That may be a special exception, as Amtrak can advertise on their own stations, in trains. Perhaps that connection extends to almost all public transit.
Also they may not have had to pay for it, as there are plenty of mutally beneficial relations that amtrak establishes between public transit systems, especially commuter railroads.
I meant to hit preview, and I meant to close the italics. DOH!
And I hate slashdot's 2 min delay!
and huh? A user had given a moderation of Interesting (+1) to your comment, Good...progressive., attached to QuickTime 7 Released, HD Movie Trailers Available. That moderation has now been undone, probably due to the user posting in the discussion after moderating in it. Your comment is currently scored Normal (1).
I would choose QT over Real any day QT is definitely better overall, although Real gives me an impression of being better over extremely low bandwidth links.
but it might not support H.264 inside.mov This is probably unlikely. The container is supposed to provide locality on the streams, and seeking for really really dumb streams. Basically, the QT driver should simply unpack the stream and just forward it to any codec that it knows how to forward to. In windows, that compatibility is given by directshow, which means if you have a dshow based codec, and a dshow based container player, then you should be able to run the movie. This is highly funny as QT for windows has a dshow output, therefore if QT6 were to output the streams into dshow, the dshow filter (assuming there is one for AVC) should be able to play the video anyway. If this is not the case, then it is probably apple crippling their software again.
MPEG-4 is confusing. I agree. I keep thinking I understand it, and then I realize I do not. I think I will stay with my understanding that it is a "smart" video stream, halfway between a container and a codec.
I've read elsewhere of people using H.264 video + MP3 audio inside AVI containers, which is so incredibly non-standard that it makes me dizzy. I will disagree with you here. AVI is a crappy container, but you can plugin any 1 video stream, and any one audio stream into it, and it will work. AVC video should not be any less standard than the a DIVX5 avi.
Now what I wonder is where is the innovation in containers. I think it was mkv that was supposed to allow "menu" streams. If they did that, they would be the king of formats.
Speaking of which, given that most bootleg anime now comes in MKV and OGM, I am beginning to suspect that these formats are quite technically superior to anything else outthere, that they will take over. Any thoughts?
1) The stupidity (psychology) of the average American citizen (SUV driver).
This is only an argument against local mass transit systems. Absolutely not the case for the intercity trains. Most people hate driving from DC to Boston, and if the trains were a bit faster and did not have a reputation of being slow, all of them would start using the train.
You are lucky. You do not frequently travel from a non-hub city to another non-hub city.
My two city pairs have practically no airports (2 flights a day), are overpriced, and require two layovers, or 1 6 hour layover. In either case, the train is about a quarter of cost, and about 2 hours faster than the fastest you could ever take on the plane.
When both of your cities are not hubs, you begin to realize how crappy and expensive flying can be.
I think the blame in this case goes to poor usage, cheap maintenance and Amtrak tries to keep operating costs low, thus running train at low speeds.
No the main problem lies in the bureaucracy. When the goverment has to vote on how many miles of catenary you can fix this year you have a problem. When you have to fix the tracks so that local railroads can use them and do not pay you for the service, you have a problem. When slow moving freight trains have priority on the tracks, and the government needs to vote to expand the railway from a single line, you have a problem. Poor usage is not the case. Trains in the northeast run at capacity, and during bad weather when airlines cancel, they can pack above capacity.
Also a little bit of advertisement will help. Being a government agency, either they are not allowed to advertize, or they have to seek special funds, as advertizing is definitely not in their budget.
No one's interested in high speed trains because most companies can rent a car, write it off, and drive at well above the speed limit (say, 80 miles/hour or so), and make better time than an Amtrak could...
Not so. If you try to do this from say Philadelphia, to say New Haven, you will quickly notice that drivi g around New York takes a crapload of time, and driving through New York can take even longer.
On many stretches, Amtrak trains, even non high speed, can easily go about 110 to 130 mph, and regularly go about 90 mph.
And on a different note, where exactly is hudson county, and which light rail are you talking about. Are you getting a Path connection, or is it a separate system?
Well, its better than their funding for next year: $0. Actually Amtrak is pretty much being eliminated at this point. Sad, as there are plenty of people working for Amtrak who really care. More details on the National Association of Railroad Passengers page.
second only to Flash. This means that you can create H.264 content and have the reasonable expectation that people will be able to view it.
I will have to disagree there. The number of people viewing avi's and wmv's are much larger than the number of people who are ready to view qt's. In the nom-mac world it is more likely that a person can view realvideo than quicktime. And in that sense, wmv is the second most installed content delivery platform, with realplayer the third, putting quicktime in the fourth place.
It is a shame actually. Quicktime is a much better format than avi. Heck, the only thing worse than avi are separate streams. However, mkv and ogg seem to be at or above the level of quicktime, but obviously they are non-commercial container formats, and thus are not really platforms, but just formats.
Anyway, that was just a flame back.
I am more interested in discussing this line: QuickTime was showing it's age -- QuickTime 6's MPEG-4 implementation was a joke, mostly because of assumptions made with QuickTime 1 that no longer hold true.
So is quicktime7 a new format or not? Your response seems to imply that it is not. But then in that case if they had released these trailers before qt7, the only people who would not be able to play them are the people with Quicktime players, while people with mplayer will have had no problems.
Also I am beginning to doubt my own understanding of MPEG-4. As I understand MPEG-4 specifies a bunch of codecs that can be wrapped in an MPEG-4 header. This basically allows for seeking inside the mpeg-4 track. Then Qt/Ogg/mkv sit on top and handle multiple mpeg-4 streams, and use the mpeg4 information for seeking multiple streams. Is my understanding correct?
Thus did I understand your overall conclusion correctly: Quicktime7 is the same old qt format, but the program now has a better MPEG-4 stream handling plus an AVC handler?
Even after conversion of 1080i into 1080p, the 720p looks better. I notice the fuzz, even after all the blendings and corrections. I am actually disappointed to hear that there are channels broadcasting interlaced HDTV, as I hoped that interlacing will go away as inferior technology.
Well....given that I do not have quicktime7, nor does it seem that my linux libquicktime implementation handles this new modified quicktime format (or perhaps it has streaming problems?), I did not actually see the trailers.
For the rest of the conversation I will assume that Apple changed the qt format, as I never had any problems playing quicktimes before.
However, due to description that the videos are 1280x768 (720p), I would like to thank the people at Apple who did not do the idiotic thing and run this at 1960x1080 (1040i) interlaced, which looks damn awful on most computer screens.
Also I would like to point out to the author of the article that one does not need a 30" cinema screen to see this in all its glory, as even my 10.4" laptop can handle 1280x768, and I have seen a 7.6" screen that handles the same resolution.
Lastly I would like to ask the Mac experts about H.264. It seems that this codec is nothing new, and ffmpeg has supported it for a couple of years now. Why could this not be placed into an older qt version? Or is it just that it was not? Why H.264 is such big news?
I like how you began with a dumbass. Given the fact that you did not even attempt to think about what I said, nor did you present anything interesting to my query, nor did you show why my question is meaningless, it seems that you are not very intelligent yourself. But given this is slashdot, I will cut you some slack, and present my point of view details.
I do agree, the codebases did split, that does not mean that they do not want the changes to go back and forth. KHTML group would like to integrate a bunch of things that have been done by the Apple's fork. They should be able to do so without too much pain. Unfortunately it seems that the changes to the code are not easily mergeable. And although it is not required by GPL, the ability to put the code back is definitely implied by it.
So the reason I asked the question is to hear what the GPL gurus have to say about the conflicting issues of ability to merge changes, and ability to fork. Obviously the fork is absolutely required for the software to stay truly free, but how much obligation should there be for the fork to pass their changes back. Should it depend on how much the split is? What about if the code is moving into extremely different directions?
And for the parent: "APple needed something and KDE wanted something else" is not quite right. Both KDE and Apple want a good html renderer that supports the standards. Given that this is happening just in KHTML, there is very little reason why the codebases are splitting. In fact the only reason is that Apple is not patient enough with KHTML, which is to some extent not that great of the reason, even if it is understandable.
You did not catch the difference between people and people. You are not people. Your mom/gf/whoever did not buy it because she thinks that the system is simple and does not crash, but because someone else said that an iPod is the coolest music player.
The number of people who buy them due to coolness factor seems to be much larger than the people who buy them due to good tech.
Just a little anecdote: I know someone in a business school who bought an iPod, because he needed it to keep his status. He does not even care about music, but he went onto iTunes and bought all the appropriate songs, which he apparently needed to have. And now he walks around with the white earphones in his ears, and the music is not even playing. He just has it there in case someone asks to hear a specific song or something. Go figure.
Someone is going to post under this that this guy is an idiot and is not representative of most people. To you I answer with "duh". My point is not that everyone is like him, but rather that iPods are in fact status symbols in some circles, which have nothing to do with how good they are. iPods occupy the same place as expensive watches: they are really accurate, but that is not why a lot of people buy them.
Actually, you are probably wrong unless you are in japan. Most people do not care which phone you have. Oooh it has a camera, just like three thousand other types of cell phones.
The argument is actually reverse: I believe that people are buying iPods so they can be seen in public with their fancy music player, with OMG-white earphones so everybody sees how cool they are.
(Yeah, iPods are good technology, but when was the last time that people actually bought good technology for just being good technology.)
Since I know you are a layer, but not my lawer, I am asking for an opinion and not advice.
If I read this bill correctly, I may be given 3 year in jail for bringing a digital camera (yay 30 secs of crappy quicktime) into a theater, even if I am not using it, but it is visible?
I would not be surprised that the Business school people probably keep their accounting on personal laptops. Nothing IT can do about that, if they do not have the power to bend all computers in the school to their will, which they should not (bad it policy is worse then no it policy)
The thing that IT is making sure of however is that the passwords are used only via the main kerb. CMU had plenty of problems of people giving passwords to OLR, housing in order to use online services. The current policy is that there must be no site that asks for password, instead site must forward to a ticket granting site. That is good policy. But it still does not help the secretary on a windows laptop with bonzibuddy preloaded.
On andrew, I am quite certain there is a point where your network is shut off first, no warnings given. I have experienced it before there were any set limits for bandwidth, so I do not know what that specific amount is. The stated policy is 1GB. I have been warned twice: when I averaged a little more per day (I had a traffic manager running, averaging about 1.0-1.2 GB a day, running p2p), and the second time (on a totally different registration) I have gone over 2.5 GB that day, and got warned instantly (did not get cut though). Perhaps they no longer cut the network like they used to.
For CS I am not sure, but I know people who use the network extensively, and they said they did not have to do anything special. However, what might have confused me, is that those people are probably on I2. So in other words, I have no clue about the real policy for CS machines is.
The whole issue with Kerb4 is that there were some major problems with Kerb5 playing with AFS or something like that. Andrew AFS is still AFAIK based on Kerb4 tokens.
The bandwidth policies are not policies, but rather good adminning. If you have an andrew machine, and it uses up too much bandwidth, which I believe is 5GB per day, or over 1GB average for 5 days, your pipe is shut off (there may or may not be a warning). This does not happen with CS machines, where andrew assumes by default that those machines have academic use. They will investigate though.
CS quotas are much easier to change than andrew quotas. In andrew to get a bigger quota, you need to start a project, and get faculty to undersign it. Then you need to submit it to a bunch of people. Also among the same lines, it is much easier to contribute software on cs. Sometimes you can just send an email, and the people in CS will install the software on whatever servers you need. (assuming you want some kind of software on public servers). Also the CS has maintained clusters used for educational purposes, which should technically be under the administration of a professor. None of these services are provided by andrew.
Single logon is not just Kerb and AFS being cool. IIRC the system is set up to trust all the different domains. Hence a much greater interoperability. Therefore things like zephyr can work across domains as well. Although, I should not make such a grand claim. Perhaps it is just AFS/Kerb being cool.
CS should fall only to social engineering attacks taking advantage of dorks Heh, I would be more worried about personally managed machines. Although those should not contain too much sensitive info, as most secretarial machines are cs managed.
That may be true if they were the only ones doing that. However that is not the case. All academic departments at CMU have their own networks. IT owns cmu.edu and andrew.cmu.edu, which provide connectivity, cluster services, student AFS space, and generally everything that has to do with undergrads. CS department on the other hand has its own space, and much more lax rules. Many people in CS have root access to their machines, and no bandwidth policies, arbitrary quotas on AFS servers, etc.
All of these are highly integrated, and frequently run on the single kerberos realm provided by IT. (You can log in and read files in CS with your Andrew account, etc)
It would be nice to have a single system, but the number of requests will be highly uneven, and it would be a nightmare to figure out who pays for what. Especially in terms of software. Should IT buy pro-e for the whole school, when only engineering requires it.
And really, this breach has nothing to do with bad network policy. Sure someone broke into an insecure computer, and probably downloaded the access database that was used to store some personal info. This will make the administrator annoyed, but not responsible. And definitely not as angry as when the same file has been lifted off an AFS without knowing someone's password.
Slashdot Mirror
That may be a special exception, as Amtrak can advertise on their own stations, in trains. Perhaps that connection extends to almost all public transit.
Also they may not have had to pay for it, as there are plenty of mutally beneficial relations that amtrak establishes between public transit systems, especially commuter railroads.
I meant to hit preview, and I meant to close the italics. DOH!
And I hate slashdot's 2 min delay!
and huh?
A user had given a moderation of Interesting (+1) to your comment, Good...progressive., attached to QuickTime 7 Released, HD Movie Trailers Available. That moderation has now been undone, probably due to the user posting in the discussion after moderating in it. Your comment is currently scored Normal (1).
I would choose QT over Real any day
.mov
QT is definitely better overall, although Real gives me an impression of being better over extremely low bandwidth links.
but it might not support H.264 inside
This is probably unlikely. The container is supposed to provide locality on the streams, and seeking for really really dumb streams. Basically, the QT driver should simply unpack the stream and just forward it to any codec that it knows how to forward to. In windows, that compatibility is given by directshow, which means if you have a dshow based codec, and a dshow based container player, then you should be able to run the movie. This is highly funny as QT for windows has a dshow output, therefore if QT6 were to output the streams into dshow, the dshow filter (assuming there is one for AVC) should be able to play the video anyway. If this is not the case, then it is probably apple crippling their software again.
MPEG-4 is confusing. I agree. I keep thinking I understand it, and then I realize I do not. I think I will stay with my understanding that it is a "smart" video stream, halfway between a container and a codec.
I've read elsewhere of people using H.264 video + MP3 audio inside AVI containers, which is so incredibly non-standard that it makes me dizzy.
I will disagree with you here. AVI is a crappy container, but you can plugin any 1 video stream, and any one audio stream into it, and it will work. AVC video should not be any less standard than the a DIVX5 avi.
Now what I wonder is where is the innovation in containers. I think it was mkv that was supposed to allow "menu" streams. If they did that, they would be the king of formats.
Speaking of which, given that most bootleg anime now comes in MKV and OGM, I am beginning to suspect that these formats are quite technically superior to anything else outthere, that they will take over. Any thoughts?
1) The stupidity (psychology) of the average American citizen (SUV driver).
This is only an argument against local mass transit systems. Absolutely not the case for the intercity trains. Most people hate driving from DC to Boston, and if the trains were a bit faster and did not have a reputation of being slow, all of them would start using the train.
Only idiots, and large group travellers will take an SUV, since then it becomes cheaper. But even then it is not the case, as amtrak frequently has large group promotions such as this one where you can get 90% off for third, fourth, fifth and sixth persons in the NE corridor (except peak weekend travel times). Too bad most people do not know about these.
You are lucky. You do not frequently travel from a non-hub city to another non-hub city.
My two city pairs have practically no airports (2 flights a day), are overpriced, and require two layovers, or 1 6 hour layover. In either case, the train is about a quarter of cost, and about 2 hours faster than the fastest you could ever take on the plane.
When both of your cities are not hubs, you begin to realize how crappy and expensive flying can be.
This is not the case in entire USA. NYC and its $20 dollar a day parking will make anyone who can be a subway rider into a subway rider.
I think the blame in this case goes to poor usage, cheap maintenance and Amtrak tries to keep operating costs low, thus running train at low speeds.
No the main problem lies in the bureaucracy. When the goverment has to vote on how many miles of catenary you can fix this year you have a problem. When you have to fix the tracks so that local railroads can use them and do not pay you for the service, you have a problem. When slow moving freight trains have priority on the tracks, and the government needs to vote to expand the railway from a single line, you have a problem. Poor usage is not the case. Trains in the northeast run at capacity, and during bad weather when airlines cancel, they can pack above capacity.
Also a little bit of advertisement will help.
Being a government agency, either they are not allowed to advertize, or they have to seek special funds, as advertizing is definitely not in their budget.
No one's interested in high speed trains because most companies can rent a car, write it off, and drive at well above the speed limit (say, 80 miles/hour or so), and make better time than an Amtrak could...
Not so. If you try to do this from say Philadelphia, to say New Haven, you will quickly notice that drivi g around New York takes a crapload of time, and driving through New York can take even longer.
On many stretches, Amtrak trains, even non high speed, can easily go about 110 to 130 mph, and regularly go about 90 mph.
And on a different note, where exactly is hudson county, and which light rail are you talking about. Are you getting a Path connection, or is it a separate system?
Well, its better than their funding for next year: $0. Actually Amtrak is pretty much being eliminated at this point. Sad, as there are plenty of people working for Amtrak who really care. More details on the National Association of Railroad Passengers page.
Yeah. That works without trouble. Apparently libquicktime does not do streams.
second only to Flash. This means that you can create H.264 content and have the reasonable expectation that people will be able to view it.
I will have to disagree there. The number of people viewing avi's and wmv's are much larger than the number of people who are ready to view qt's. In the nom-mac world it is more likely that a person can view realvideo than quicktime. And in that sense, wmv is the second most installed content delivery platform, with realplayer the third, putting quicktime in the fourth place.
It is a shame actually. Quicktime is a much better format than avi. Heck, the only thing worse than avi are separate streams. However, mkv and ogg seem to be at or above the level of quicktime, but obviously they are non-commercial container formats, and thus are not really platforms, but just formats.
Anyway, that was just a flame back.
I am more interested in discussing this line:
QuickTime was showing it's age -- QuickTime 6's MPEG-4 implementation was a joke, mostly because of assumptions made with QuickTime 1 that no longer hold true.
So is quicktime7 a new format or not? Your response seems to imply that it is not. But then in that case if they had released these trailers before qt7, the only people who would not be able to play them are the people with Quicktime players, while people with mplayer will have had no problems.
Also I am beginning to doubt my own understanding of MPEG-4. As I understand MPEG-4 specifies a bunch of codecs that can be wrapped in an MPEG-4 header. This basically allows for seeking inside the mpeg-4 track. Then Qt/Ogg/mkv sit on top and handle multiple mpeg-4 streams, and use the mpeg4 information for seeking multiple streams. Is my understanding correct?
Thus did I understand your overall conclusion correctly: Quicktime7 is the same old qt format, but the program now has a better MPEG-4 stream handling plus an AVC handler?
Even after conversion of 1080i into 1080p, the 720p looks better. I notice the fuzz, even after all the blendings and corrections. I am actually disappointed to hear that there are channels broadcasting interlaced HDTV, as I hoped that interlacing will go away as inferior technology.
Well....given that I do not have quicktime7, nor does it seem that my linux libquicktime implementation handles this new modified quicktime format (or perhaps it has streaming problems?), I did not actually see the trailers.
For the rest of the conversation I will assume that Apple changed the qt format, as I never had any problems playing quicktimes before.
However, due to description that the videos are 1280x768 (720p), I would like to thank the people at Apple who did not do the idiotic thing and run this at 1960x1080 (1040i) interlaced, which looks damn awful on most computer screens.
Also I would like to point out to the author of the article that one does not need a 30" cinema screen to see this in all its glory, as even my 10.4" laptop can handle 1280x768, and I have seen a 7.6" screen that handles the same resolution.
Lastly I would like to ask the Mac experts about H.264. It seems that this codec is nothing new, and ffmpeg has supported it for a couple of years now. Why could this not be placed into an older qt version? Or is it just that it was not? Why H.264 is such big news?
I like how you began with a dumbass. Given the fact that you did not even attempt to think about what I said, nor did you present anything interesting to my query, nor did you show why my question is meaningless, it seems that you are not very intelligent yourself. But given this is slashdot, I will cut you some slack, and present my point of view details.
I do agree, the codebases did split, that does not mean that they do not want the changes to go back and forth. KHTML group would like to integrate a bunch of things that have been done by the Apple's fork. They should be able to do so without too much pain. Unfortunately it seems that the changes to the code are not easily mergeable. And although it is not required by GPL, the ability to put the code back is definitely implied by it.
So the reason I asked the question is to hear what the GPL gurus have to say about the conflicting issues of ability to merge changes, and ability to fork. Obviously the fork is absolutely required for the software to stay truly free, but how much obligation should there be for the fork to pass their changes back. Should it depend on how much the split is? What about if the code is moving into extremely different directions?
And for the parent: "APple needed something and KDE wanted something else" is not quite right. Both KDE and Apple want a good html renderer that supports the standards. Given that this is happening just in KHTML, there is very little reason why the codebases are splitting. In fact the only reason is that Apple is not patient enough with KHTML, which is to some extent not that great of the reason, even if it is understandable.
So is GPL v3 going to fix this issue?
Or will any attempt to do so put too much burden on the users/modifiers of code?
buy none white ones.
:)
Heh. I know a couple of goths who were very unhappy that iPods only came in white. They were very pleased when the black iPod came out.
I think I ruined the day when I mentioned that having an iPod will make them a lot more like everyone else.
You did not catch the difference between people and people. You are not people. Your mom/gf/whoever did not buy it because she thinks that the system is simple and does not crash, but because someone else said that an iPod is the coolest music player.
The number of people who buy them due to coolness factor seems to be much larger than the people who buy them due to good tech.
Just a little anecdote: I know someone in a business school who bought an iPod, because he needed it to keep his status. He does not even care about music, but he went onto iTunes and bought all the appropriate songs, which he apparently needed to have. And now he walks around with the white earphones in his ears, and the music is not even playing. He just has it there in case someone asks to hear a specific song or something. Go figure.
Someone is going to post under this that this guy is an idiot and is not representative of most people. To you I answer with "duh". My point is not that everyone is like him, but rather that iPods are in fact status symbols in some circles, which have nothing to do with how good they are. iPods occupy the same place as expensive watches: they are really accurate, but that is not why a lot of people buy them.
Actually, you are probably wrong unless you are in japan. Most people do not care which phone you have. Oooh it has a camera, just like three thousand other types of cell phones.
The argument is actually reverse: I believe that people are buying iPods so they can be seen in public with their fancy music player, with OMG-white earphones so everybody sees how cool they are.
(Yeah, iPods are good technology, but when was the last time that people actually bought good technology for just being good technology.)
Since I know you are a layer, but not my lawer, I am asking for an opinion and not advice.
If I read this bill correctly, I may be given 3 year in jail for bringing a digital camera (yay 30 secs of crappy quicktime) into a theater, even if I am not using it, but it is visible?
Is that right?
But it would take some pretty spectacular changes to get me to switch.
Same here. Are they planning to release a linux version?
I would not be surprised that the Business school people probably keep their accounting on personal laptops. Nothing IT can do about that, if they do not have the power to bend all computers in the school to their will, which they should not (bad it policy is worse then no it policy)
The thing that IT is making sure of however is that the passwords are used only via the main kerb. CMU had plenty of problems of people giving passwords to OLR, housing in order to use online services. The current policy is that there must be no site that asks for password, instead site must forward to a ticket granting site. That is good policy. But it still does not help the secretary on a windows laptop with bonzibuddy preloaded.
CS or andrew?
On andrew, I am quite certain there is a point where your network is shut off first, no warnings given. I have experienced it before there were any set limits for bandwidth, so I do not know what that specific amount is. The stated policy is 1GB. I have been warned twice: when I averaged a little more per day (I had a traffic manager running, averaging about 1.0-1.2 GB a day, running p2p), and the second time (on a totally different registration) I have gone over 2.5 GB that day, and got warned instantly (did not get cut though). Perhaps they no longer cut the network like they used to.
For CS I am not sure, but I know people who use the network extensively, and they said they did not have to do anything special. However, what might have confused me, is that those people are probably on I2. So in other words, I have no clue about the real policy for CS machines is.
The whole issue with Kerb4 is that there were some major problems with Kerb5 playing with AFS or something like that. Andrew AFS is still AFAIK based on Kerb4 tokens.
The bandwidth policies are not policies, but rather good adminning. If you have an andrew machine, and it uses up too much bandwidth, which I believe is 5GB per day, or over 1GB average for 5 days, your pipe is shut off (there may or may not be a warning). This does not happen with CS machines, where andrew assumes by default that those machines have academic use. They will investigate though.
CS quotas are much easier to change than andrew quotas. In andrew to get a bigger quota, you need to start a project, and get faculty to undersign it. Then you need to submit it to a bunch of people. Also among the same lines, it is much easier to contribute software on cs. Sometimes you can just send an email, and the people in CS will install the software on whatever servers you need. (assuming you want some kind of software on public servers). Also the CS has maintained clusters used for educational purposes, which should technically be under the administration of a professor. None of these services are provided by andrew.
Single logon is not just Kerb and AFS being cool. IIRC the system is set up to trust all the different domains. Hence a much greater interoperability. Therefore things like zephyr can work across domains as well. Although, I should not make such a grand claim. Perhaps it is just AFS/Kerb being cool.
CS should fall only to social engineering attacks taking advantage of dorks
Heh, I would be more worried about personally managed machines. Although those should not contain too much sensitive info, as most secretarial machines are cs managed.
That may be true if they were the only ones doing that. However that is not the case. All academic departments at CMU have their own networks. IT owns cmu.edu and andrew.cmu.edu, which provide connectivity, cluster services, student AFS space, and generally everything that has to do with undergrads. CS department on the other hand has its own space, and much more lax rules. Many people in CS have root access to their machines, and no bandwidth policies, arbitrary quotas on AFS servers, etc.
All of these are highly integrated, and frequently run on the single kerberos realm provided by IT. (You can log in and read files in CS with your Andrew account, etc)
It would be nice to have a single system, but the number of requests will be highly uneven, and it would be a nightmare to figure out who pays for what. Especially in terms of software. Should IT buy pro-e for the whole school, when only engineering requires it.
And really, this breach has nothing to do with bad network policy. Sure someone broke into an insecure computer, and probably downloaded the access database that was used to store some personal info. This will make the administrator annoyed, but not responsible. And definitely not as angry as when the same file has been lifted off an AFS without knowing someone's password.