By any stretch of the imagination, these aren't dumb terminals they're talking about. Hell, the vt220 I have downstairs is a smart terminal by most definitions, and the vt520 definitely is. Modern thin clients are smart, fast, and capable of local graphics processing. What they're not capable of is storage, resident OS, or local compute cycles.
The good folks at the WST (and anyone else) should try setting TERM=dumb in a Unix shell session, and find out just how dumb a dumb terminal is.
First of all, you're missing the other costs I mentioned. That quarter million in gear costs probably comes out to 0.5-1.0MM/yr when everything else is added in. Now when you consider that 87% of the email we receive (by quantity, not bandwidth, which would be slightly higher) is spam, you have to ask how much we'd be spending on email per year if we didn't have spam to deal with.
In other words, we're not spending money protecting an asset, we're WASTING at least HALF A MILLION DOLLARS PER YEAR, by allowing (and paying for!) criminal activity. I don't care if you're a startup or Microsoft or GE--wasting half a mil on anything is considered bad business.
Two very good points. Ignoring my own personal opinion, you're entirely correct--we don't need to penalise
every
spammer out there, just a significant percentage of them, right across the board. Nailing the top three every four years isn't going to do any damage, but nailing 80 of the top 200 every year without fail would be a better deterrent. Similarly, your criminal vs. civil prosecution argument supports the same point, except for one point: It's hard to collect on a civil lawsuit, particularly if the defendant is in Florida and has put all of his money into his house. Once found guilty and tossed in a cold cell, it's easier to make a dirtbag stay in that cell.
You suggest that we could probably get free software that would do the same as our commercial stuff. I'm not a huge fan of the "commercial software on MS platform" model, but I have to say that there is nothing in the OSS world that does the same as our email filters. Regardless, the software is (relatively) cheap--I think it's about $20k of the total solution.
Your 'brick and paint' argument makes no sense. Graffiti doesn't consume bricks. It does consume paint of course, and probably does affect the price of paint. The difference is that as the brick-building owner, it's SOME paint that they're vandalising my property with. In the case of spam, it's MY bandwidth they're vandalising my network with.
In our case, 87% of the inbound mail to the company is spam. Eighty-FUCKING-seven percent. That's not bandwidth either--given the prevalence of attachments (jpgs mostly) in spam these days, the actual bandwidth consumed by inbound spam as a percentage of the total inbound mail bandwidth is slightly higher--about 89-91% (varies month to month).
So yes, our total bandwidth cost is heavily influenced by spam. Not cost per megabyte per second, but total dollars spent is significantly higher than it would be if it weren't for criminal fuckers who deserve to be taken out and flogged in public.
You don't want him out, you want him in. Put him in jail. Put him behind bars. Put him in the same conditions he's forced upon the Guantanamo Bay detainees (without benefit of due process, open trials, etc.)
Don't let him off--hold the SOB accountable for his crimes.
Seriously, I usually laugh my ass off when I watch Brazil, and I've watched it a lot.
It's a black comedy, a parody of the present, that never takes itself as seriously as you think it does. It has some really brilliant performances by some of the finest actors around. However, that's not why I laugh.
I laugh because it's true. It's so very close to the truth, and for two decades now, I've watched the gap between the movie and reality shrink. If I didn't laugh, I'd go completely insane.
Truly one of the finest movies of the 20th century. Or at least one of the most twisted.
The fundamental aim of lawsuits is to make spam unprofitable. This is unfortunately not going to work unless EVERY spammer gets sued in a consistent manner, and loses consistently. As long as lawsuits are only sparsely executed against the biggest targets, there's going to be a far too tempting amount of 'under-the-radar' space.
Throwing these scumbags in jail for decades at a time for fraud and theft and vandalism is the only real solution
Re:Unfortunatly it is the only way to go.
on
MySpace Sues Spam King
·
· Score: 2, Insightful
This was an old and flawed argument in 1993. AOL pays the postal services to deliver their crap. Spammers make the recipient pay for their crap. If it weren't for AOL and junk mail in general, you would pay MORE for your postal service. If it weren't for Scott Richter and the like, you would pay LESS for your internet service. (and it would be significantly better as well.)
One of my customers is a moderately large company, with about 4000 employees. They have spent about a quarter million dollars on anti-spam software and hardware, which doesn't include the cost of bandwidth, storage, man-hours, or lost productivity.
They are working quite hard at it. It was one of the major drives behind the OpenSolaris project, in fact. On the one hand, running Solaris10 on commodity hardware is good for the workstation world, where Solaris on Sparc has been falling behind as of late. On the other hand, getting Solaris into the hands of as many hobbyists as possible (through OpenSolaris, due to all the multimedia tools) is the best thing they can do.
Solaris is the dominant OS in the oil company datacentres of the world. Windows is the dominant desktop. Linux is making inroads on the desktop, and is a complete bit-player on the server side, in this industry. In commerce, AIX is still dominant, and Linux is unheard of. Telecom companies, admittedly, are getting more friendly with Linux.
Solaris is not only alive, but will remain that way for a while.
Actually, I chose those two examples quite carefully. The nature of gravity at a very low level is still pretty vague--and yet the results of gravity are something we live with every second of our lives. The fact that it's a theory doesn't preclude something from holding us onto the earth.
Light, on the other hand, has a very well understood model--and yet it's only a model. As you said, "we can say with some certainty that visible light is electromagnetic radiation and the frequency is 10 to the 14 through 16 with a wavelength between 400 to 700 nm."
That's a theoretical model. It's a model that fits our observations, but even EMR is only a theory, not a proof.
I'll assume you've missed all of the Solaris 10 hype, and are genuinely curious. That said, there are a lot of interesting things in Solaris 10.
First of all, it is robust and reliable to a degree that Linux still doesn't achieve in a general-purpose environment. It's also immensely scaleable--dealing gracefully with as big of a machine as you want to throw at it. In terms of technology, Solaris 10 was a complete rewrite, and in many ways was a rethinking of Unix. It provides service-level fault tolerance (via SMF, which replaces the traditional/etc/init.d method of starting services). There's dtrace which can trace anything in the computer (massively, incredibly more powerful than strace or truss). Zones are an implementation of virtual machines, and allow for complete isolation of environments all under one kernel. Related to that is the scheduler, which allows a very granular means of resource allocation to a process or application. Also, Brandz will let you run Linux code under Solaris, within a zone. I know of developers who are using this, because it lets them run dtrace against their Linux code for debugging and optimisation. Finally there's ZFS, which is truly a new filesystem--the first in a long time on any platform. It combines filesystem operations with volume management, and results in a filesystem that has been abstracted from the hardware it's running on.
These are just the highlights of the most robust Unix out there right now.
What Solaris 10 will NOT buy you though, is the same end-user experience of Linux. The graphics routines, multimedia applications, and audio support just aren't at the same level in Solaris yet. That's changing fast enough, but it hasn't caught up yet.
So IT has an interesting difference from most other parts of a company.
An ordinary user can't take down a phone system by accidentally hitting three buttons on their phone. They can't take down the power to the building by turning on the wrong power switches in their office. However, they CAN potentially take down a significant part of computer infrastructure by opening the wrong attachment, and thanks to the virus and spyware-infected world of computing, that's relatively likely.
With this sort of power one would normally expect an equal amount of knowledge and responsibility, but it's not really their fault that they've had that much power thrust into their hands. Also, they have their own jobs which (generally) require their time and attention, and they shouldn't have to become an IT expert to do their own work.
Furthermore, there are some truly unrealistic demands placed on IT, probably moreso than most departments. People who phone and say, "my computer is broken--fix it RIGHT NOW!!!" may be completely unable to do their work (which is bad), but are unwilling to look at a larger picture of, 'I can fix your computer, or I can fix all outbound mail being sent to our largest partner.'
Bottom line is that computers are both essential to business, and also the most unreliable and time-intensive part of modern corporate infrastructure.
However, none of that makes any difference in the end. If you're employed to support people, then you'd damned well better support them professionally. Keep your bitching to jokes within the team and keep them to a minimum. Treat people with all the respect in the world and always work to the best of your abilities for them, even if they don't deserve it. That's what professionalism is all about.
You seem to misunderstand what theory means in the context of scientific rigour.
Evolution is a theory. Global warming is a theory. Gravity is a theory too! So is light.
"Respected members of the scientific community" stopped making such statements as soon as contrary 'evidence' was examined. It's becoming very clear that global warming is happening--that doesn't make it less of a theory, by the way. The relative causes are still getting sorted out, but anyone who actually takes the time to study and understand the evidence can't avoid the obvious and clear conclusion.
If you're so much smarter than them, then why do you think the parent is a woman, when the article clearly states otherwise?
Re:From vi to vim, now back again to vi
on
The Birth of vi
·
· Score: 1
I would argue against that claim.
When I say "strict vi compatability," I mean that I shouldn't know that I'm using anything other than plain ol' vanilla vi. vim's compatability mode is...weak. Lots of things that might be closer, but still aren't vi.
Re:From vi to vim, now back again to vi
on
The Birth of vi
·
· Score: 1
Step away from Linux, and you'll discover that it's the oddity in the crowd.
Solaris: vi HP-UX: vi AIX: vi IRIX: vi *BSD: nvi (not exactly vi, but much closer than vim)
Why vim took over the Linux world baffles me. It always takes me a week on a new Linux machine to figure out how to turn off all of the unwelcome 'on by default' features. If vim is invoked as vi, it should revert to strict vi compatibility mode by default.
Well unfortunately, Coke really does deserve to be in the bad corner.
Do some digging on Coke in India (especially Kerala), and find out just how much damage they've done. Fertile ground has been turned into parched earth, groundwater has been contaminated across the country, and there's some question about the incidence of birth defects near the bottling plants. Unfortunately, it's far enough away from the Western World(tm), and also India's special interest groups tend to stray farther from the truth than they would here, so it's not the debacle it should be.
As an aside, Pepsi is right up there in bad behaviour as well. Both companies were temporarily banned from production in Kerala, in August of 2006, but the ban was overturned.
No no, I'm not claiming that technical superiority will lead to market share. That's something that is determined primarily by marketing (sadly).
However, what I'm saying is that Solaris holds enough of an existing marketshare and has enough interesting stuff to capture a significant amount of the hobbyist interest. That doesn't often lead to real marketshare, but is another world unto itself. Furthermore, it's one that is primarily (or at least significantly) driven by technology (and thanks to Richard Stallman, licensing minutae).
For the most part, Linux and Solaris still don't compete. They're isolated by role, without a huge amount of crossover (with the exception of technical workstations, where Linux is kicking Solaris around the block).
If you read the article, you'd find that this is the first time. Clinton enacted a law, and Bush (!) has enforced it. From here on in, it will happen every year, but this is the first.
I agreed with everything you mentioned, except this:
"This literary device has been completely abandoned since the '80s, I think, and nowadays the author tends to throw the reader "in media res" and add details as the story moves along."
I think you need to expand your reading sphere a bit. Many authors still use this technique, in fact most of the ones I read. Maybe avoiding it is something that's limited to post-Gibson SF authors.
Some examples of currently writing authors who use this technique: Murakami, Gaiman, Marquez. They may not be quite as dull and plodding about it as the lectures of the '50s, but to be fair, the _good_ authors of the '50s weren't dull and plodding about it either.
Well, I've already replied to one of your posts tonight--why not make a pair?
"OpenSolaris hasn't been big enough to experience fragmentation yet, so you can't judge how well OpenSolaris will hold up at scale of development community..."
This is entirely accurate. However, I'm inclined to believe that fragmentation will be less of a factor for OpenSolaris, due to the fact that it ultimately feeds back into a single definitive snapshot. Solaris will always be the watermark for OpenSolaris, so as new projects get brought in, they'll be (1) incorporated into Solaris and become part of the canon, (2) left out, and gradually fade away, or (3) be important enough to cause a fork. The key is that option (1) doesn't exist per se for Linux, so option (3) gets invoked far more often in that realm. There will be some fragmentation, yes, but OpenSolaris is constantly being refreshed from Solaris (and also feeding back into it), so the potential for fragmentation, in my opinion, is far less.
Of course, it's early. We'll find out in two years if I'm right.
Slashdot Mirror
It's always a tough decision between those two options.
In this case, I think it's crack.
By any stretch of the imagination, these aren't dumb terminals they're talking about. Hell, the vt220 I have downstairs is a smart terminal by most definitions, and the vt520 definitely is. Modern thin clients are smart, fast, and capable of local graphics processing. What they're not capable of is storage, resident OS, or local compute cycles.
The good folks at the WST (and anyone else) should try setting TERM=dumb in a Unix shell session, and find out just how dumb a dumb terminal is.
First of all, you're missing the other costs I mentioned. That quarter million in gear costs probably comes out to 0.5-1.0MM/yr when everything else is added in. Now when you consider that 87% of the email we receive (by quantity, not bandwidth, which would be slightly higher) is spam, you have to ask how much we'd be spending on email per year if we didn't have spam to deal with.
In other words, we're not spending money protecting an asset, we're WASTING at least HALF A MILLION DOLLARS PER YEAR, by allowing (and paying for!) criminal activity. I don't care if you're a startup or Microsoft or GE--wasting half a mil on anything is considered bad business.
- every
spammer out there, just a significant percentage of them, right across the board. Nailing the top three every four years isn't going to do any damage, but nailing 80 of the top 200 every year without fail would be a better deterrent. Similarly, your criminal vs. civil prosecution argument supports the same point, except for one point: It's hard to collect on a civil lawsuit, particularly if the defendant is in Florida and has put all of his money into his house. Once found guilty and tossed in a cold cell, it's easier to make a dirtbag stay in that cell.You suggest that we could probably get free software that would do the same as our commercial stuff. I'm not a huge fan of the "commercial software on MS platform" model, but I have to say that there is nothing in the OSS world that does the same as our email filters. Regardless, the software is (relatively) cheap--I think it's about $20k of the total solution.
Your 'brick and paint' argument makes no sense. Graffiti doesn't consume bricks. It does consume paint of course, and probably does affect the price of paint. The difference is that as the brick-building owner, it's SOME paint that they're vandalising my property with. In the case of spam, it's MY bandwidth they're vandalising my network with.
In our case, 87% of the inbound mail to the company is spam. Eighty-FUCKING-seven percent. That's not bandwidth either--given the prevalence of attachments (jpgs mostly) in spam these days, the actual bandwidth consumed by inbound spam as a percentage of the total inbound mail bandwidth is slightly higher--about 89-91% (varies month to month).
So yes, our total bandwidth cost is heavily influenced by spam. Not cost per megabyte per second, but total dollars spent is significantly higher than it would be if it weren't for criminal fuckers who deserve to be taken out and flogged in public.
You don't want him out, you want him in. Put him in jail. Put him behind bars. Put him in the same conditions he's forced upon the Guantanamo Bay detainees (without benefit of due process, open trials, etc.)
Don't let him off--hold the SOB accountable for his crimes.
Seriously, I usually laugh my ass off when I watch Brazil, and I've watched it a lot.
It's a black comedy, a parody of the present, that never takes itself as seriously as you think it does. It has some really brilliant performances by some of the finest actors around. However, that's not why I laugh.
I laugh because it's true. It's so very close to the truth, and for two decades now, I've watched the gap between the movie and reality shrink. If I didn't laugh, I'd go completely insane.
Truly one of the finest movies of the 20th century. Or at least one of the most twisted.
The fundamental aim of lawsuits is to make spam unprofitable. This is unfortunately not going to work unless EVERY spammer gets sued in a consistent manner, and loses consistently. As long as lawsuits are only sparsely executed against the biggest targets, there's going to be a far too tempting amount of 'under-the-radar' space.
Throwing these scumbags in jail for decades at a time for fraud and theft and vandalism is the only real solution
This was an old and flawed argument in 1993.
AOL pays the postal services to deliver their crap. Spammers make the recipient pay for their crap.
If it weren't for AOL and junk mail in general, you would pay MORE for your postal service. If it weren't for Scott Richter and the like, you would pay LESS for your internet service. (and it would be significantly better as well.)
One of my customers is a moderately large company, with about 4000 employees. They have spent about a quarter million dollars on anti-spam software and hardware, which doesn't include the cost of bandwidth, storage, man-hours, or lost productivity.
They are working quite hard at it. It was one of the major drives behind the OpenSolaris project, in fact. On the one hand, running Solaris10 on commodity hardware is good for the workstation world, where Solaris on Sparc has been falling behind as of late. On the other hand, getting Solaris into the hands of as many hobbyists as possible (through OpenSolaris, due to all the multimedia tools) is the best thing they can do.
So yes, they're workin on it.
You must live in a fascinating world.
Solaris is the dominant OS in the oil company datacentres of the world. Windows is the dominant desktop. Linux is making inroads on the desktop, and is a complete bit-player on the server side, in this industry. In commerce, AIX is still dominant, and Linux is unheard of. Telecom companies, admittedly, are getting more friendly with Linux.
Solaris is not only alive, but will remain that way for a while.
Actually, I chose those two examples quite carefully. The nature of gravity at a very low level is still pretty vague--and yet the results of gravity are something we live with every second of our lives. The fact that it's a theory doesn't preclude something from holding us onto the earth.
Light, on the other hand, has a very well understood model--and yet it's only a model. As you said, "we can say with some certainty that visible light is electromagnetic radiation and the frequency is 10 to the 14 through 16 with a wavelength between 400 to 700 nm."
That's a theoretical model. It's a model that fits our observations, but even EMR is only a theory, not a proof.
I'll assume you've missed all of the Solaris 10 hype, and are genuinely curious. That said, there are a lot of interesting things in Solaris 10.
/etc/init.d method of starting services). There's dtrace which can trace anything in the computer (massively, incredibly more powerful than strace or truss). Zones are an implementation of virtual machines, and allow for complete isolation of environments all under one kernel. Related to that is the scheduler, which allows a very granular means of resource allocation to a process or application. Also, Brandz will let you run Linux code under Solaris, within a zone. I know of developers who are using this, because it lets them run dtrace against their Linux code for debugging and optimisation.
First of all, it is robust and reliable to a degree that Linux still doesn't achieve in a general-purpose environment. It's also immensely scaleable--dealing gracefully with as big of a machine as you want to throw at it. In terms of technology, Solaris 10 was a complete rewrite, and in many ways was a rethinking of Unix. It provides service-level fault tolerance (via SMF, which replaces the traditional
Finally there's ZFS, which is truly a new filesystem--the first in a long time on any platform. It combines filesystem operations with volume management, and results in a filesystem that has been abstracted from the hardware it's running on.
These are just the highlights of the most robust Unix out there right now.
What Solaris 10 will NOT buy you though, is the same end-user experience of Linux. The graphics routines, multimedia applications, and audio support just aren't at the same level in Solaris yet. That's changing fast enough, but it hasn't caught up yet.
So IT has an interesting difference from most other parts of a company.
An ordinary user can't take down a phone system by accidentally hitting three buttons on their phone. They can't take down the power to the building by turning on the wrong power switches in their office. However, they CAN potentially take down a significant part of computer infrastructure by opening the wrong attachment, and thanks to the virus and spyware-infected world of computing, that's relatively likely.
With this sort of power one would normally expect an equal amount of knowledge and responsibility, but it's not really their fault that they've had that much power thrust into their hands. Also, they have their own jobs which (generally) require their time and attention, and they shouldn't have to become an IT expert to do their own work.
Furthermore, there are some truly unrealistic demands placed on IT, probably moreso than most departments. People who phone and say, "my computer is broken--fix it RIGHT NOW!!!" may be completely unable to do their work (which is bad), but are unwilling to look at a larger picture of, 'I can fix your computer, or I can fix all outbound mail being sent to our largest partner.'
Bottom line is that computers are both essential to business, and also the most unreliable and time-intensive part of modern corporate infrastructure.
However, none of that makes any difference in the end. If you're employed to support people, then you'd damned well better support them professionally. Keep your bitching to jokes within the team and keep them to a minimum. Treat people with all the respect in the world and always work to the best of your abilities for them, even if they don't deserve it. That's what professionalism is all about.
1) Probably.
2) Who cares?
If your music is so bad that Timbaland is producing a copy of it, you should take up mime or tiddlywinks as a means of creative expression.
Consider Nelly Furtado: Intelligent, talented, creative musician who has been turned into a shite-generating whore. All thanks to Timbaland.
You seem to misunderstand what theory means in the context of scientific rigour.
Evolution is a theory. Global warming is a theory. Gravity is a theory too! So is light.
"Respected members of the scientific community" stopped making such statements as soon as contrary 'evidence' was examined. It's becoming very clear that global warming is happening--that doesn't make it less of a theory, by the way. The relative causes are still getting sorted out, but anyone who actually takes the time to study and understand the evidence can't avoid the obvious and clear conclusion.
If you're so much smarter than them, then why do you think the parent is a woman, when the article clearly states otherwise?
I would argue against that claim.
When I say "strict vi compatability," I mean that I shouldn't know that I'm using anything other than plain ol' vanilla vi. vim's compatability mode is...weak. Lots of things that might be closer, but still aren't vi.
Step away from Linux, and you'll discover that it's the oddity in the crowd.
Solaris: vi
HP-UX: vi
AIX: vi
IRIX: vi
*BSD: nvi (not exactly vi, but much closer than vim)
Why vim took over the Linux world baffles me. It always takes me a week on a new Linux machine to figure out how to turn off all of the unwelcome 'on by default' features. If vim is invoked as vi, it should revert to strict vi compatibility mode by default.
Yep. Usenet.
Well unfortunately, Coke really does deserve to be in the bad corner.
Do some digging on Coke in India (especially Kerala), and find out just how much damage they've done. Fertile ground has been turned into parched earth, groundwater has been contaminated across the country, and there's some question about the incidence of birth defects near the bottling plants. Unfortunately, it's far enough away from the Western World(tm), and also India's special interest groups tend to stray farther from the truth than they would here, so it's not the debacle it should be.
As an aside, Pepsi is right up there in bad behaviour as well. Both companies were temporarily banned from production in Kerala, in August of 2006, but the ban was overturned.
No no, I'm not claiming that technical superiority will lead to market share. That's something that is determined primarily by marketing (sadly).
However, what I'm saying is that Solaris holds enough of an existing marketshare and has enough interesting stuff to capture a significant amount of the hobbyist interest. That doesn't often lead to real marketshare, but is another world unto itself. Furthermore, it's one that is primarily (or at least significantly) driven by technology (and thanks to Richard Stallman, licensing minutae).
For the most part, Linux and Solaris still don't compete. They're isolated by role, without a huge amount of crossover (with the exception of technical workstations, where Linux is kicking Solaris around the block).
If you read the article, you'd find that this is the first time. Clinton enacted a law, and Bush (!) has enforced it. From here on in, it will happen every year, but this is the first.
I agreed with everything you mentioned, except this:
"This literary device has been completely abandoned since the '80s, I think, and nowadays the author tends to throw the reader "in media res" and add details as the story moves along."
I think you need to expand your reading sphere a bit. Many authors still use this technique, in fact most of the ones I read. Maybe avoiding it is something that's limited to post-Gibson SF authors.
Some examples of currently writing authors who use this technique: Murakami, Gaiman, Marquez. They may not be quite as dull and plodding about it as the lectures of the '50s, but to be fair, the _good_ authors of the '50s weren't dull and plodding about it either.
Well, I've already replied to one of your posts tonight--why not make a pair?
"OpenSolaris hasn't been big enough to experience fragmentation yet, so you can't judge how well OpenSolaris will hold up at scale of development community..."
This is entirely accurate. However, I'm inclined to believe that fragmentation will be less of a factor for OpenSolaris, due to the fact that it ultimately feeds back into a single definitive snapshot. Solaris will always be the watermark for OpenSolaris, so as new projects get brought in, they'll be (1) incorporated into Solaris and become part of the canon, (2) left out, and gradually fade away, or (3) be important enough to cause a fork. The key is that option (1) doesn't exist per se for Linux, so option (3) gets invoked far more often in that realm. There will be some fragmentation, yes, but OpenSolaris is constantly being refreshed from Solaris (and also feeding back into it), so the potential for fragmentation, in my opinion, is far less.
Of course, it's early. We'll find out in two years if I'm right.