Re:Complexity is a good thing?
on
ZDNet Reviews iMovie
·
· Score: 5, Interesting
iMovie is movie making for the AOL crowd. FCP is for the, well, Mac crowd.
I agree and I don't. A division of my company does video integrations for broadcasters and post production companies. Two years ago those places were overflowing with Avids, a few Expresses but mostly Media Composers.
Now it's G4s with FCP as far as the eye can see.
Final Cut Pro on a Power Mac (about $6,000 total) is replacing Avid Media Composer systems (around $100,000) in professional settings.
Although I do snicker a little at the amount of time you took composing what was evidently a well thought-out-- albeit completely misguided-- response to what was basically a throwaway comment on my part.
I'm pretty sure that waiving your right to counsel invalidates any subsequent claim to inadequate counsel. If the judge didn't believe that this kid understood the consequences of self-representation without legal advice, he wouldn't have allowed the kid to waive his right. Once that happened, though, it was all over from a legal standpoint.
I think humor is one of the last things left that our peculiar western society allows us to make purely subjective value judgements about. Things that were once morally unthinkable are now valid lifestyle choices. Prejudice, even if based on experience or education, is unacceptable. Wanna shut down the conversation at a dinner party? Start expressing the opinion that certain things are just absolutely right and others are absolutely wrong.
But funny is always open to interpretation. If it ain't funny, you're free to say so. Loudly, if you want. Won't offend anybody.
So yeah, I guess you could say I subscribe to a universal humor scale whereby everything is judged. It's called my sense of humor, and I consider it absolute.
uuh...NO. they arent a backup medium. ever read the warranty information on the box ? they are most certainly NOT a backup medium. CDR's, AIT/DLT tapes are a backup medium.
uuh...NO. CDRs and magnetic tapes have lifespans measured in mere decades.
This is a backup medium. If it can't last 10,000 years submerged in salt water or in a mountain cave, I'm not interested.
How are they going to prevent customers from using the generic name?
As I understand it, we don't have to. According to our marketing department, we have to maintain a consistent use, within the company and with customers and partners, of the full name of the product. If we maintain that consistently, our trademark should be enforcable.
Part of the principle of fair use is that it's okay to quote from a published source, even at length, if your purpose in quoting is to comment directly on the material quoted.
So no, they shouldn't have to ask before using Slashdot content for their own means... if, as in this case, their purpose is to comment on what they quote.
In the words of a good friend, "It's all just ones and zeros."
And in the words of somebody undoubtedly very wise, "I ain't payin' that kind of money for a bunch o' ones and zeros, no matter how much time y'all spent puttin' 'em in just the right order."
Since the server is currently curled up in the corner, whimpering and mewling, I can't cut-and-paste and must paraphrase from memory. But as I remember it, the best one-liner from this article went something like this:
"When it comes to OS X and Linux, it seems that the market logic is AND, not OR."
(You may now begin the tired and meaningless flames about the difference between AND, OR, and XOR. No one will read them.;-) )
You apparently have never worked for a real company. If you had, you would know that, even if you ignore the costs associated with coming up with a new name in the first place, changing the name of a product can cost tens of thousands of dollars or more. Your suggestion that our marketing people are morons, based on the fact that they aren't suggesting a name change, is specious at best.
The difference is that we're trying really hard to make sure the generic term never enters the common use as a name for our product. If it does, we're boned.
"Windows", whether or not it's trademarkable is seldom called "Microsoft Windows" in the common household. It is "Windows" plain and simple. While this is MS's fault for choosing something that would *obviously* be shortened to "Windows" and thus be un-trademarkable, the public still associates "Windows" with MS.
You're absolutely right about this, in my opinion. What's more, my company is trying to learn from MS's mistake. We are soon going to release a product with the name "XYZ Genericword." (Obviously that's not the real name, but it's a generic nontrademarkable word preceeded by a trademarked three-letter acronym.)
During the last couple of months, we've gotten lazy and started referring to the product by the generic word only, leaving out the acronym. This sent our marketing people in orbit. Now they're on a crusade. Every time somebody uses the generic name alone in an email, or even verbally, they get corrected, loudly and publicly. Because we all know that if our product name gets reduced to that single generic word, we're up a creek as far as trademark protection goes.
It's not "training," really. In the demo I saw, they gave VideoLogger four or five video frames that had Saddam Hussein in them, and drew little boxes around his face to identify it, then assigned a keyword to it.
Then they ran some news footage through the system that had other pictures of Hussein in it. VideoLogger picked him out and assigned the keyword "Saddam Hussein" to the clip. It got did this on face recognition, not speech or CC recognition, because the video clip was from the Russian TV news!
It was pretty cool, even though it was just a demo.
All and all it seems like the big boys(IBM, EMC, Sun, STK, etc)
Just FYI, Sun doesn't actually make their high-end storage product. I think they call it the StorEdge 9900 or something but it's actually a rebranded Hitachi Data Systems 9960.
Funny thing about HDS. When you buy one of their 9960 systems-- a minimum investment of about $250,000-- you get a guarantee. If you ever lose any data at all on that storage system due to hardware or firmware fault, HDS will give you 30% of your purchase price back.
According to one of the senior HDS VPs that I spoke to last month, they've never had to pay out that penalty clause.
Bell Labs? Are you high? The system you're talking about is commercially available: it's called Virage VideoLogger. (I'd provide a link, but the Virage web site sucks so bad.... Just go to www.virage.com [the www is mandatory].)
VideoLogger has neato features like speech-to-text, speaker identification, face recognition, and keyframe extraction. All of those things happen in real time, if the PC is fast enough for it.
Combined with a half-decent RDBMS back-end, you can do stuff like search on "Saddam Hussein" and get back a reference to a clip that includes a picture of him, but not his actual name anywhere in the voiceover or the CC data. It's pretty cool.
It's also, like, $60,000 a copy, or something.
No, I don't work for Virage, and I've never had a business relationship with them. I've seen their stuff demoed, though.
more cache makes a difference to a smaller and smaller subset of programs
Question: what impact would (say) 8 MB of unified secondary cache have on a system running (say) 60 or so processes concurrently? (Not literally concurrently, of course, but scheduled in slices.)
While optimizing your code for cache friendliness helps, all the world is not a batch-scheduled supercomputer. If your main loop fits in cache, that's great... until the OS preempts your program and replaces all of its neatly organized cache lines with a bunch of stuff from the MP3 player that's running in the background.
I don't know enough about processor and cache internals, but it seems to me that overall system performance could be dramatically improved if the number of cache misses due to context switches were reduced.
Besides, the "cache is expensive" argument is just a matter of economies of scale. They said the same thing about RAM, and it was true until we started cranking out so much of the stuff that they're practically giving it away. I found a 64 MB stick in my cereal this morning.
I don't disagree with you at all, in principle. But look at the state of our industry. How successful do you think we'd be trying to implement a standard practice based on "discipline on the part of those writing shared libraries?"
The other day, while I was at work a plumber visited my house to fix some pipes. While he was there, he borrowed my wife and gave her incurable sexual diseases.... I'd feel the same way if he fixed my computer by installing M$.
This really isn't worth the trouble of a response, but here ya go anyway.
Your analogy of a plumber "borrowing" your wife is just silly, but I can't tell from your post whether you were kidding or what. That would obviously be a crime of assault, in this case sexual. That's totally different from everything we've been talking about, which is questions of property.
But to go with the second part of your analogy, installing software that interferes with your use of the computer (say it's buggy and it causes the computer to crash [even more] often) could be the crime of computer trespass.
In reading up on the McOwen case for another post, I read the Georgia Computer Systems Protection Act. Since I happen to have it open in another window:
"Any person who uses a computer or computer network with knowledge that such use is without authority and with the intention of:
(1) Deleting or in any way removing, either temporarily or permanently, any computer program or data from a computer or computer network;
(2) Obstructing, interrupting, or in any way interfering with the use of a computer program or data; or
(3) Altering, damaging, or in any way causing the malfunction of a computer, computer network, or computer program, regardless of how long the alteration, damage, or malfunction persists
shall be guilty of the crime of computer trespass."
So if somebody, Foo Inc., installs buggy third-party software alongside their application, and that buggy third-party software causes your computer to crash, then (if you lived in Georgia) you could make the case that Foo Inc. is guilty of computer trespass.
But there are two reasons why that wouldn't hold up. The standard, as spelled out in the statute, is that the person or party accused must have (1) used the computer without authorization, and (2) intended to interfere with your use of the computer, and/or cause the computer to malfunction.
For the first part, you authorized Foo Inc. to install software on your computer when you ran the installer. While it's true that the installer also installed third-party software, the party actually doing the installing is Foo Inc., and therefore they're the only ones that needed authorization.
But more important is the second part: they have to have intended to make your computer malfunction, or to interfere with your use of the computer. In other words, for Foo Inc. to be guilty of computer trespass, they had to have been acting maliciously. Of course, that's not the case.
So sorry. They aren't guilty of a crime. (At least, admittedly, in Georgia. But other computer crime statutes are similarly structured.)
I think this the case was based on 16-9-93(b) of the Georgia Computer System Protection Act. The anandtech.com article you pointed to refers to section (a) part 3, but I think they're misreading the statute.
Section (a) of that part defines the crime of computer theft. The whole definition is too long to excerpt here (read the whole act at this address), but the gist is that computer theft is the crime of using a computer without authority in an attempt to appropriate, obtain, or convert property illegally. In other words, if you commit theft, and you use a computer, without authorization, to do it, then you're guilty of computer theft. In that way, the legal definition of computer theft similar to that of the crime of armed robbery. No robbery, no armed robbery. Likewise, no theft, no computer theft.
Section (b), on the other hand, describes the crime of computer trespass. This crime covers using a computer without authorization to delete data, interfere with the normal use of the computer system, or alter or damage the computer. This is clearly where the meat of the McOwen case lay: his installing of the Distributed.net client without permission had the side effect (arguably) of interfering with the normal operations of the computer systems in question.
McOwen was only charged with one count of computer theft, but seven counts of computer trespass. The count of computer theft was probably justified by the fact that Distributed.net offers a $1,000 prize related to the cracking of RC5, thereby implying that McOwen acted to use a computer without authorization to illegally obtain property (the $1,000). The property in question, here, is the $1,000 prize. Not the computer capacity that he (allegedly) stole.
So this case, while interesting in its implications, has nothing to do with whether or not computer capacity (i.e., CPU cycles) is property.
Really interesting case, though. In a way, it's too bad McOwen decided to plead it out instead of taking it to trial. The judge's ruling would have been just fascinating.
That's a really good point, about cache misses and performance. Of course, the current crop of desktop CPUs feature something like 256 KB or 512 KB of cache, and that's just silly. The two-year-old R12000-based servers and workstations at work have 8 MB of secondary cache per CPU.
Look into the college that sued its employee when they installed a grid computing program to crack the RSA algorithm while everyone was on Christmas break.
This isn't very specific; Google doesn't know what to do with this. Can you please provide any specific information about the case?
Nothin'. Office 2001 for Mac OS 9 worked just like this. They had this little thing called the first-run utility, or something like that, that would scatter libraries throughout the System Folder.
It was kinda neat, too, to the extent that it would also perform sanity checks at application start up. If a critical component was missing or inconsistent, Office would fire off the first-run utility and fix it. Because only Office used those libraries, it never devolved into an argument between two programs trying to "repair" each other's libraries.
Under OS X, though, all the bits and pieces that would have been stored in the System Folder under OS 9 are stored in the application bundle itself. So each app has its own private copies of shared libraries.
As we've discussed many times, it's a trade-off between conserving disk and RAM (common libraries) and maximizing long-term system reliability (private libraries).
Component reuse is good in theory, but bad in practice because of versioning problems.
Component reuse is good in theory if and only if disk and RAM are scarce. As this scarcity lessens (hell, I have 640 MB of RAM in the iMac I'm using to write this, just because it was so damn cheap) the need for shared libraries becomes less important.
Of course, the trade-off is having every process on your system load its own copy of libc.
Slashdot Mirror
iMovie is movie making for the AOL crowd. FCP is for the, well, Mac crowd.
I agree and I don't. A division of my company does video integrations for broadcasters and post production companies. Two years ago those places were overflowing with Avids, a few Expresses but mostly Media Composers.
Now it's G4s with FCP as far as the eye can see.
Final Cut Pro on a Power Mac (about $6,000 total) is replacing Avid Media Composer systems (around $100,000) in professional settings.
I don't think anybody saw that coming.
(sigh)
You have taken all the fun out of it.
Although I do snicker a little at the amount of time you took composing what was evidently a well thought-out-- albeit completely misguided-- response to what was basically a throwaway comment on my part.
I'm pretty sure that waiving your right to counsel invalidates any subsequent claim to inadequate counsel. If the judge didn't believe that this kid understood the consequences of self-representation without legal advice, he wouldn't have allowed the kid to waive his right. Once that happened, though, it was all over from a legal standpoint.
Unless there's an appeal a-cookin', of course.
I think humor is one of the last things left that our peculiar western society allows us to make purely subjective value judgements about. Things that were once morally unthinkable are now valid lifestyle choices. Prejudice, even if based on experience or education, is unacceptable. Wanna shut down the conversation at a dinner party? Start expressing the opinion that certain things are just absolutely right and others are absolutely wrong.
But funny is always open to interpretation. If it ain't funny, you're free to say so. Loudly, if you want. Won't offend anybody.
So yeah, I guess you could say I subscribe to a universal humor scale whereby everything is judged. It's called my sense of humor, and I consider it absolute.
Nyah.
(HHOS)
uuh...NO. they arent a backup medium. ever read the warranty information on the box ? they are most certainly NOT a backup medium. CDR's, AIT/DLT tapes are a backup medium.
uuh...NO. CDRs and magnetic tapes have lifespans measured in mere decades.
This is a backup medium. If it can't last 10,000 years submerged in salt water or in a mountain cave, I'm not interested.
(HHOS)
HUMOUR
Two words:
NOT FUNNY.
Admittedly, humor of the absurd can be pretty damn funny. But this comment wasn't absurd. It was just dumb. Eye-rollingly, head-shakingly dumb.
How are they going to prevent customers from using the generic name?
As I understand it, we don't have to. According to our marketing department, we have to maintain a consistent use, within the company and with customers and partners, of the full name of the product. If we maintain that consistently, our trademark should be enforcable.
But who knows?
Part of the principle of fair use is that it's okay to quote from a published source, even at length, if your purpose in quoting is to comment directly on the material quoted.
So no, they shouldn't have to ask before using Slashdot content for their own means... if, as in this case, their purpose is to comment on what they quote.
In the words of a good friend, "It's all just ones and zeros."
And in the words of somebody undoubtedly very wise, "I ain't payin' that kind of money for a bunch o' ones and zeros, no matter how much time y'all spent puttin' 'em in just the right order."
Since the server is currently curled up in the corner, whimpering and mewling, I can't cut-and-paste and must paraphrase from memory. But as I remember it, the best one-liner from this article went something like this:
;-) )
"When it comes to OS X and Linux, it seems that the market logic is AND, not OR."
(You may now begin the tired and meaningless flames about the difference between AND, OR, and XOR. No one will read them.
You apparently have never worked for a real company. If you had, you would know that, even if you ignore the costs associated with coming up with a new name in the first place, changing the name of a product can cost tens of thousands of dollars or more. Your suggestion that our marketing people are morons, based on the fact that they aren't suggesting a name change, is specious at best.
To quote Homer: "Gentlemen, you have your moron."
The difference is that we're trying really hard to make sure the generic term never enters the common use as a name for our product. If it does, we're boned.
"Windows", whether or not it's trademarkable is seldom called "Microsoft Windows" in the common household. It is "Windows" plain and simple. While this is MS's fault for choosing something that would *obviously* be shortened to "Windows" and thus be un-trademarkable, the public still associates "Windows" with MS.
You're absolutely right about this, in my opinion. What's more, my company is trying to learn from MS's mistake. We are soon going to release a product with the name "XYZ Genericword." (Obviously that's not the real name, but it's a generic nontrademarkable word preceeded by a trademarked three-letter acronym.)
During the last couple of months, we've gotten lazy and started referring to the product by the generic word only, leaving out the acronym. This sent our marketing people in orbit. Now they're on a crusade. Every time somebody uses the generic name alone in an email, or even verbally, they get corrected, loudly and publicly. Because we all know that if our product name gets reduced to that single generic word, we're up a creek as far as trademark protection goes.
It's not "training," really. In the demo I saw, they gave VideoLogger four or five video frames that had Saddam Hussein in them, and drew little boxes around his face to identify it, then assigned a keyword to it.
Then they ran some news footage through the system that had other pictures of Hussein in it. VideoLogger picked him out and assigned the keyword "Saddam Hussein" to the clip. It got did this on face recognition, not speech or CC recognition, because the video clip was from the Russian TV news!
It was pretty cool, even though it was just a demo.
All and all it seems like the big boys(IBM, EMC, Sun, STK, etc)
Just FYI, Sun doesn't actually make their high-end storage product. I think they call it the StorEdge 9900 or something but it's actually a rebranded Hitachi Data Systems 9960.
Funny thing about HDS. When you buy one of their 9960 systems-- a minimum investment of about $250,000-- you get a guarantee. If you ever lose any data at all on that storage system due to hardware or firmware fault, HDS will give you 30% of your purchase price back.
According to one of the senior HDS VPs that I spoke to last month, they've never had to pay out that penalty clause.
Bell Labs? Are you high? The system you're talking about is commercially available: it's called Virage VideoLogger. (I'd provide a link, but the Virage web site sucks so bad.... Just go to www.virage.com [the www is mandatory].)
VideoLogger has neato features like speech-to-text, speaker identification, face recognition, and keyframe extraction. All of those things happen in real time, if the PC is fast enough for it.
Combined with a half-decent RDBMS back-end, you can do stuff like search on "Saddam Hussein" and get back a reference to a clip that includes a picture of him, but not his actual name anywhere in the voiceover or the CC data. It's pretty cool.
It's also, like, $60,000 a copy, or something.
No, I don't work for Virage, and I've never had a business relationship with them. I've seen their stuff demoed, though.
more cache makes a difference to a smaller and smaller subset of programs
Question: what impact would (say) 8 MB of unified secondary cache have on a system running (say) 60 or so processes concurrently? (Not literally concurrently, of course, but scheduled in slices.)
While optimizing your code for cache friendliness helps, all the world is not a batch-scheduled supercomputer. If your main loop fits in cache, that's great... until the OS preempts your program and replaces all of its neatly organized cache lines with a bunch of stuff from the MP3 player that's running in the background.
I don't know enough about processor and cache internals, but it seems to me that overall system performance could be dramatically improved if the number of cache misses due to context switches were reduced.
Besides, the "cache is expensive" argument is just a matter of economies of scale. They said the same thing about RAM, and it was true until we started cranking out so much of the stuff that they're practically giving it away. I found a 64 MB stick in my cereal this morning.
I don't disagree with you at all, in principle. But look at the state of our industry. How successful do you think we'd be trying to implement a standard practice based on "discipline on the part of those writing shared libraries?"
The other day, while I was at work a plumber visited my house to fix some pipes. While he was there, he borrowed my wife and gave her incurable sexual diseases.... I'd feel the same way if he fixed my computer by installing M$.
This really isn't worth the trouble of a response, but here ya go anyway.
Your analogy of a plumber "borrowing" your wife is just silly, but I can't tell from your post whether you were kidding or what. That would obviously be a crime of assault, in this case sexual. That's totally different from everything we've been talking about, which is questions of property.
But to go with the second part of your analogy, installing software that interferes with your use of the computer (say it's buggy and it causes the computer to crash [even more] often) could be the crime of computer trespass.
In reading up on the McOwen case for another post, I read the Georgia Computer Systems Protection Act. Since I happen to have it open in another window:
"Any person who uses a computer or computer network with knowledge that such use is without authority and with the intention of:
(1) Deleting or in any way removing, either temporarily or permanently, any computer program or data from a computer or computer network;
(2) Obstructing, interrupting, or in any way interfering with the use of a computer program or data; or
(3) Altering, damaging, or in any way causing the malfunction of a computer, computer network, or computer program, regardless of how long the alteration, damage, or malfunction persists
shall be guilty of the crime of computer trespass."
So if somebody, Foo Inc., installs buggy third-party software alongside their application, and that buggy third-party software causes your computer to crash, then (if you lived in Georgia) you could make the case that Foo Inc. is guilty of computer trespass.
But there are two reasons why that wouldn't hold up. The standard, as spelled out in the statute, is that the person or party accused must have (1) used the computer without authorization, and (2) intended to interfere with your use of the computer, and/or cause the computer to malfunction.
For the first part, you authorized Foo Inc. to install software on your computer when you ran the installer. While it's true that the installer also installed third-party software, the party actually doing the installing is Foo Inc., and therefore they're the only ones that needed authorization.
But more important is the second part: they have to have intended to make your computer malfunction, or to interfere with your use of the computer. In other words, for Foo Inc. to be guilty of computer trespass, they had to have been acting maliciously. Of course, that's not the case.
So sorry. They aren't guilty of a crime. (At least, admittedly, in Georgia. But other computer crime statutes are similarly structured.)
I think this the case was based on 16-9-93(b) of the Georgia Computer System Protection Act. The anandtech.com article you pointed to refers to section (a) part 3, but I think they're misreading the statute.
Section (a) of that part defines the crime of computer theft. The whole definition is too long to excerpt here (read the whole act at this address), but the gist is that computer theft is the crime of using a computer without authority in an attempt to appropriate, obtain, or convert property illegally. In other words, if you commit theft, and you use a computer, without authorization, to do it, then you're guilty of computer theft. In that way, the legal definition of computer theft similar to that of the crime of armed robbery. No robbery, no armed robbery. Likewise, no theft, no computer theft.
Section (b), on the other hand, describes the crime of computer trespass. This crime covers using a computer without authorization to delete data, interfere with the normal use of the computer system, or alter or damage the computer. This is clearly where the meat of the McOwen case lay: his installing of the Distributed.net client without permission had the side effect (arguably) of interfering with the normal operations of the computer systems in question.
McOwen was only charged with one count of computer theft, but seven counts of computer trespass. The count of computer theft was probably justified by the fact that Distributed.net offers a $1,000 prize related to the cracking of RC5, thereby implying that McOwen acted to use a computer without authorization to illegally obtain property (the $1,000). The property in question, here, is the $1,000 prize. Not the computer capacity that he (allegedly) stole.
So this case, while interesting in its implications, has nothing to do with whether or not computer capacity (i.e., CPU cycles) is property.
Really interesting case, though. In a way, it's too bad McOwen decided to plead it out instead of taking it to trial. The judge's ruling would have been just fascinating.
That's a really good point, about cache misses and performance. Of course, the current crop of desktop CPUs feature something like 256 KB or 512 KB of cache, and that's just silly. The two-year-old R12000-based servers and workstations at work have 8 MB of secondary cache per CPU.
You'd rather need to update every single application after (to take a timely example) a zlib bug gets discovered?
Given the choice between that and the current state of shared code in consumer systems? Yes.
Look into the college that sued its employee when they installed a grid computing program to crack the RSA algorithm while everyone was on Christmas break.
This isn't very specific; Google doesn't know what to do with this. Can you please provide any specific information about the case?
Nothin'. Office 2001 for Mac OS 9 worked just like this. They had this little thing called the first-run utility, or something like that, that would scatter libraries throughout the System Folder.
It was kinda neat, too, to the extent that it would also perform sanity checks at application start up. If a critical component was missing or inconsistent, Office would fire off the first-run utility and fix it. Because only Office used those libraries, it never devolved into an argument between two programs trying to "repair" each other's libraries.
Under OS X, though, all the bits and pieces that would have been stored in the System Folder under OS 9 are stored in the application bundle itself. So each app has its own private copies of shared libraries.
As we've discussed many times, it's a trade-off between conserving disk and RAM (common libraries) and maximizing long-term system reliability (private libraries).
Component reuse is good in theory, but bad in practice because of versioning problems.
Component reuse is good in theory if and only if disk and RAM are scarce. As this scarcity lessens (hell, I have 640 MB of RAM in the iMac I'm using to write this, just because it was so damn cheap) the need for shared libraries becomes less important.
Of course, the trade-off is having every process on your system load its own copy of libc.