So reducing someone's market for their IP constitutes theft in any reasonable ethical system?
Doesn't that mean that offering a competing product also constitutes theft in such a system? Or any of the other legal things one can do which might result in a lessening of the market for a particular product?
It's closer to English than, say, "ethernet" is, and just about as widely known by this particular audience.
And it does actually serve a distinctive purpose in the language. It indicates a much more fundamental, visceral understanding of something than the austere "knowing" of facts.
I'm not a Heinleinite, and am a prescriptivist, but I find this to be entirely valid usage.
This has been your completely off-topic pedantry for the day.
You're right. The only sense in which root is "disabled" in osx is that it has no valid password unless someone explicitly sets one. So supplying a whole new set of accounts does obviate that.
Fortunately, the other point is still somewhat valid: sshd (and afpd, and all other services) are turned off by default. Anyone who enables any of them could get bitten by this, so it's still a problem, but the vast majority of users will leave them off and be invulnerable.
Apple has essentially made the design choice to default to a system which trusts the local dhcp server. Which is problematic much of the time, but convenient if you'd like to just unbox a new shipment of macs for your lab and plug them in, without needing any further client-side config.
This means that the dhcp server can provide authoritative information about anything ldap handles, including user accounts. So Mallory can use a rogue dhcp server to give herself a root account on your system.
But unless I'm mistaken, the default configuration still doesn't allow her to do anything with it. sshd and afpd are turned off by default, so even having a root account doesn't get you anything unless you physically sit down at the box and log in locally.
I think I'd prefer that the system defaulted to not trusting other hosts for anything beyond network numbers, but I don't think that issue will lead immediately to a rash of rooted osx machines.
I'm a big fan of technical solutions over administrative solutions. But in this case, technology has failed to solve the problem, and gives every evidence that it will continue to fail to solve the problem. Many of the world's greatest minds and largest budgets have been bent on technical solutions to spam, with very limited success.
Your assertions that spam is effective and easy don't distinguish it from other crimes. Mugging people is a fast, easy way to make money. We could either encourage a technical solution to the problem (perhaps based around guns which use mind-reading devices to determine the purity of the intent of their wielder, and only fire in approved circumstances), or we could go with the administrative solution of saying it's illegal to mug people.
Would the technical solution be better? Hell yes! Once it exists. Is the administrative solution better than nothing for the years or decades until the technical solution happens? Yep.
Unfortunately, all the improvements to SMTP that I've heard proposed would still be rendered moot by the insecurity of Windows.
Strong authentication to a mail server that knows you personally? Unforgeable headers? Hash cash? Great ideas, but not ones that will have any effect on millions of compromised Windows systems each sending a small number of messages properly through their own mail servers.
Do you have some improvements in mind which would obviate the zombie-army problem? I'd love to hear them.
With any current X11 server of which I know, no. Apple made the mixed decision to implement a windowing system which is largely vector based, and which uses opengl textures for compositing of ui elements. Which is completely different from the behaviour of extant X11 servers.
I say "mixed" because Apple's approach has some fairly significant overhead for even the simplest of operations, but very complex operations (like Expose) are pretty much free beyond that.
You also see with with the "genie effect" by which windows minimize into the Dock, the "cube" effect of fast user switching, and true window transparency. They're basically showing off all the neat stuff that's easy to do, to counterbalance the fact that the whole windowing system is slow on older systems.
Xfree86 or a window manager could hack in some poor imitation of the effect, just as some terminals lamely imitate transparency. But doing it "for reals" would involve some huge bottom-up changes.
Well, that's a _much_ more marginal case than ceasing security updates for an OS that was current last week.
AFP over TCP was introduced with that, macos 8? Meaning 1996ish? Given the immense degree to which IP is better suited to ethernet than DDP is, expecting people to transition within a span of seven years isn't exactly demanding.
Dual display and video mirroring: Simultaneously supports up to 1024 by 768 pixels on the built-in display and up to 2048 by 1536 pixels on an external display, both at millions of colors
Even more people use cars, telephones, and clothes every day. Do you also feel that we need standardized education on crankshaft design, that all high school graduates should know the current provided by dialtone from an ESS7, and that all citizens should be familiar with cotton cultivation techniques?
The fact that these devices are ubiquitous is a cause and result of the fact that one does _not_ need to know arcana like "what a megabyte is" to use them.
I've seen people make this argument for at least ten years now. (And I suspect that it's been made far longer than that, and I'm just displaying my relative inexperience.)
It always turns out to be pretty shortsighted. Yes, an average new computer is overkill for most users the day it's purchased. Software is targetted at the average computer in use, new computers tend to be faster than the average, so new computers are always overkill. This is so definitional as to border on tautology.
But software continues to evolve to take advantage of new hardware resources.
It's easy to just dismiss this as laziness an inpcompetence on the part of software developers; to say that they're not really adding anything new, they can just afford to be sloppy now.
Obviously that's an oversimplification: some software becomes more demanding because it simply does things which were not practical with more limited resources. But even in the worst case, where new software is less efficient, I'd still call this a feature rather than a bug. If developers can spend less time on optimization, they can spend more time on more important things. Those efforts can go into clarity, security, consistency, or portability.
(Clearly some developers, notably those working for a certain little startup in Redmond, will not use these freed-up resources wisely, and will instead just take the opportunity to write a larger volume of bad code. But those are the same developers who probably would've buggered up the optimization process anyway, so this is really no worse.)
If we'd just stopped when we had enough hardware to speedily run common user applications, it would be too minimal to handle taxing things like tcp/ip, which you might have found to be handy.
And allow me the honor of being the next in a long line of people you'll meet who dislike tabs. They are a poor solution to the problem, and unquestionably implemented at the wrong layer.
If what you want is a single-gesture way of switching linearly through the windows in the current application, you want command-~, something all Cocoa applications get for free.
And tabs do have a cost to those who don't use them: the opportunity cost of the development time spent on them. Tabs don't implement themselves, and I'd almost certainly rather have whatever other feature those developers were creating instead.
Do you respond to news items about new operating systems with comments that operating systems have existed for decades? Or to articles on processors by pointing out that the abacus was around millennia ago?
The task is not new, and has been tackled before. This article discusses and compares the most current tools for the task. This is earthshattering to about the same degree that most technology news is.
640x480x32 appears to be 7372800 bits a second. This is considerably more data than one can reliably pass through usb1.1, despite its theoretical bandwidth of 11Mbps.
Those usb camera that do 640x480 are managing this by doing primitive hardware compression in the camera before stuffing it down the usb line. But of course they're cheap, so they're not doing good enough compression that you want to just send it over the net raw, you need to compress it again in software. At which point you're stacking lossy codecs, and you get crappier video.
Firewire camera have the bandwidth available (not only through sheer speed, but through isosynchronous transfers, which usb2 still lacks) to just send untouched 640x480x32, and your software has a clean shot at it.
Well, you could use a g4 tower as a midpoint between an imac and g5, giving you both expandability and some more speed. (Yes, they're still available, and range between $1200 and $1500).
My claims about the g5's speed were based upon the information revealed yesterday, which implies that they're actually significantly faster even than those dual 3ghz amd cpus. Of course, we'll need to wait to see them in person and verify how true that turns out to be, but the current data seem very promising.
And yes, if assembling a machine is something that you find to be fun, then obviously that's not a cost. And without that cost, I think it's safe to say that a personally-built machine will always be less expensive than any prebuilt one.
Last year, my attempt to follow the several pages of instructions for installing Mathematica on linux failed. I eventually spent several unpleasant hours hacking through 50K of interdependent awk scripts, reworking some of them and manually handling some of the things they were failing to do. I got it working eventually, but it wasn't fun. And I certainly can't remember what all I did, so if I had to install it again tomorrow, I'd have to repeat the whole painful process.
Imagine how happy I was when I installed it the following day on macosx, and found that instructions consisted of "Drag the Mathematica icon to the Applications folder."
Unless that motherboard has a lot more than I'd guess, you've left out gigabit ethernet, firewire 400 and 800, optical audio in and out, and amplified analog audio out.
Perhaps more importantly, that system would be about half the speed of the mac to which you're comparing it.
Don't need GigE, optical audio, and all that speed? Then you should be comparing to an imac, which is still cheaper than what you've listed.
Certainly more importantly, you haven't accounted for your time spent selecting, purchasing, and assembling the parts. What's your consulting rate? Several hours work at a few hundred an hour dwarfs the cost of the hardware entirely.
One word for four is exactly the same phonetically as the the word for death: shi.
But the glyphs for them are completely different. This happens a lot in Japanese, which has a very limited sound palette and an incredibly rich character palette, so most people don't think much of it. (Because of this, the idea of a pun is pretty much meaningless in Japanese; every word sounds exactly like a dozen others anyway, but they're all conceptually distinct. (Though oddly enough, the exact same set of circumstances has lead to the Chinese being very punny.))
This does mean that four has a very mild reputation as an unlucky number, about comparable to thirteen in American culture. But I don't think anybody bothers to rename products to avoid it.
I suspect that Canon's reasoning was similar to Netscape's: "bigger numbers better."
Yeah, that function of Terminal is odd, and hard to give a description or name, but surprisingly handy. It allows you to simultaneously see and interact with two arbitrary segments of scrollback within a single window.
This facilitates doing a variant on a long series of steps you did a while earlier, comparing previous output of an operation to current output, or just reading deep scrollback without missing new data. It's not a life-changing feature, but it is regularly useful.
iTerm has given me the one thing that I've found seriously lacking in Terminal.app: configuration of what colors are used to display ANSI "colors". No more screaming yellow or illegible dark blue for me, thanks.
Unfortunately, iTerm does have a few limitations and bugs:
- while the xterm-experienced will like PgUp/PgDown going straight through, and using shift for local scrolling, I'd really like to see this togglable.
- no Home/End functionality, with our without shift.
- no local Find.
- it "helpfully" doesn't include whitespace when copying out of its windows. Actually, I did want that linefeed, thanks.
- periodically decides it wants to just sit and suck all my cpu until I kill it.
- font settings don't stick between launches.
I've also found that Terminal.app's split-window function is surprisingly useful. And unique, in my experience.
Slashdot Mirror
So reducing someone's market for their IP constitutes theft in any reasonable ethical system?
Doesn't that mean that offering a competing product also constitutes theft in such a system? Or any of the other legal things one can do which might result in a lessening of the market for a particular product?
It's closer to English than, say, "ethernet" is, and just about as widely known by this particular audience.
And it does actually serve a distinctive purpose in the language. It indicates a much more fundamental, visceral understanding of something than the austere "knowing" of facts.
I'm not a Heinleinite, and am a prescriptivist, but I find this to be entirely valid usage.
This has been your completely off-topic pedantry for the day.
You're right. The only sense in which root is "disabled" in osx is that it has no valid password unless someone explicitly sets one. So supplying a whole new set of accounts does obviate that.
Fortunately, the other point is still somewhat valid: sshd (and afpd, and all other services) are turned off by default. Anyone who enables any of them could get bitten by this, so it's still a problem, but the vast majority of users will leave them off and be invulnerable.
Apple has essentially made the design choice to default to a system which trusts the local dhcp server. Which is problematic much of the time, but convenient if you'd like to just unbox a new shipment of macs for your lab and plug them in, without needing any further client-side config.
This means that the dhcp server can provide authoritative information about anything ldap handles, including user accounts. So Mallory can use a rogue dhcp server to give herself a root account on your system.
But unless I'm mistaken, the default configuration still doesn't allow her to do anything with it. sshd and afpd are turned off by default, so even having a root account doesn't get you anything unless you physically sit down at the box and log in locally.
I think I'd prefer that the system defaulted to not trusting other hosts for anything beyond network numbers, but I don't think that issue will lead immediately to a rash of rooted osx machines.
I'm a big fan of technical solutions over administrative solutions. But in this case, technology has failed to solve the problem, and gives every evidence that it will continue to fail to solve the problem. Many of the world's greatest minds and largest budgets have been bent on technical solutions to spam, with very limited success.
Your assertions that spam is effective and easy don't distinguish it from other crimes. Mugging people is a fast, easy way to make money. We could either encourage a technical solution to the problem (perhaps based around guns which use mind-reading devices to determine the purity of the intent of their wielder, and only fire in approved circumstances), or we could go with the administrative solution of saying it's illegal to mug people.
Would the technical solution be better? Hell yes! Once it exists. Is the administrative solution better than nothing for the years or decades until the technical solution happens? Yep.
Unfortunately, all the improvements to SMTP that I've heard proposed would still be rendered moot by the insecurity of Windows.
Strong authentication to a mail server that knows you personally? Unforgeable headers? Hash cash? Great ideas, but not ones that will have any effect on millions of compromised Windows systems each sending a small number of messages properly through their own mail servers.
Do you have some improvements in mind which would obviate the zombie-army problem? I'd love to hear them.
With any current X11 server of which I know, no. Apple made the mixed decision to implement a windowing system which is largely vector based, and which uses opengl textures for compositing of ui elements. Which is completely different from the behaviour of extant X11 servers.
I say "mixed" because Apple's approach has some fairly significant overhead for even the simplest of operations, but very complex operations (like Expose) are pretty much free beyond that.
You also see with with the "genie effect" by which windows minimize into the Dock, the "cube" effect of fast user switching, and true window transparency. They're basically showing off all the neat stuff that's easy to do, to counterbalance the fact that the whole windowing system is slow on older systems.
Xfree86 or a window manager could hack in some poor imitation of the effect, just as some terminals lamely imitate transparency. But doing it "for reals" would involve some huge bottom-up changes.
The "damn thing" was Soundjam, and it's what Apple purchased and turned into iTunes.
Absolutely not! That would be totally unprecedented!
Well, that's a _much_ more marginal case than ceasing security updates for an OS that was current last week.
AFP over TCP was introduced with that, macos 8? Meaning 1996ish? Given the immense degree to which IP is better suited to ethernet than DDP is, expecting people to transition within a span of seven years isn't exactly demanding.
From http://www.apple.com/powerbook/specs.html :
Dual display and video mirroring: Simultaneously supports up to 1024 by 768 pixels on the built-in display and up to 2048 by 1536 pixels on an external display, both at millions of colors
Even more people use cars, telephones, and clothes every day. Do you also feel that we need standardized education on crankshaft design, that all high school graduates should know the current provided by dialtone from an ESS7, and that all citizens should be familiar with cotton cultivation techniques?
The fact that these devices are ubiquitous is a cause and result of the fact that one does _not_ need to know arcana like "what a megabyte is" to use them.
The fact that no one uses it significantly alters the amount of security it provides.
I've seen people make this argument for at least ten years now. (And I suspect that it's been made far longer than that, and I'm just displaying my relative inexperience.)
It always turns out to be pretty shortsighted. Yes, an average new computer is overkill for most users the day it's purchased. Software is targetted at the average computer in use, new computers tend to be faster than the average, so new computers are always overkill. This is so definitional as to border on tautology.
But software continues to evolve to take advantage of new hardware resources.
It's easy to just dismiss this as laziness an inpcompetence on the part of software developers; to say that they're not really adding anything new, they can just afford to be sloppy now.
Obviously that's an oversimplification: some software becomes more demanding because it simply does things which were not practical with more limited resources. But even in the worst case, where new software is less efficient, I'd still call this a feature rather than a bug. If developers can spend less time on optimization, they can spend more time on more important things. Those efforts can go into clarity, security, consistency, or portability.
(Clearly some developers, notably those working for a certain little startup in Redmond, will not use these freed-up resources wisely, and will instead just take the opportunity to write a larger volume of bad code. But those are the same developers who probably would've buggered up the optimization process anyway, so this is really no worse.)
If we'd just stopped when we had enough hardware to speedily run common user applications, it would be too minimal to handle taxing things like tcp/ip, which you might have found to be handy.
And allow me the honor of being the next in a long line of people you'll meet who dislike tabs. They are a poor solution to the problem, and unquestionably implemented at the wrong layer.
If what you want is a single-gesture way of switching linearly through the windows in the current application, you want command-~, something all Cocoa applications get for free.
And tabs do have a cost to those who don't use them: the opportunity cost of the development time spent on them. Tabs don't implement themselves, and I'd almost certainly rather have whatever other feature those developers were creating instead.
Do you respond to news items about new operating systems with comments that operating systems have existed for decades? Or to articles on processors by pointing out that the abacus was around millennia ago?
The task is not new, and has been tackled before. This article discusses and compares the most current tools for the task. This is earthshattering to about the same degree that most technology news is.
640x480x32 appears to be 7372800 bits a second. This is considerably more data than one can reliably pass through usb1.1, despite its theoretical bandwidth of 11Mbps.
Those usb camera that do 640x480 are managing this by doing primitive hardware compression in the camera before stuffing it down the usb line. But of course they're cheap, so they're not doing good enough compression that you want to just send it over the net raw, you need to compress it again in software. At which point you're stacking lossy codecs, and you get crappier video.
Firewire camera have the bandwidth available (not only through sheer speed, but through isosynchronous transfers, which usb2 still lacks) to just send untouched 640x480x32, and your software has a clean shot at it.
Well, you could use a g4 tower as a midpoint between an imac and g5, giving you both expandability and some more speed. (Yes, they're still available, and range between $1200 and $1500).
My claims about the g5's speed were based upon the information revealed yesterday, which implies that they're actually significantly faster even than those dual 3ghz amd cpus. Of course, we'll need to wait to see them in person and verify how true that turns out to be, but the current data seem very promising.
And yes, if assembling a machine is something that you find to be fun, then obviously that's not a cost. And without that cost, I think it's safe to say that a personally-built machine will always be less expensive than any prebuilt one.
There is an "Eject" item in the contextual menu.
And there is an eject item in the Finder's File menu which is available when exclusively removable volumes are selected.
And there is a reliable and obvious key equivalent for that menu item (command-E).
And, in addition to these, there is the additional shortcut of dragging the volume to the trash.
Last year, my attempt to follow the several pages of instructions for installing Mathematica on linux failed. I eventually spent several unpleasant hours hacking through 50K of interdependent awk scripts, reworking some of them and manually handling some of the things they were failing to do. I got it working eventually, but it wasn't fun. And I certainly can't remember what all I did, so if I had to install it again tomorrow, I'd have to repeat the whole painful process.
Imagine how happy I was when I installed it the following day on macosx, and found that instructions consisted of "Drag the Mathematica icon to the Applications folder."
Have you not seen fink? Or is there some way in which you find it to not be apt-enough?
Unless that motherboard has a lot more than I'd guess, you've left out gigabit ethernet, firewire 400 and 800, optical audio in and out, and amplified analog audio out.
Perhaps more importantly, that system would be about half the speed of the mac to which you're comparing it.
Don't need GigE, optical audio, and all that speed? Then you should be comparing to an imac, which is still cheaper than what you've listed.
Certainly more importantly, you haven't accounted for your time spent selecting, purchasing, and assembling the parts. What's your consulting rate? Several hours work at a few hundred an hour dwarfs the cost of the hardware entirely.
One word for four is exactly the same phonetically as the the word for death: shi.
But the glyphs for them are completely different. This happens a lot in Japanese, which has a very limited sound palette and an incredibly rich character palette, so most people don't think much of it. (Because of this, the idea of a pun is pretty much meaningless in Japanese; every word sounds exactly like a dozen others anyway, but they're all conceptually distinct. (Though oddly enough, the exact same set of circumstances has lead to the Chinese being very punny.))
This does mean that four has a very mild reputation as an unlucky number, about comparable to thirteen in American culture. But I don't think anybody bothers to rename products to avoid it.
I suspect that Canon's reasoning was similar to Netscape's: "bigger numbers better."
Yeah, that function of Terminal is odd, and hard to give a description or name, but surprisingly handy. It allows you to simultaneously see and interact with two arbitrary segments of scrollback within a single window.
This facilitates doing a variant on a long series of steps you did a while earlier, comparing previous output of an operation to current output, or just reading deep scrollback without missing new data. It's not a life-changing feature, but it is regularly useful.
iTerm has given me the one thing that I've found seriously lacking in Terminal.app: configuration of what colors are used to display ANSI "colors". No more screaming yellow or illegible dark blue for me, thanks.
Unfortunately, iTerm does have a few limitations and bugs:
- while the xterm-experienced will like PgUp/PgDown going straight through, and using shift for local scrolling, I'd really like to see this togglable.
- no Home/End functionality, with our without shift.
- no local Find.
- it "helpfully" doesn't include whitespace when copying out of its windows. Actually, I did want that linefeed, thanks.
- periodically decides it wants to just sit and suck all my cpu until I kill it.
- font settings don't stick between launches.
I've also found that Terminal.app's split-window function is surprisingly useful. And unique, in my experience.