UNIX has this wonderful habit of trying to protect users from their own stupidity without recognising its own.
He seems to have confused the "imaginary" file system that is his OS X folders, with the actual file system underneath.
So, that quote should more accurately be:
"MacOS has this wonderful habit of trying to protect users from their own stupidity, and thus ensures that they remain stupid and completley dependant on it."
That's the way good port jobs are done. After all, Photoshop on Windows doesn't force you to use Alt-O (Alt being in the same position as Cmd is on a Mac keyboard) to open documents, does it? It doesn't make you use a Mac-esque open dialog, does it? Of course not. Windows users would be annoyed, and rightly so.
Photoshop on Windows... a GOOD port? That's a joke, right?
Photoshop on Windows has to have one of the most horrid UIs I've ever used. Half-baked MDI, nonstandard widgets, those AWFUL floating toolbars (shudder, try maximizing an image window sometime), and 50 billion tabs more or less arbitrarly grouped... And since all the good filters are plugins, they're usually grouped in the menu by author rather than by function.
IMO, Corel PhotoPaint has a much slicker and well-done UI. It's much more usable on Windows, probably becuase it was actually designed for that platform rather than being a shoddy port from Mac. Funny thing is that in addition to being able to seamlessly read Photoshop files and use Photoshop plugins, it's customizable enough that there's even a "feature" to emulate the Photoshop UI...
Similarly, GIMP flourishes best in a UNIX environment, where window managers are designed to manage groups of top-level windows owned by a single process.
The point I want to make is that absolute proof does only exist in mathematics. Any other discipline is strongly affected by personal beliefs, as the above example shows, where the scientists not even think about Plato's location of all this.
Yeah, but Plato's math was off by a factor of 10 due to bad translations from Egyptian. I thought everybody knew that.:D
I thought there was some tritium (deteurium++) involved somewhere in the process too... Of course, the last article I read about cold fusion theory was about 10 years ago.
You're half right. You won't really "break the pad", but you might still be able to decode the message (eventually); especially if the algorithm used to generate the pad is known. If you can come up with a list of every possible pad that might have been used, you can brute-force it.
If you use shortcuts to generate the pad, then this is true. However, an "ideal" OTP is completely random and is the same length as the original message. Actually, at that point, the distinction between the encrypted message and the pad goes away and you end up with two high-entropy blocks that happen to combine to form the original message. That can't be bruteforced with only one of the pads, if you try you'll end up decoding every possible message that could fit in the size of the original.
Of course, if your method for keeping one of the two pads secure is that good, why not just keep the original message in the secure location to begin with? Seems kind of pointless to me. Maybe if you're sending a message and use two relatively secure channels, there's less likelihood of them both being compromised...
Could also be useful as a method of stenography. Generate LOTS of random data to hide them in, and it's really easy to generate decoy pads that decode to completely harmless messages. I still can't see it being any more useful than regular crypto.
if your hard-drive is filled with what appears to be random garbage, but contains multiple encrypted slices (that cannot be detected without their respective magic keys)... there is an open source project that does this (i forget the name)
I don't know about a Nethack movie. Consider that most people get killed before the five minute mark. Will people pay admission for that?
I don't know, for most experienced players a game of 90-120 minutes isn't too uncommon. Would be the perfect length for a movie:)
It would be an interesting plot twist: The protagonist has been collecting lots of cool items, meeting interesting characters, killing monsters, when suddenly out of nowhere a cockatrice falls from the ceiling and petrifies him.
Or better yet, a death ray from a wand ricochets off the dungeon wall and we never find out who the killer is:)
For the sort of simple, single-user applications that MySQL is frequently used for, I don't understand why SQLite isn't used more often.
One word: Marketing. Nobody has heard of SQLite. I discovered it only recently -- somebody made a passing reference to it in relation to some embedded device development. I looked it up and it looks really slick. I'm planning on giving it a go the next time I have a need for a small database-driven app.
Nice. Is this something that came with 5.x, or is it in the late 4.x as well? I don't think I had sound mixing when I last had FreeBSD installed with sound (4.8, I think). I still use FreeBSD on my laptop, but I haven't bothered compiling in sound support.
It's been in 4.x for a while (I first remember using it around 4.6 -- according to cvsweb it's been in stable since 4.4), but it works better in 5.x.
In 4.x it's mostly manual. You do
sysctl hw.snd.pcm0.vchans=8
Then you get/dev/dsp0.0, dsp0.1, all the way to 0.7 that you can use as sound devices. Not quite seamless but it's still good for pointing XMMS at one of them, artsd/esd at another, and leaving/dev/dsp free for things hardcoded to use that.
I've been using 5.x almost exclusively for almost a year and it's much more useful there. All it takes is:
sysctl hw.snd.maxautovchans=8
(or whatever other number you feel like). 5.x has device cloning, so everything that opens/dev/dsp gets its own virtual channel, which is automatically allocated. Completely transparent and almost as easy as Windows 2k/XP. You do have to turn it on (man sysctl.conf to see how to make it permanent) but it's still a million times easier than fiddling with ALSA mixer plugins:-D
Assuming the **AA don't manage to mandate software installation on all student-owned campus PCs, this ACNS thing is probably going to be defeated by fileswapping LAN parties, or by SneakerNet.
But if you try the same on Linux or *BSD, the last program that tried to access/dev/dsp will hang politely while waiting for the first one to let go of the device[1].
No, FreeBSD has had vchans (low-latency kernel software mixing) for a long time now, so sound from multiple sources "just works" like it does in Windows -- even on el cheapo cards with only 1 hardware channel.
Linus or somebody seems to have some philospohical objection to mixing sound in the kernel, which is why Linux users are mostly stuck with crappy userland mixers or the "buy a better sound card" rhetoric.
When I am encoding a video on my Windows machine I can't do anything else on my computer, if I try to click something it takes about 30 seconds for the menu to popup.
That's a result of Windows's semi brain-dead priority system. Pop open task manager, find the encoding process (the one that's gobbling all your CPU), right click on it and set priority to 'Low' or 'Idle'.
I used to play Unreal Tournament while encoding videos all the time without a problem. It encodes slower at low priority but UT didn't suffer much of a performance hit. Made me miss my old dual-CPU system though.
Unless of course you're running that 9X/ME crap, which even most Windows users these days realize sucks.
If the bad guys have to come up with fakes of decently-designed ID cards, they have to make a lot more contacts and take a lot more actions. Every one of these is another opportunity for counterintelligence forces to detect their activity.
The problem is that most of the bad guys involved would have been able to get real ID cards through legitimate channels anyway. After all, if we knew for a fact that they were bad guys, we would have gone after them in the first place.
I mean let's face it, when it comes to suicide bombings there aren't many repeat offenders.
It's not illegal by the FCC because anything goes in 2.4 GHz so long as you don't go over the power limits... there's no bandwidth-footprint limit that keeps you from using everything between the lines.
So if my neighbors get one of these I just need something that will broadcast random noise at the maximum allowable power level over the whole 2.4Ghz band, with a directional antenna. Then we'll see how long it takes for them to give up and take it back to the store because it doesn't work.
An EULA is a purported agreement between you and the vendor, but hardware is sold, not licenced; and even if you lied to the vendor about what you were going to do with it after you bought it in order to be allowed to buy it, the transaction still stands.
If only it were that simple. I think the reasoning goes that the hardware is yours, but the firmware is protected by copyright and you can't use (license) it without agreeing to the EULA. Completely asinie, but that's what the lawyers seem to think.
Of course there may be a loophole there for devices like DVD players. The original argument that perverted copyright law from restricting distribution to restricing use was that you couldn't use software without making a copy (into RAM). However in most devices, the firmware executes directly from flash and never gets copied into RAM, so it's possible that things like EULAs could be completely invalid for them.
The problem is not on Gtk. Gtk makes use of a double buffer to avoid flickering. That's not bad, but it brings an X Window design decission to the surface.
When using X over a network, it always transfers ALL the screen to the clients, even the static parts.
I'm not so sure it's really X's fault. Qt-based apps over a remote X connection on a LAN can be difficult to tell that they're not running locally. They're very fast, even when using pixmap and alpha-blended themes (Xrender at work?). Over a dial-up or slow network they're not exactly fast, but are generally still quite usable.
Gtk-based programs (2.2 anyway, haven't tried 2.4 yet;) seem to be quite laggy even on a moderately fast LAN. Over a slow connection I usually give up and try to extract the data I need using the CLI if possible (though I've used gimp through Xvnc a few times because it was much quicker). Even using a theme like ThinIce it's still painful, especially redraws. Full-window drag seems to interact badly with gtk, I've seen it take 5-10 seconds just to redraw widgets LOCALLY after moving a window across.
I don't know if Qt does double-buffering or not, but I've never noticed any flickering. Not trying to knock gtk but it does not seem to be well-optimized for remote X connections.
Not to mention the millions of hidden features in that game. I wasted many hours just playing the 'secret' Scorched Earth clone embedded in it...
Re:No infringement required; allegations are enoug
on
EU Passes Nasty IP Law
·
· Score: 1
This gets back to the argument I made years ago. Capitalism doesn't work in a world of plenty... This will work no better than communism's attempt to impose a communist economic system on a domain where it wasn't applicable (a domain of scarcity)
Microsoft has pandered to broken applications for far too long. Maybe if they finally get over their "backwards compatibility at all costs" attitude, they'll get around to fixing some of the fundamental flaws in their OS.
I highly doubt that Linux authors would think twice about breaking buggy apps to force the issue.
Slashdot Mirror
"MacOS has this wonderful habit of trying to protect users from their own stupidity, and thus ensures that they remain stupid and completley dependant on it."
Photoshop on Windows has to have one of the most horrid UIs I've ever used. Half-baked MDI, nonstandard widgets, those AWFUL floating toolbars (shudder, try maximizing an image window sometime), and 50 billion tabs more or less arbitrarly grouped... And since all the good filters are plugins, they're usually grouped in the menu by author rather than by function.
IMO, Corel PhotoPaint has a much slicker and well-done UI. It's much more usable on Windows, probably becuase it was actually designed for that platform rather than being a shoddy port from Mac. Funny thing is that in addition to being able to seamlessly read Photoshop files and use Photoshop plugins, it's customizable enough that there's even a "feature" to emulate the Photoshop UI...
Similarly, GIMP flourishes best in a UNIX environment, where window managers are designed to manage groups of top-level windows owned by a single process.
Yeah, but Plato's math was off by a factor of 10 due to bad translations from Egyptian. I thought everybody knew that. :D
I thought there was some tritium (deteurium++) involved somewhere in the process too... Of course, the last article I read about cold fusion theory was about 10 years ago.
You're half right. You won't really "break the pad", but you might still be able to decode the message (eventually); especially if the algorithm used to generate the pad is known. If you can come up with a list of every possible pad that might have been used, you can brute-force it.
If you use shortcuts to generate the pad, then this is true. However, an "ideal" OTP is completely random and is the same length as the original message. Actually, at that point, the distinction between the encrypted message and the pad goes away and you end up with two high-entropy blocks that happen to combine to form the original message. That can't be bruteforced with only one of the pads, if you try you'll end up decoding every possible message that could fit in the size of the original.
Of course, if your method for keeping one of the two pads secure is that good, why not just keep the original message in the secure location to begin with? Seems kind of pointless to me. Maybe if you're sending a message and use two relatively secure channels, there's less likelihood of them both being compromised...
Could also be useful as a method of stenography. Generate LOTS of random data to hide them in, and it's really easy to generate decoy pads that decode to completely harmless messages. I still can't see it being any more useful than regular crypto.
Rubberhose
I don't know about a Nethack movie. Consider that most people get killed before the five minute mark. Will people pay admission for that?
:)
:)
I don't know, for most experienced players a game of 90-120 minutes isn't too uncommon. Would be the perfect length for a movie
It would be an interesting plot twist: The protagonist has been collecting lots of cool items, meeting interesting characters, killing monsters, when suddenly out of nowhere a cockatrice falls from the ceiling and petrifies him.
Or better yet, a death ray from a wand ricochets off the dungeon wall and we never find out who the killer is
I always thought that Nethack would make a good movie.
The biggest question is who to cast as '@'
For the sort of simple, single-user applications that MySQL is frequently used for, I don't understand why SQLite isn't used more often.
One word: Marketing. Nobody has heard of SQLite. I discovered it only recently -- somebody made a passing reference to it in relation to some embedded device development. I looked it up and it looks really slick. I'm planning on giving it a go the next time I have a need for a small database-driven app.
The article mentioned clustering PCs and having the entire database stored in memory.
Hope you have a REALLY good UPS. Might want to throw in a diesel generator too just for good measure...
Nice. Is this something that came with 5.x, or is it in the late 4.x as well? I don't think I had sound mixing when I last had FreeBSD installed with sound (4.8, I think). I still use FreeBSD on my laptop, but I haven't bothered compiling in sound support.
/dev/dsp0.0, dsp0.1, all the way to 0.7 that you can use as sound devices. Not quite seamless but it's still good for pointing XMMS at one of them, artsd/esd at another, and leaving /dev/dsp free for things hardcoded to use that.
/dev/dsp gets its own virtual channel, which is automatically allocated. Completely transparent and almost as easy as Windows 2k/XP. You do have to turn it on (man sysctl.conf to see how to make it permanent) but it's still a million times easier than fiddling with ALSA mixer plugins :-D
It's been in 4.x for a while (I first remember using it around 4.6 -- according to cvsweb it's been in stable since 4.4), but it works better in 5.x.
In 4.x it's mostly manual. You do
sysctl hw.snd.pcm0.vchans=8
Then you get
I've been using 5.x almost exclusively for almost a year and it's much more useful there. All it takes is:
sysctl hw.snd.maxautovchans=8
(or whatever other number you feel like). 5.x has device cloning, so everything that opens
Assuming the **AA don't manage to mandate software installation on all student-owned campus PCs, this ACNS thing is probably going to be defeated by fileswapping LAN parties, or by SneakerNet.
Or 802.11 Ghetto-net...
But if you try the same on Linux or *BSD, the last program that tried to access /dev/dsp will hang politely while waiting for the first one to let go of the device[1].
No, FreeBSD has had vchans (low-latency kernel software mixing) for a long time now, so sound from multiple sources "just works" like it does in Windows -- even on el cheapo cards with only 1 hardware channel.
Linus or somebody seems to have some philospohical objection to mixing sound in the kernel, which is why Linux users are mostly stuck with crappy userland mixers or the "buy a better sound card" rhetoric.
When I am encoding a video on my Windows machine I can't do anything else on my computer, if I try to click something it takes about 30 seconds for the menu to popup.
That's a result of Windows's semi brain-dead priority system. Pop open task manager, find the encoding process (the one that's gobbling all your CPU), right click on it and set priority to 'Low' or 'Idle'.
I used to play Unreal Tournament while encoding videos all the time without a problem. It encodes slower at low priority but UT didn't suffer much of a performance hit. Made me miss my old dual-CPU system though.
Unless of course you're running that 9X/ME crap, which even most Windows users these days realize sucks.
If the bad guys have to come up with fakes of decently-designed ID cards, they have to make a lot more contacts and take a lot more actions. Every one of these is another opportunity for counterintelligence forces to detect their activity.
The problem is that most of the bad guys involved would have been able to get real ID cards through legitimate channels anyway. After all, if we knew for a fact that they were bad guys, we would have gone after them in the first place.
I mean let's face it, when it comes to suicide bombings there aren't many repeat offenders.
It's not illegal by the FCC because anything goes in 2.4 GHz so long as you don't go over the power limits... there's no bandwidth-footprint limit that keeps you from using everything between the lines.
So if my neighbors get one of these I just need something that will broadcast random noise at the maximum allowable power level over the whole 2.4Ghz band, with a directional antenna. Then we'll see how long it takes for them to give up and take it back to the store because it doesn't work.
Is it just me or does "use tax" sound a whole lot like "tea tax"?
Hey, you can't have that! That snorkel's been just like a snorkel to me!
An EULA is a purported agreement between you and the vendor, but hardware is sold, not licenced; and even if you lied to the vendor about what you were going to do with it after you bought it in order to be allowed to buy it, the transaction still stands.
If only it were that simple. I think the reasoning goes that the hardware is yours, but the firmware is protected by copyright and you can't use (license) it without agreeing to the EULA. Completely asinie, but that's what the lawyers seem to think.
Of course there may be a loophole there for devices like DVD players. The original argument that perverted copyright law from restricting distribution to restricing use was that you couldn't use software without making a copy (into RAM). However in most devices, the firmware executes directly from flash and never gets copied into RAM, so it's possible that things like EULAs could be completely invalid for them.
I know that the parent post is a cliche, but I was about to ask this for real.
It's hard to tell from the shots that they had in the article, but that screen layout sure looks a lot like QTopia to me...
So, maybe.
Remember the rule of thumb: it's ok to see breasts on TV as long as they are mangled beyond recognition.
The problem is not on Gtk. Gtk makes use of a double buffer to avoid flickering. That's not bad, but it brings an X Window design decission to the surface.
;) seem to be quite laggy even on a moderately fast LAN. Over a slow connection I usually give up and try to extract the data I need using the CLI if possible (though I've used gimp through Xvnc a few times because it was much quicker). Even using a theme like ThinIce it's still painful, especially redraws. Full-window drag seems to interact badly with gtk, I've seen it take 5-10 seconds just to redraw widgets LOCALLY after moving a window across.
When using X over a network, it always transfers ALL the screen to the clients, even the static parts.
I'm not so sure it's really X's fault. Qt-based apps over a remote X connection on a LAN can be difficult to tell that they're not running locally. They're very fast, even when using pixmap and alpha-blended themes (Xrender at work?). Over a dial-up or slow network they're not exactly fast, but are generally still quite usable.
Gtk-based programs (2.2 anyway, haven't tried 2.4 yet
I don't know if Qt does double-buffering or not, but I've never noticed any flickering. Not trying to knock gtk but it does not seem to be well-optimized for remote X connections.
Not to mention the millions of hidden features in that game. I wasted many hours just playing the 'secret' Scorched Earth clone embedded in it...
This gets back to the argument I made years ago. Capitalism doesn't work in a world of plenty... This will work no better than communism's attempt to impose a communist economic system on a domain where it wasn't applicable (a domain of scarcity)
:)
Hey! Stop reading my mind
Microsoft has pandered to broken applications for far too long. Maybe if they finally get over their "backwards compatibility at all costs" attitude, they'll get around to fixing some of the fundamental flaws in their OS.
I highly doubt that Linux authors would think twice about breaking buggy apps to force the issue.