ..... most of which are actually a lot less security-critical when you aren't on the Internet. Unstable gets updated whenever new packages are created {random intervals} and to fix any breakages caused by introducing new packages {soon afterwards if you're lucky}. Of course, broken stuff doesn't make it into Ubuntu -- their devs will fix it for Debian if necessary and possible, otherwise use an older version.
I'd reckon that, if you weren't on a broadband connection, you probably could get by without using every single update that Ubuntu offer you, just upgrading when the next CD set is released.
This is why I have ".inc" on my AddType application/x-httpd-php line. I also have ".asp" and ".cfm", but I don't really expect anybody to be fooled for a moment.
.inc.php for an extension on your includes {unix doesn't care about double-barrelled extensions; in fact, it doesn't care about extensions at all, a dot is just a legal character in a filename} then you won't have to worry about this sort of thing. And the chances are that the worst thing you can expose is a MySQL username and password ..... in all probability, "root" and "". Since MySQL usernames and passwords are not the same as unix usernames and passwords, this isn't nearly as vulnerable as it sounds ..... the root MySQL account can only connect from localhost, and still gets all activities logged.
Of course, if you use
In some parts of the world, broadband internet has a somewhat lower priority than things like clean drinking water and efficient sewers. {Even though you have to admit the logistics are simpler}.
The bandwidth of a Ford Transit packed with CD-Rs should certainly not be underestimated!
I hear your argument, and it is a compelling one. But then, there are a lot of people out there who don't understand the politics. {Like anything left-of-centre, there are competing factions with a corresponding gamut of opinions. For all I know, the right-wingers may have divisions of their own too.} It may well be worth encouraging them to take at least a few baby-steps into the world of Open Source.
And let's face it, the politics {evenm though they tend to become remarkably easy to understand when Gates's Hired Goons start threatening to bust you for piracy} aren't everything. Once they've played a few games of Frozen Bubble, or burned a copy of a DVD+RW {from the TV recorder} onto a DVD+R {for keeps} using K3B, who's going to go back to Windows?
You physically own the disc or whatever other article on which the software is stored. Therefore, you are privy to the software; it is not secret from you {though you may be bound to keep it secret from others}.
You'll have a job enforcing it, though.
Not true. Ignorance of the facts is a defence. If the person watching TV from outside the house can show they did not know that the set was unlicenced, they are not committing a crime.