Well, except 100Mbits networks can support such activity without any problem. You can even feed a decent video stream over 10Mbit. Or 2.5Mbit DSL. Video conferencing could have taken off at any point the last three decades. It hasnt happened. That you cant transmit 'full motion, full screen high definition' Just Aint The Problem. It just isnt interesting or practical for most people. It isnt interesting or practical in low res, nor is it interesting or practical in high definition. And for the JenniCam people of the world they already have it.
Massivly paralell computing clusters would be one field where you could use it. Or for running all the computers as a single system image if someone felt like developing an OS that could handle it. Except I dont really think that many students would care.
Backup servers is the only place where I've found any use for it, but then they have to take multiple 100Mbit streams at a time. And while not Gigabit, Fibre Channel is useful for consolidated storage. But then I dont think many students would like to pay $20 per gigabyte per month in storage costs.
Personally, apart from infrastructure needs, I dont think there are many applications for Gigabit. The bandwidth is there for most things already. It's only when you start shifting the infrastructure from client/server/peer-to-peer to something completely different that there will be a use for it. And the bandwidth wont be the most serious problem in many cases anyway; latency is far worse.
I agree. I rarely burn above 12-16x since the media seems to start going corrupt above that. Not very often, and not very much, but often enough to make any savings on burntime go up in smoke due to re-burn time. One broken mp3 might not really matter, but one broken rpm package is really really annoying.
If there is only windows on the desktop there will only be windows on the servers. And on the PDA's. And on the watches. Without desktop competition there's nothing to stop Microsoft from wiping the other markets from competitors. The one who controls the desktop can make damn sure there isnt a thing in the world which will connect to the server. And how useful is that server when there's nothing to display its data because no desktop will talk to it? Whoever controls the desktop controls the server controls the desktop. It goes both ways and if you control one you can control the other.
And who needs NFS if there arent any clients that use NFS?
"Some of your partners--AOL Time Warner among them -- have recently started selling individual downloads of songs.
They're experimenting with that, as are others. And this has become a much more active conversation now about how to sell single tracks or entire albums online.
But I mean they're selling MP3s. They're selling files that aren't tethered at all for 99 cents a download.
Right, that's a permanent download."
Huh. Anyone know what that's about or have any URL? That sounds like something that would actually be remotely interesting.
The problem is that today it's close to impossible to avoid infringing within some fields. Either you do a search, and you come up with several patents that may or may not be remotely related to your product. So, what do you do then? Call the patent owner and ask 'excuse me does your patent cover what we're doing?'. Of course it does. So then you'll have to get the patent overturned. Or you pay patent fees for something which isnt really covering what you're doing. Or you face willful infringement which is going to hurt a possible case later. Any way you do it you lose if you search. And the thing with these patents is you _cant_ design around the blocking patents because they cover doing something _at all_, not the method for doing it. You _cant_ stay out of trouble, and you _are_ going to infringe for any thing more complicated than 'hello world', depending on how broad patents are allowed to stand in court.
So, the only way you win in this game is if you know nothing and nobody sues you. And if they do call on you and you know nothing you have a better chance of dealing with it then.
Actually, that might have been doable while there were a few thousands of patents. Today it's pretty close to impossible within certain fields since the patents are overly broad and you dont know if they apply without getting an actual judgement in court.
Not to mention, a lot of corporations dont want their employees to do patent research (unless they're filing patents themselves); willful infringment (ie: saw that patent, didnt think it applied) puts you in a much worse position in court than infringing without knowing a patent existed at all.
It is a complete lie of course. The only thing that can happen is that the OEM loses the right to distribute the GPL software they've tried to include in a proprietary piece of intellectual property. Of course, that may make their own intellectual property worthless if it's dependent upon the GPL software.
Kindof like if they got hold of MS sourcecode and used that code in a program they made and got caught at it. I sortof doubt that MS gives them right to distribute Microsoft code just because their product depends on it...
You mean; The world economy is faked with American money (that may or may not exist) and American companies (who may, or may not, have profits). Foreign companies that want to be more than a tiny, localized entity need to take a class in shredding and cooking their books to do buisness with Americans and American companies... etc.
In case you hadnt noticed, with the way the American economy is currently heading, foreign customers may turn out to be quite important to Microsoft. You know, the kind of customers whose checks dont bounce.
Lets see... who you gonna call when Open Source stuff breaks? IBM? HP/Compaq? Sun? Redhat? SuSE? Caldera? One of the thousands of consultancy companies that would love to sell you time? And, you know, if one of them wont solve it, you can call another one.
So... who you gonna call when Microsoft stuff breaks and Microsoft refuses to fix it? The DOJ?
This is a fairly common mistake that annoys me a bit. There are a lot of linux/unix people who pop up an xterm and show people how to do things. Yes, that's the easiest way to do it, it's the fastest way to do it, vi and bash are rarely broken and you can back out of the changes because you know what changes you made. That's how I do it. That's probably how you do it. And for anyone doing it every day on dozens of machines it's the most natural thing in the world.
But that's _not_ how to show newbies how to do it.
Never, _ever_ pop up an xterm while demoing. Use the GUI tools. These days they can do pretty much everything a newbie needs to do and they're not intimidating in the same way.
In my experience the 'support' factor goes down when installing Linux for them. If she's used to Word, Excel, email and web browsing she'll do just fine with a modern Linux.
Set it up, tell her not to login as root, fix her up with a good desktop config for what she wants to do, and an easy way to restore the config to default, and she just cant mess it up (which is a factor that makes 'mom' types more comfortable with their computers, in my experience).
You'll probably get a few calls after powerdowns about fsck and such, but ext3 usually handles it without any manual recovery (and you could probably hack the initscripts to blaze through an fsck -y anyway).
For a novice desktop user there is no steep learning curve. It's not until you get to 'power-windows-user' that you get a more serious learning curve.
Have you considered Samsung Contact (formerly HP Openmail)? As far as Exchange replacements it should be a viable alternative. Runs on Solaris, Linux, HP-UX or AIX on the server side and supports pretty much everything Exchange does on the client side (and of course it supports most other email clients).
Of course, if you dont need a feature for feature match with Exchange there are unlimited cheap alternatives for mail servers.
News flash: You hit right on without knowing it. In your comparison, threading equals multitasking without memory protection. It Just Doesnt Work Very Well.
The overhead of a fork on an OS that does copy on write for the forking is minimal. And it outweighs the cost of dealing with threads by far. Fork costs more, but mutexes cost as well, usually to the extent that you lose the advantage of having multiple concurrent (or not so concurrent, after all) threads of execution.
Actually, I think the only one able to possibly do a worse job at Star Wars would be Spielberg. He'd insist on replacing Jar-Jar and the bots with Cute Kids (tm) Saving the Situation (tm), which IMO is the only conceivable way it could get worse. Gaaaack.
Well, to be fair, Spielberg did a good job on Minority Report, which was among the absolutely best films that he's ever done. And he's done a few other fairly good ones lately. But I can think of several directors who could do a better job than either Spielberg or Lucas.
I wouldnt say that gprof is useless... threading, however, comes very close to it.
Threading is useful in the instance where you have an application that needs to scale with SMP and which you cannot, for whatever reason, fork. But the accompanying pain of being forced to pay extremely close attention and mutex lock the code all over makes it not worth it for most situations.
Use fork. Use other IPC methods if necessary. But dont thread or you'll spend an order of magnitude more time debugging.
True. However, it isnt really used that often, and I'd say the OpenBSD folks have ignored their own best practices of turning off all unnecessary features in this case.
The reason people think this was handled badly is because most people arent affected. As far as I can tell, pretty much everyone who isnt running xBSD hasnt been affected, and a workaround by changing one line in the config is _far_ easier than upgrading to a sortof broken version, followed by yet another upgrade to a less broken version.
The reason law generally doesnt allow traps to be set is pretty simple. Imagine Mr Razor Stereo forgets to remove the razors when leaving the car for service, or imagine he gets stopped by Customs when going to Mexico. Ow. Imagine Mr. Exploding Wallet drops his wallet, and a friendly person picks it up, intending to return it to Lost Effects at the police, or something. Boom.
Traps have a tendency not to be picky, so while allowing them could deter a few crimes, I think the average trap would be more likely to catch someone without criminal intent.
You deserve? You _deserve_? Well, there are some people who maybe _deserved_ to live. They dont, however, because neither they nor their governments could afford the medicines due to the patents.
You _deserve_ nothing. Patents are a fiction that's been made up to promote the development of science for the benefit of society. The benefit to society is the basic justification of patents, and that benefit is by now very doubtful. Perhaps society would be better served by removing patent protection entirely and funding development in alternative ways. The inability of the medical industry to handle the ethical burdens of the issues may eventually make that necessary.
In practice I agree that specific methods should be patentable. Your specific method to clone a tumor suppressor gene should be patentable but anyone should likewise be able to devise other, cheaper, methods to clone the same gene. This is how traditional patents have worked. You can patent the specific method to do something, which is far less damaging than being able to patent what amounts to the actual goal. And only, _only_, if these are non-obvious methods to a person with experience in the field. That is, if you're racing against someone else to develop a specific method first, forget it. It's _obviously_ not inventive and 'deserving' of patent protection, or there wouldnt _be_ a race.
Actually, in such a case, Amgen would be forced to sue the individual patients for allowing their cells to violate Amgens patents. In fact, they should probably sue everyone else whose body produces the infringing proteins and ask law enforcement agencies to forcibly stop this violation.
Wouldnt it make even more sense to not grant patents on genes in the first place, since rather than promoting progress they appear to throw up roadblocks that have to be worked around?
Actually, I find fewer and fewer sites that dont quite work with Mozilla based browsers, in fact, by now it's down to about ordinary breakage around the web. I doubt there'd be a deluge; considering most sites will work without a hitch, it's more likely the complaints will go to the site in question.
The advantages of Mozilla are also things that the ordinary user benefits from. Popup blocking, especially, makes the web a more pleasant experience.
Yeah, well, it can still 'break' massively. Try having nfs mounted home directories and poweroff a workstation without a shutdown... zzzzap, completely whacked gnome-conf. GConfd wont even run until you clear out its directory, and no error messages, and naturally no apps relying on gnome-conf will start without it.
Slashdot Mirror
Well, except 100Mbits networks can support such activity without any problem. You can even feed a decent video stream over 10Mbit. Or 2.5Mbit DSL. Video conferencing could have taken off at any point the last three decades. It hasnt happened. That you cant transmit 'full motion, full screen high definition' Just Aint The Problem. It just isnt interesting or practical for most people. It isnt interesting or practical in low res, nor is it interesting or practical in high definition. And for the JenniCam people of the world they already have it.
Massivly paralell computing clusters would be one field where you could use it. Or for running all the computers as a single system image if someone felt like developing an OS that could handle it. Except I dont really think that many students would care.
Backup servers is the only place where I've found any use for it, but then they have to take multiple 100Mbit streams at a time. And while not Gigabit, Fibre Channel is useful for consolidated storage. But then I dont think many students would like to pay $20 per gigabyte per month in storage costs.
Personally, apart from infrastructure needs, I dont think there are many applications for Gigabit. The bandwidth is there for most things already. It's only when you start shifting the infrastructure from client/server/peer-to-peer to something completely different that there will be a use for it. And the bandwidth wont be the most serious problem in many cases anyway; latency is far worse.
I agree. I rarely burn above 12-16x since the media seems to start going corrupt above that. Not very often, and not very much, but often enough to make any savings on burntime go up in smoke due to re-burn time. One broken mp3 might not really matter, but one broken rpm package is really really annoying.
If there is only windows on the desktop there will only be windows on the servers. And on the PDA's. And on the watches. Without desktop competition there's nothing to stop Microsoft from wiping the other markets from competitors. The one who controls the desktop can make damn sure there isnt a thing in the world which will connect to the server. And how useful is that server when there's nothing to display its data because no desktop will talk to it? Whoever controls the desktop controls the server controls the desktop. It goes both ways and if you control one you can control the other.
And who needs NFS if there arent any clients that use NFS?
They're experimenting with that, as are others. And this has become a much more active conversation now about how to sell single tracks or entire albums online.
But I mean they're selling MP3s. They're selling files that aren't tethered at all for 99 cents a download.
Right, that's a permanent download."
Huh. Anyone know what that's about or have any URL? That sounds like something that would actually be remotely interesting.
The problem is that today it's close to impossible to avoid infringing within some fields. Either you do a search, and you come up with several patents that may or may not be remotely related to your product. So, what do you do then? Call the patent owner and ask 'excuse me does your patent cover what we're doing?'. Of course it does. So then you'll have to get the patent overturned. Or you pay patent fees for something which isnt really covering what you're doing. Or you face willful infringement which is going to hurt a possible case later. Any way you do it you lose if you search. And the thing with these patents is you _cant_ design around the blocking patents because they cover doing something _at all_, not the method for doing it. You _cant_ stay out of trouble, and you _are_ going to infringe for any
thing more complicated than 'hello world', depending on how broad patents are allowed to stand in court.
So, the only way you win in this game is if you know nothing and nobody sues you. And if they do call on you and you know nothing you have a better chance of dealing with it then.
Actually, that might have been doable while there were a few thousands of patents. Today it's pretty close to impossible within certain fields since the patents are overly broad and you dont know if they apply without getting an actual judgement in court.
Not to mention, a lot of corporations dont want their employees to do patent research (unless they're filing patents themselves); willful infringment (ie: saw that patent, didnt think it applied) puts you in a much worse position in court than infringing without knowing a patent existed at all.
It is a complete lie of course. The only thing that can happen is that the OEM loses the right to distribute the GPL software they've tried to include in a proprietary piece of intellectual property. Of course, that may make their own intellectual property worthless if it's dependent upon the GPL software.
Kindof like if they got hold of MS sourcecode and used that code in a program they made and got caught at it. I sortof doubt that MS gives them right to distribute Microsoft code just because their product depends on it...
You mean; The world economy is faked with American money (that may or may not exist) and American companies (who may, or may not, have profits). Foreign companies that want to be more than a tiny, localized entity need to take a class in shredding and cooking their books to do buisness with Americans and American companies... etc.
In case you hadnt noticed, with the way the American economy is currently heading, foreign customers may turn out to be quite important to Microsoft. You know, the kind of customers whose checks dont bounce.
Lets see... who you gonna call when Open Source stuff breaks? IBM? HP/Compaq? Sun? Redhat? SuSE? Caldera? One of the thousands of consultancy companies that would love to sell you time? And, you know, if one of them wont solve it, you can call another one.
So... who you gonna call when Microsoft stuff breaks and Microsoft refuses to fix it? The DOJ?
And how hard is that? Fire up red-carpet, click subscribe on the channel you want, mark the software she wants and click install...
Or download it and double-click on the rpm (your dist should have a gui rpm handler installed).
It's not harder than windows. Unless you make it so.
This is a fairly common mistake that annoys me a bit. There are a lot of linux/unix people who pop up an xterm and show people how to do things. Yes, that's the easiest way to do it, it's the fastest way to do it, vi and bash are rarely broken and you can back out of the changes because you know what changes you made. That's how I do it. That's probably how you do it. And for anyone doing it every day on dozens of machines it's the most natural thing in the world.
But that's _not_ how to show newbies how to do it.
Never, _ever_ pop up an xterm while demoing. Use the GUI tools. These days they can do pretty much everything a newbie needs to do and they're not intimidating in the same way.
In my experience the 'support' factor goes down when installing Linux for them. If she's used to Word, Excel, email and web browsing she'll do just fine with a modern Linux.
Set it up, tell her not to login as root, fix her up with a good desktop config for what she wants to do, and an easy way to restore the config to default, and she just cant mess it up (which is a factor that makes 'mom' types more comfortable with their computers, in my experience).
You'll probably get a few calls after powerdowns about fsck and such, but ext3 usually handles it without any manual recovery (and you could probably hack the initscripts to blaze through an fsck -y anyway).
For a novice desktop user there is no steep learning curve. It's not until you get to 'power-windows-user' that you get a more serious learning curve.
Have you considered Samsung Contact (formerly HP Openmail)? As far as Exchange replacements it should be a viable alternative. Runs on Solaris, Linux, HP-UX or AIX on the server side and supports pretty much everything Exchange does on the client side (and of course it supports most other email clients).
Of course, if you dont need a feature for feature match with Exchange there are unlimited cheap alternatives for mail servers.
News flash: You hit right on without knowing it. In your comparison, threading equals multitasking without memory protection. It Just Doesnt Work Very Well.
The overhead of a fork on an OS that does copy on write for the forking is minimal. And it outweighs the cost of dealing with threads by far. Fork costs more, but mutexes cost as well, usually to the extent that you lose the advantage of having multiple concurrent (or not so concurrent, after all) threads of execution.
Actually, I think the only one able to possibly do a worse job at Star Wars would be Spielberg. He'd insist on replacing Jar-Jar and the bots with Cute Kids (tm) Saving the Situation (tm), which IMO is the only conceivable way it could get worse. Gaaaack.
Well, to be fair, Spielberg did a good job on Minority Report, which was among the absolutely best films that he's ever done. And he's done a few other fairly good ones lately. But I can think of several directors who could do a better job than either Spielberg or Lucas.
"Some problems are conceptually parellel; it almost always easist to write a procedure in a way that mirrors the way it's conceptualized."
In that case... fork and use IPC. It's not substantially more expensive and you wont have to ensure your parallel code is thread safe.
I wouldnt say that gprof is useless... threading, however, comes very close to it.
Threading is useful in the instance where you have an application that needs to scale with SMP and which you cannot, for whatever reason, fork. But the accompanying pain of being forced to pay extremely close attention and mutex lock the code all over makes it not worth it for most situations.
Use fork. Use other IPC methods if necessary. But dont thread or you'll spend an order of magnitude more time debugging.
True. However, it isnt really used that often, and I'd say the OpenBSD folks have ignored their own best practices of turning off all unnecessary features in this case.
The reason people think this was handled badly is because most people arent affected. As far as I can tell, pretty much everyone who isnt running xBSD hasnt been affected, and a workaround by changing one line in the config is _far_ easier than upgrading to a sortof broken version, followed by yet another upgrade to a less broken version.
The reason law generally doesnt allow traps to be set is pretty simple. Imagine Mr Razor Stereo forgets to remove the razors when leaving the car for service, or imagine he gets stopped by Customs when going to Mexico. Ow. Imagine Mr. Exploding Wallet drops his wallet, and a friendly person picks it up, intending to return it to Lost Effects at the police, or something. Boom.
Traps have a tendency not to be picky, so while allowing them could deter a few crimes, I think the average trap would be more likely to catch someone without criminal intent.
You deserve? You _deserve_? Well, there are some people who maybe _deserved_ to live. They dont, however, because neither they nor their governments could afford the medicines due to the patents.
You _deserve_ nothing. Patents are a fiction that's been made up to promote the development of science for the benefit of society. The benefit to society is the basic justification of patents, and that benefit is by now very doubtful. Perhaps society would be better served by removing patent protection entirely and funding development in alternative ways. The inability of the medical industry to handle the ethical burdens of the issues may eventually make that necessary.
In practice I agree that specific methods should be patentable. Your specific method to clone a tumor suppressor gene should be patentable but anyone should likewise be able to devise other, cheaper, methods to clone the same gene. This is how traditional patents have worked. You can patent the specific method to do something, which is far less damaging than being able to patent what amounts to the actual goal. And only, _only_, if these are non-obvious methods to a person with experience in the field. That is, if you're racing against someone else to develop a specific method first, forget it. It's _obviously_ not inventive and 'deserving' of patent protection, or there wouldnt _be_ a race.
Actually, in such a case, Amgen would be forced to sue the individual patients for allowing their cells to violate Amgens patents. In fact, they should probably sue everyone else whose body produces the infringing proteins and ask law enforcement agencies to forcibly stop this violation.
Apparently not. Or they would probably do that.
Wouldnt it make even more sense to not grant patents on genes in the first place, since rather than promoting progress they appear to throw up roadblocks that have to be worked around?
Actually, I find fewer and fewer sites that dont quite work with Mozilla based browsers, in fact, by now it's down to about ordinary breakage around the web. I doubt there'd be a deluge; considering most sites will work without a hitch, it's more likely the complaints will go to the site in question.
The advantages of Mozilla are also things that the ordinary user benefits from. Popup blocking, especially, makes the web a more pleasant experience.
Yeah, well, it can still 'break' massively. Try having nfs mounted home directories and poweroff a workstation without a shutdown... zzzzap, completely whacked gnome-conf. GConfd wont even run until you clear out its directory, and no error messages, and naturally no apps relying on gnome-conf will start without it.