I don't understand OLPC - at all. [...] Sure, it's great to give a kid a computer. Isn't it better to give the kid some medicine, drinking water, or food?
It is a very subtle and idealistic concept. The entire idea behind it is that the real problem of the third world isn't about medicine, drinking water, or food -- the OLPC is really targeted at kids that have their survival needs taken care of. It's based on the assumption that the real problem is a lack of education and access to information. That if you could give children these two, they would be able to obtain better medicine, drinking water and food.
You guys may find it hard to believe, but there are places that are three days away from ELECTRICITY.
That's why it can be wound up.
A kid spending his day farming isn't going to say, "man, I could really go for a/. break right now."
Really? I suppose he wouldn't be too interested in the Natalie Portman jokes or iPhone banter, but neither most poor people nor most slashdotters are so insular and parochial. The OLPC and the Internet facilitate people talking to people, and is thus an absolute good.
This attitude that all slashdotters are ignorant of poverty, and that all the poor people on Earth have no interest in technology or the availability of information is deeply snobbish, imho. I don't know if this was exactly your point, but I just wanna be on the record as against it;).
Labels ensure that the only top-tier, signed artists get play in each of the mediums you've mentioned. They serve to keep the little guy out.
If a film or television director hears your song and wants it, he's going to get it. The real problem is that very few big movies are directed by people that wanna get new, edgy, or slightly-less-than-commercial music from untried acts. The call to the music supervisor is generally "Get me something that's by a hot band to put over the snowboarding scene," the actual band, performer, or subject is really secondary to the commercial aspect. Only occasionally do you hear "Get me band x's song y".
This is the area most in flux. "Publishing" can mean many things, and there's no reason that copyright law couldn't be simplified to make collecting royalties something much more manageable
I agree, but this reform had better happen soon if people distributing on the Internet want ay protection at all.
Physical media is dying. You're just going to have to deal with that. It makes absolutely no sense to continue distributing music on CDs when it's possible to distribute bit-perfect copies at a lower cost over the web.
Eh, I'm not alone. I guess we'll just have to see, but for just tracks of audio, you're right. It's left to see if the recording industry won't start offering more value for the physical media than they do now, like lyrics embedded as a standard, easier ripping, videos, promo materials... huh, you're right, it's dead.
Again, labels are the reason why giant budgets are required to promote songs. They serve to reinforce the status quo, which means Joe Anotherguitarplayer can't compete.
Well, you can't make musician A disappear from the Earth by heavily promoting musician B. Again the real problem is that people don't always have strong feelings about particular songs or artists, but just wanna hear music, any music, maybe restricted to a genre but that's it.
And, well, guess what, if you want 100 million people to hear the hook to your song or even read the name of your band, that's going to cost you millions of dollars in media buys, sorry. The average music listener doesn't spend an hour a day on band home pages looking for good music. There's this whole line of work called "A&R" with professionals to do that for them. You may not like what they come up with, so you're welcome to start your own music website, too.
See #2.
In flux, yes, but a big fact remains that if you create music that only exists as a recording, i.e. you don't have a live act, it's going to be very difficult for you to make a living doing your art. The thing left unstated in these debates oftentimes is that the people advocating more freedom really don't care if they only do their music as a hobby. I myself am certain quality would suffer under such a regime, but again this remains to be seen.
When a band can distribute its albums by posting a zip file on a web site, there's a lot less incentive to turn to labels.
Great, they can put their song into people's ears. In terms of what the recording industry does, that just leaves:
Getting the song placed in motion pictures, television, and other media
Publishing the song, allowing the writers to collect royalties from covers, remixes and samples
Packaging and selling physical media for those of us who demand it (like, say, me)
Promoting the song so that items 1-4 actually happen, thus driving demand for live performances
Making money, thus making the job of "professional recording artist" an actual proposition
Now to be fair, a lot of these can be handled by a good agent and a good manager, and as we see, many of the recording labels are reforming themselves as agents and management companies. Thus the new revenue comes from licensing and live performances, instead of CDs.
Characters often read books on TNG and in the movies, and not just antiques. I seem to remember a few VIP presentations (one in ST6 stands out in my mind) that were actually delivered with a presentation pad and an easel, which I think made an important statement about the plain utility of such things.
California's testing and qualification of voting machines -- this time around, at least -- was quite formidable, including tiger teams to hack the machines. The report from the testers was the reason they all got sent back in the first place.
Computer time always counts in the number of SI seconds since the Epoch, which is 0:00:00, January 1, 1970 UTC.
There's a lot of subtlety in this, given that UTC didn't exist in 1970, and what they did was simply decide that January 1, 1972 00:00 at Greenwich was equal to 63,072,000 (2 years), so representations of times before 1970 are somewhat ill-defined -- good enough for printing timecards, but maybe not for recording astronomical events.
Also, if we're talking Unix time, modern Unix hosts set their RTC to UTC provided by NTP; the clock only counts a continuous number of seconds if you leave it alone, and when leap seconds happen, the Unix time is discontinuous at that moment.
Example: If you update your clock every 60 seconds (let's say at 0:30 every minute), and you do an update at 11:59:30PM before a leap second, and 12:00:30AM after one, the time_t value for the first and second checks will differ by 61 seconds, even though only 60 elapsed. There is a Unix epoch and a Unix timescale, but there's no real effort to maintain it, and POSIX doesn't even require you to keep your clock accurate.
It's a very interesting reading, I think you have your thesis!
This discussion of course brings up the case of Kismet the Robot, to which many fully mature adults display an emotional response. They may KNOW that the robot has no emotions itself, but a smiley face and big eyes that respond to even a small repertory of facial responses in the human is enough to create the impression of sympathy, and enough to elicit an unconscious smile back from the human.
Getting people to attribute agency or emotions onto inanimate objects has a long history of course, just ask a Totem Pole;). A lot of people are convinced their computer has a personality, after all.
I think the the real information that Facebook vends isn't your interests so much as who you know.
There's this quote from Hannah Arendt that Facebook always makes me think of:
The Okhrana, the Czarist predecessor of the GPU, is reported to have invented a filing system in which every suspect was noted on a large card in the center of which his name was surrounded by a red circle; his political friends were designated by smaller red circles and his nonpolitical acquaintances by green ones; brown circles indicated persons in contact with friends of the suspect but not known to him personally; cross-relationships between the suspect's friends, political and nonpolitical, and the friends of his friends were indicated by lines between the respective circles. [...] this is the utopian goal of the totalitarian secret police: a look at the gigantic map on the office wall should suffice at any given moment to establish, not who is who or who thinks what, but who is related to whom and in what degree or kind of intimacy.
But this is private enterprise, so they may no knock down your door, but what if someone identified as your best friend were to publicly criticize your employer for some policy? They could demand you provide information on them (and in an at-will hire state your job would be on the line), or they could characterize you as a potential risk due to your association and this might cause you to lose a promotion. If someone hasn't already written a facebook API that quantifies an employee's risk to an organization based on his friends, I'm sure it's coming.
I think I have a friendster account, I made it in 2005 when all the pretty young things in Hollywood were using it for their "social" (sing. n.). I haven't used it since the week I made it, I doubt I could even log into it now (everybody went to myspace after about a month, and then to facebook, and nobody ever checks their page anyways). I suspect the number of accounts many social networking sites report come with the same caveat. I think the real question is "How many logins have at least one access per week? Per day?" The real gem is owning a website that people use every day and keep current with real information.
The salesforce.com numbers are the most interesting, since the contacts that particular corpus tracks are recorded by paying subscribers, and are used in commercial transactions, and not by vapid teeners to promote their band or unite with other Gossip Girls fans.
Of course, this information is really useful if you're selling cheap amps or advertising time for the CW, but considering the quality of the source, you might as well be using wikipedia to find the exchange rate of the dollar.
It sounds like this trojan comes with a local privilege escalation vulnerability otherwise this also depends on users on Macs having root level access.
Stare argumentum; this executable in question makes no use of an exploit, the OS behaves exactly as the user commands.
OS X most certainly does have a root user, it's *interactive* logins by root that are disabled by default.
Not just interactive logins, logins period. There is no process you can undertake by which you will be recognized as real user 0 without setuid(), thus you already need to be euid 0, and thus you must be either a sudoer and recently authenticated or running a binary owned by root. I think the distinction is semantic and doesn't advance on the original point the poster made. "Users" on Macs don't have root-level access, they only have the privilege of running a program with euid of root, given they enter their password. That's very different from the implied "they all run in admin mode" of the parent.
To my great shock and amazement it decodes 4 channels from current DVDs just fine.
You have to hand it to those Dolby guys, they really figured out a good system with Dolby Stereo. We still do Dolby SR/Stereo printmasters the same day we do the 5.1 printmasters, with the same mixers, and generally with the director present; it's just as authoritative artistically as the 5.1, and it's still just about the most compatible format there is for a wide class of room types and playback systems. It lacks a little awesomeness in a large room, but for any 800sq/ft or smaller room, or any overly long or broad room, I think it's the best compromise you can have. Split surrounds are a waste unless your head is at least 5-10 away from the closest one, IMHO.
It sounds like this trojan comes with a local privilege escalation vulnerability otherwise this also depends on users on Macs having root level access.
Mac OS X has no root user by default, and as a potential victim you need to type in an admin password in order to install the trojan. Even if "open safe files" was turned on (which it isn't any more), it won't install until you double-click on the package installer, click thru the installer, and type in an admin password.
I hear Leopard adds full Python support to XCode, though I haven't had a change to test it out. Even without, I haven't yet found anything I can't do with Python on OS X that I can do with Objective-C.
Not just Python support, friend, but a complete bridge into the Cocoa classes. Take a look. I'm a Rubinista myself, but this stuff is pretty slick. All I write is in-house stuff, and this stuff is a serious gift. Who needs a CLR when you can just call/usr/bin/ruby?
Almost 6 years since they built the serious pain-in-the-butt Java-Cocoa bridge, and no popular java applications written for OS X. Still waiting, Java developers. Of course, it's all deprectated now, because after 6 years no one used it. I don't know if they kept statistics on it, but I bet you could count the number of times someone typed extends com.apple.cocoa.foundation on one hand. If java had taken off as a good way to write applications for the platform, then maybe things would've been different, but there really haven't been any. None, really, that I can think of.
The java devs are not completely forgotten, Apple still has webobjects, but just how much energy are they supposed to put into keeping the Java kit up-to-date just to keep java developers happy, particularly if they don't share any of their happiness with common users? Aside from good press from the grey old men of Java, what exactly does Java bring to Mac OS X? If I were allocating resources in Cupertino, hell yes I would get Time Machine debugged before I spent a man-hour on Java 6.
I swear the only time I hear about this issue is when Gosling blogs on it, and then it is copied off to the rest of the intarweb as an utterance from the mountaintop. Is this what we Ruby on Rails developers sound like?:)
Don't get me wrong man, all of that's completely correct, it's just irrelevant.
For me the original issue was "is Mac OS X going to remain a relevant platform for development without a managed runtime?" To which I respond in the affirmative..NET is just Microsoft's way of getting to the same place Apple is; where Apple controls their hardware platform, Microsoft abstracts it away, and where Vista uses the.NET CLR to provide runtime security and introspection, OS X simply does it in the Objective-C runtime and in the kernel with a different overall security model and tools like DTrace.
I guess.NET is provably more safe then running on the registers, but so was Smalltalk and Lisp, and look where they are now! Vista/CLR and OS X/Objective-C use very different techniques to provide an application environment that is, for all intents and purposes, from the perspective of the user, equivalent. The user experience is all that matters in my formulation of the issue, and users on OS X do not report the problems that come with unsafe languages running on, say, XP. All of this rather offensive talk of safe references, runtime introspection, and strict typing has nothing to do directly with the end-user experience, and most of the security and runtime benefits of a managed environment can be provided by other mechanisms, be they in the language or the client OS.
Developers can have an easy time or hard time, but if users like the overall platform, the devs will put up with a little pain to deliver the application in the users preferred platform. This is very much MS's modus operandi compared to Apple's. The former thinks "Make the devs happy and the users will buy what the devs build on," and this has worked historically. Apple's thinking is "Make the users happy and the devs will write on our platform," and this works too.
Like Python (first class and installed by default for years)?
Or the newcomer, Ruby?
It's really not exactly the same, but it's probably sufficient. Just to play around today I built a little Ruby application against CoreData for saving a few related tables and saved thumbnail images, and it only took me about 10 lines of code on top of the boilerplate (and schema definitions, Interface Builder stuff, etc.) Archiving/Unarchiving the image and making the edit view receive it from a drag-n-drop event didn't require any code either, I was happy to notice. The Ruby/Python system is extremely good for making a quickie one-off program for yourself -- a demo project in the dev package is an RSS reader that pulls pictures down and lets you apply CoreImage filters among other things, very styling, beautiful UI and extremely fast to code, it's probably 100 lines. I might not ever write a CLI ruby script again, the Cocoa links and IB stuff are too easy.
The parent wasn't really talking about this, though. The desire is for some kind of first-class "environment" that doesn't just "bridge" into the OS framework but is actually used to implement it. Bound into this is "safe language" snobbery: there's this attitude among a certain class of programmer that if a language lets you put a *star in front of something, the language is inferior, will cause NORAD to launch missiles, will trash your ~/, and "crashes alot". The Siracusa quote on the issue might be:
In Objective-C, on the other hand, there is no gate [compared to C#]; it's all just one big, grassy field. Raw C code is just a keystroke away at all times. Again, this arrangement is a strength today. But in the future, I think the desire for "safety by default" will eventually win out, and Objective-C's intimate relationship with C will be perceived as a dangerous relic.
Of course, among the end users, Objective-C isn't perceived as much of anything. It ain't perfect but crashing and security aren't really issues on Mac OS X, and are quite effectively addressed (for the time being) in other ways. I've never ever ever heard an end user observe that programs on Mac OS X "crash more" or "install adware more" than programs on XP or Vista. This could change, but when is it going to happen and how?
That said, even when Apple came out with it's "safe" development environment and runtime, what devs would use it? I would, maybe Siracusa would, we're little ISVs. Maybe hobbyists and students and ISVs writing FTP software would, but nobody buys computers to run FTP programs or hobbyist projects; people buy computers to run Word, and Illustrator, and Final Cut Pro, and Cubase, and on and on. These vendors don't give a flying fuck about.NET, and won't probably ever, and we have yet to see any.NET or managed-runtime competitors for them.
I think the whole "safe language" debate is just a bunch of devs arguing over wether or not MS and Sun have discovered the "next magic bullet." Which, if you get my reference, lets you know what I think.
Slashdot Mirror
It is a very subtle and idealistic concept. The entire idea behind it is that the real problem of the third world isn't about medicine, drinking water, or food -- the OLPC is really targeted at kids that have their survival needs taken care of. It's based on the assumption that the real problem is a lack of education and access to information. That if you could give children these two, they would be able to obtain better medicine, drinking water and food.
You guys may find it hard to believe, but there are places that are three days away from ELECTRICITY.That's why it can be wound up.
A kid spending his day farming isn't going to say, "man, I could really go for aReally? I suppose he wouldn't be too interested in the Natalie Portman jokes or iPhone banter, but neither most poor people nor most slashdotters are so insular and parochial. The OLPC and the Internet facilitate people talking to people, and is thus an absolute good.
This attitude that all slashdotters are ignorant of poverty, and that all the poor people on Earth have no interest in technology or the availability of information is deeply snobbish, imho. I don't know if this was exactly your point, but I just wanna be on the record as against it ;).
The gravimetric distortions are only a problem if you miss the annual baryon sweeps. The real concern is chromometric distortions and temporal wakes.
And MacGuyver..
If a film or television director hears your song and wants it, he's going to get it. The real problem is that very few big movies are directed by people that wanna get new, edgy, or slightly-less-than-commercial music from untried acts. The call to the music supervisor is generally "Get me something that's by a hot band to put over the snowboarding scene," the actual band, performer, or subject is really secondary to the commercial aspect. Only occasionally do you hear "Get me band x's song y".
This is the area most in flux. "Publishing" can mean many things, and there's no reason that copyright law couldn't be simplified to make collecting royalties something much more manageableI agree, but this reform had better happen soon if people distributing on the Internet want ay protection at all.
Physical media is dying. You're just going to have to deal with that. It makes absolutely no sense to continue distributing music on CDs when it's possible to distribute bit-perfect copies at a lower cost over the web.Eh, I'm not alone. I guess we'll just have to see, but for just tracks of audio, you're right. It's left to see if the recording industry won't start offering more value for the physical media than they do now, like lyrics embedded as a standard, easier ripping, videos, promo materials... huh, you're right, it's dead.
Again, labels are the reason why giant budgets are required to promote songs. They serve to reinforce the status quo, which means Joe Anotherguitarplayer can't compete.Well, you can't make musician A disappear from the Earth by heavily promoting musician B. Again the real problem is that people don't always have strong feelings about particular songs or artists, but just wanna hear music, any music, maybe restricted to a genre but that's it.
And, well, guess what, if you want 100 million people to hear the hook to your song or even read the name of your band, that's going to cost you millions of dollars in media buys, sorry. The average music listener doesn't spend an hour a day on band home pages looking for good music. There's this whole line of work called "A&R" with professionals to do that for them. You may not like what they come up with, so you're welcome to start your own music website, too.
See #2.In flux, yes, but a big fact remains that if you create music that only exists as a recording, i.e. you don't have a live act, it's going to be very difficult for you to make a living doing your art. The thing left unstated in these debates oftentimes is that the people advocating more freedom really don't care if they only do their music as a hobby. I myself am certain quality would suffer under such a regime, but again this remains to be seen.
Great, they can put their song into people's ears. In terms of what the recording industry does, that just leaves:
Now to be fair, a lot of these can be handled by a good agent and a good manager, and as we see, many of the recording labels are reforming themselves as agents and management companies. Thus the new revenue comes from licensing and live performances, instead of CDs.
At the risk of being pedantic, Bush the Elder was president when the Soviet Union collapsed.
....Ahem
Characters often read books on TNG and in the movies, and not just antiques. I seem to remember a few VIP presentations (one in ST6 stands out in my mind) that were actually delivered with a presentation pad and an easel, which I think made an important statement about the plain utility of such things.
Allovertheworld!
Hikeeba!
MITCHELL!
I'm about to google it, but what's database normalization?
California's testing and qualification of voting machines -- this time around, at least -- was quite formidable, including tiger teams to hack the machines. The report from the testers was the reason they all got sent back in the first place.
There's a lot of subtlety in this, given that UTC didn't exist in 1970, and what they did was simply decide that January 1, 1972 00:00 at Greenwich was equal to 63,072,000 (2 years), so representations of times before 1970 are somewhat ill-defined -- good enough for printing timecards, but maybe not for recording astronomical events.
Also, if we're talking Unix time, modern Unix hosts set their RTC to UTC provided by NTP; the clock only counts a continuous number of seconds if you leave it alone, and when leap seconds happen, the Unix time is discontinuous at that moment.
Example: If you update your clock every 60 seconds (let's say at 0:30 every minute), and you do an update at 11:59:30PM before a leap second, and 12:00:30AM after one, the time_t value for the first and second checks will differ by 61 seconds, even though only 60 elapsed. There is a Unix epoch and a Unix timescale, but there's no real effort to maintain it, and POSIX doesn't even require you to keep your clock accurate.
Disney's board also contains the CEO of a Blu-Ray member. His name's Steve Jobs, I forget the name of his company at the moment...
It's a very interesting reading, I think you have your thesis!
This discussion of course brings up the case of Kismet the Robot, to which many fully mature adults display an emotional response. They may KNOW that the robot has no emotions itself, but a smiley face and big eyes that respond to even a small repertory of facial responses in the human is enough to create the impression of sympathy, and enough to elicit an unconscious smile back from the human.
Getting people to attribute agency or emotions onto inanimate objects has a long history of course, just ask a Totem Pole ;). A lot of people are convinced their computer has a personality, after all.
I think the the real information that Facebook vends isn't your interests so much as who you know.
There's this quote from Hannah Arendt that Facebook always makes me think of:
The Okhrana, the Czarist predecessor of the GPU, is reported to have invented a filing system in which every suspect was noted on a large card in the center of which his name was surrounded by a red circle; his political friends were designated by smaller red circles and his nonpolitical acquaintances by green ones; brown circles indicated persons in contact with friends of the suspect but not known to him personally; cross-relationships between the suspect's friends, political and nonpolitical, and the friends of his friends were indicated by lines between the respective circles. [...] this is the utopian goal of the totalitarian secret police: a look at the gigantic map on the office wall should suffice at any given moment to establish, not who is who or who thinks what, but who is related to whom and in what degree or kind of intimacy.But this is private enterprise, so they may no knock down your door, but what if someone identified as your best friend were to publicly criticize your employer for some policy? They could demand you provide information on them (and in an at-will hire state your job would be on the line), or they could characterize you as a potential risk due to your association and this might cause you to lose a promotion. If someone hasn't already written a facebook API that quantifies an employee's risk to an organization based on his friends, I'm sure it's coming.
I think I have a friendster account, I made it in 2005 when all the pretty young things in Hollywood were using it for their "social" (sing. n.). I haven't used it since the week I made it, I doubt I could even log into it now (everybody went to myspace after about a month, and then to facebook, and nobody ever checks their page anyways). I suspect the number of accounts many social networking sites report come with the same caveat. I think the real question is "How many logins have at least one access per week? Per day?" The real gem is owning a website that people use every day and keep current with real information.
The salesforce.com numbers are the most interesting, since the contacts that particular corpus tracks are recorded by paying subscribers, and are used in commercial transactions, and not by vapid teeners to promote their band or unite with other Gossip Girls fans.
Of course, this information is really useful if you're selling cheap amps or advertising time for the CW, but considering the quality of the source, you might as well be using wikipedia to find the exchange rate of the dollar.
Before this decade is out, we should dedicate ourselves to defeating Pennywise the Clown and his evil minions!
The GP:
It sounds like this trojan comes with a local privilege escalation vulnerability otherwise this also depends on users on Macs having root level access.Stare argumentum; this executable in question makes no use of an exploit, the OS behaves exactly as the user commands.
OS X most certainly does have a root user, it's *interactive* logins by root that are disabled by default.Not just interactive logins, logins period. There is no process you can undertake by which you will be recognized as real user 0 without setuid(), thus you already need to be euid 0, and thus you must be either a sudoer and recently authenticated or running a binary owned by root. I think the distinction is semantic and doesn't advance on the original point the poster made. "Users" on Macs don't have root-level access, they only have the privilege of running a program with euid of root, given they enter their password. That's very different from the implied "they all run in admin mode" of the parent.
You have to hand it to those Dolby guys, they really figured out a good system with Dolby Stereo. We still do Dolby SR/Stereo printmasters the same day we do the 5.1 printmasters, with the same mixers, and generally with the director present; it's just as authoritative artistically as the 5.1, and it's still just about the most compatible format there is for a wide class of room types and playback systems. It lacks a little awesomeness in a large room, but for any 800sq/ft or smaller room, or any overly long or broad room, I think it's the best compromise you can have. Split surrounds are a waste unless your head is at least 5-10 away from the closest one, IMHO.
Mac OS X has no root user by default, and as a potential victim you need to type in an admin password in order to install the trojan. Even if "open safe files" was turned on (which it isn't any more), it won't install until you double-click on the package installer, click thru the installer, and type in an admin password.
See Confusopoly.
It's coming in a week?
Not just Python support, friend, but a complete bridge into the Cocoa classes. Take a look. I'm a Rubinista myself, but this stuff is pretty slick. All I write is in-house stuff, and this stuff is a serious gift. Who needs a CLR when you can just call /usr/bin/ruby?
Almost 6 years since they built the serious pain-in-the-butt Java-Cocoa bridge, and no popular java applications written for OS X. Still waiting, Java developers. Of course, it's all deprectated now, because after 6 years no one used it. I don't know if they kept statistics on it, but I bet you could count the number of times someone typed extends com.apple.cocoa.foundation on one hand. If java had taken off as a good way to write applications for the platform, then maybe things would've been different, but there really haven't been any. None, really, that I can think of.
The java devs are not completely forgotten, Apple still has webobjects, but just how much energy are they supposed to put into keeping the Java kit up-to-date just to keep java developers happy, particularly if they don't share any of their happiness with common users? Aside from good press from the grey old men of Java, what exactly does Java bring to Mac OS X? If I were allocating resources in Cupertino, hell yes I would get Time Machine debugged before I spent a man-hour on Java 6.
I swear the only time I hear about this issue is when Gosling blogs on it, and then it is copied off to the rest of the intarweb as an utterance from the mountaintop. Is this what we Ruby on Rails developers sound like? :)
Don't get me wrong man, all of that's completely correct, it's just irrelevant.
For me the original issue was "is Mac OS X going to remain a relevant platform for development without a managed runtime?" To which I respond in the affirmative. .NET is just Microsoft's way of getting to the same place Apple is; where Apple controls their hardware platform, Microsoft abstracts it away, and where Vista uses the .NET CLR to provide runtime security and introspection, OS X simply does it in the Objective-C runtime and in the kernel with a different overall security model and tools like DTrace.
I guess .NET is provably more safe then running on the registers, but so was Smalltalk and Lisp, and look where they are now! Vista/CLR and OS X/Objective-C use very different techniques to provide an application environment that is, for all intents and purposes, from the perspective of the user, equivalent. The user experience is all that matters in my formulation of the issue, and users on OS X do not report the problems that come with unsafe languages running on, say, XP. All of this rather offensive talk of safe references, runtime introspection, and strict typing has nothing to do directly with the end-user experience, and most of the security and runtime benefits of a managed environment can be provided by other mechanisms, be they in the language or the client OS.
Developers can have an easy time or hard time, but if users like the overall platform, the devs will put up with a little pain to deliver the application in the users preferred platform. This is very much MS's modus operandi compared to Apple's. The former thinks "Make the devs happy and the users will buy what the devs build on," and this has worked historically. Apple's thinking is "Make the users happy and the devs will write on our platform," and this works too.
It's really not exactly the same, but it's probably sufficient. Just to play around today I built a little Ruby application against CoreData for saving a few related tables and saved thumbnail images, and it only took me about 10 lines of code on top of the boilerplate (and schema definitions, Interface Builder stuff, etc.) Archiving/Unarchiving the image and making the edit view receive it from a drag-n-drop event didn't require any code either, I was happy to notice. The Ruby/Python system is extremely good for making a quickie one-off program for yourself -- a demo project in the dev package is an RSS reader that pulls pictures down and lets you apply CoreImage filters among other things, very styling, beautiful UI and extremely fast to code, it's probably 100 lines. I might not ever write a CLI ruby script again, the Cocoa links and IB stuff are too easy.
The parent wasn't really talking about this, though. The desire is for some kind of first-class "environment" that doesn't just "bridge" into the OS framework but is actually used to implement it. Bound into this is "safe language" snobbery: there's this attitude among a certain class of programmer that if a language lets you put a *star in front of something, the language is inferior, will cause NORAD to launch missiles, will trash your ~/, and "crashes alot". The Siracusa quote on the issue might be:
In Objective-C, on the other hand, there is no gate [compared to C#]; it's all just one big, grassy field. Raw C code is just a keystroke away at all times. Again, this arrangement is a strength today. But in the future, I think the desire for "safety by default" will eventually win out, and Objective-C's intimate relationship with C will be perceived as a dangerous relic.Of course, among the end users, Objective-C isn't perceived as much of anything. It ain't perfect but crashing and security aren't really issues on Mac OS X, and are quite effectively addressed (for the time being) in other ways. I've never ever ever heard an end user observe that programs on Mac OS X "crash more" or "install adware more" than programs on XP or Vista. This could change, but when is it going to happen and how?
That said, even when Apple came out with it's "safe" development environment and runtime, what devs would use it? I would, maybe Siracusa would, we're little ISVs. Maybe hobbyists and students and ISVs writing FTP software would, but nobody buys computers to run FTP programs or hobbyist projects; people buy computers to run Word, and Illustrator, and Final Cut Pro, and Cubase, and on and on. These vendors don't give a flying fuck about .NET, and won't probably ever, and we have yet to see any .NET or managed-runtime competitors for them.
I think the whole "safe language" debate is just a bunch of devs arguing over wether or not MS and Sun have discovered the "next magic bullet." Which, if you get my reference, lets you know what I think.