The term "double opt-in" was coined by the Direct Marketing Association as an alternative to "opt-in" because it has the connotation of being overly burdensome. As for the effectiveness of MAPS, check out this ISP status page which I found at random. Here's what I gleaned from a quick look at their graphs:
The majority of their inbound mail is spam.
The MAPS RSS (relay spam stopper) accounts for the vast majority of spam stopped.
The MAPS RBL accounts for the smallest amount of spam stopped.
Some thoughts on relating this data to your post: The RSS acts directly, by blocking spam from hopelessly open relays. The RBL acts indirectly by placing pressure on the owner of a netblock. RBL nominations can be based on spam support as well as spam origination. Therefore, the fact that the RBL directly stops very little spam does not necessarily mean it's ineffective. The RBL is a deterrent to harboring spam. Maybe this is what you mean when you say "the only war they are winning is political."
The RBL has made it difficult to brazenly maintain a spam support website.
Experian has paying clients and has the right to send e-mail to people who have indicated interest.
Many spammers have paying clients. Why do you think that this legitimizes their activity? As for a "right to send email," you are talking about making a TCP connection to a box owned by someone else. You do not have an absolute right to connect to someone else's computer against his will. Even if you have "paying clients".
First the people behind MAPS work to build the Internet (with things like BIND and sendmail), then they change their mind and punch it full of holes.
It sure looks strange, when you put it that way. How about this: first they built the infrastructure, then they kept it clear of junk.
Something needs to be done about spam, but MAPS isn't the right solution.
Great. So while you're working on the right solution, please don't be offended that more practical-minded people just don't accept traffic from malicious companies.
I don't think that anyone has the right to tell a business how to run things.
Really? Let's just abolish the government then, and let businesses do whatever they want. But wait a minute. Suppose a business B1 wants to null route another business B2 because it's flooding the network with junk. Does anyone have the right to tell B1 how to run things? What if B1 shares the info with B3, B4, and B5, all of whom object to floods of junk? Don't they have the right to blackhole the offender? You can't complain because you don't think that anyone has the right to tell a business how to run things. Your position is self-contradictory.
One service (SpamCop) constantly accuses our company of sending SPAM. We respond with the IP address or the person who subscribed them to our list, date/time stamp, and the URL of the page that states the conditions of the submission/subscription...
Your company is a spam house, pure and simple. Even if the log information you offer is legit, which is open to question, using it would require the cooperation of the ISP that owns that address. It also involves substantial work on the part of the injured party, all to end up at a dialup account that was either stolen or disconnected for abuse eight hours after it was activated.
That's why the pressure will rightfully be placed on your company, because you are the people running a "single opt-in" (which means no opt-in) list. Eventually you will cause enough harm that you will be RBL'd.
I definitely agree with you that the war on spam is making the net more unfriendly, and imposing a bigger cost on technical individuals than on corporations. The situation you describe is unpleasant, and I understand why you may not want to relay through your ISP's mailserver. But here's why your anger may still be misplaced: MAPS is merely acting as a proxy for many mail admins who do not want to accept mail from dialup and similar addresses. Unfortunately, for most people who make decisions about mail handling, your IP address represents more risk than reward.
...given that MAPS and services like it, automatically blackhole email from dynamically served DNS entries...
I think you are confusing two different issues. Many sites reject mail from IP's that don't have reverse DNS, or which don't resolve to the domain they're claiming to be from. This has nothing to do with MAPS. There is also the possibility that your IP range is on the Dial-up User List (DUL). Pac Bell, in particular, tends to place networks on this list even if they're not dialup. I suggest you query the RBL to find out.
MAPS claimed they did not filter by domain name...
The MAPS RBL is a system which accepts an IP address and returns a result if the IP is associated with spam. It does not work with domain names at all.
As several pro-MAPS people have explained here in/., this is a lie. They do not filter specific IP addresses, they filter whole blocks.
Netblock != domain. The fact that MAPS blacklists spamming netblocks does not imply that they target domains. Domains exist within the context of name resolution. They have nothing to do with network architecture. It's possible to host a thousand domains at one IP address, or have one domain that includes thousands of IP addresses.
My blocked IP addresses were not even in the same subnet. They were not filtering specific IP addresses, or even whole blocks of addresses; they filtered my ISP's domain name.
The first sentence does not support the second. They probably listed some or all of the netblocks which your ISP controls. The subnet of which you speak was probably contained within one of those netblocks. Your conclusion that they "filtered" your ISP's domain name does not make sense.
That's the lie.
Before slinging insults, you might consider learning a bit more about the internet. It does not work the way you think it does.
Making this settlement goes against all their principles...
I think your concern might be excessive. On their homepage, MAPS lists three 'truces': Exactis/Experian, Media3 and Harris. The net effect is that in each case the parties came to an agreement. MAPS claims they got some concessions from Experian, though not as much as they hoped for.
As noted, unless the agreement is very broad, they can certainly name on their web site the companies they have been compelled not to block...
And the MAPS press release ends with:
And, of course, we are still free to choose to accept or reject email from them on our own personal networks.
which seems like a pretty strong hint.
And I agree with you about local, ad-hoc blocklists. This is the ultimate reality behind all the bickering over MAPS - it is just a proxy for the judgement of individual sysadmins. Ban MAPS, and everyone must spend time maintaining blocklists. And it will be easier to get on than to get off.
It's quite questionable how well MAPS blocks spam [cnet.com].
That link is worthless. It's basically an infomercial for a commercial spam filtering service. Reading between the lines, they probably have a rule-based or fuzzy logic approach that judges messages on the fly. MAPS obviously does not do that - they list IP addresses of bad guys. Based on this difference, the article claims that MAPS only blocks "2% of spam". Sounds like they generated their own spam-like messages, and tested different software's efficiency in stopping them. Of course MAPS is not going to filter a message based on spam-like content. Likewise, the article claims that MAPS wrongly rejected non-spam mail as spam. Again, MAPS does not claim that an individual message is spam. Rather, it says that it comes from a spam-friendly IP address.
As for your complaint about listing large blocks, do you think any other approach is really practical? Spam-friendly netblock owners will move the spammers around to evade the ban. I don't see how MAPS can get into the business of determining "innocence". If someone harbors spam, he is dishonest, and his statements about who uses IP addresses within his range should be seen as ruses, rather than honest reporting.
It is assenine to expect every individual who runs a mailing list to setup their list for double opt-in.
First of all, "double opt-in" is spammer jargon for opt-in. What spammers call "single opt-in" is not opt-in at all, but merely arbitrary addtion of a mail address to a list without permission of the address's owner. By using this phrase, you paint yourself as an ally or dupe of spammers.
System administrators are responsible for keeping their machines and domains running smoothly. That includes blackholing IP ranges that cause massive harm. MAPS is just a way for a bunch of sysadmins to pool their knowledge.
The past teaches us that change in complex systems (ergo: email / mailing lists / spamming practices) requires consolidated standards bodies and a unified lobbying effort for change.
Actually, MAPS is winning. SPAM has a very bad name on the internet and large corporations shun it. Spammers are driven to more extreme and more criminal tactics, which show their stripes more clearly. Spamming operations frequently involve stolen credit card numbers. If we ever transition to secure credit cards, I guess spammers will break into buildings at night to use their connectivity. Or else they might finally give up.
You could be missing the big picture. Every mail admin for a large site has to maintain some kind of blocklist to avoid being inundated with spam. Almost everyone complaining about such lists is already benefiting from them and doesn't realize it.
MAPS is a step past these ad-hoc lists because it has some semblance of due process and some method for removal. MAPS understands that most domains harboring spammers don't realize what they're doing and will change their ways.
Take away MAPS and the spamming domain or netblock will end up permanently blocked by thousands of large sites. And nobody has time to review these blocklists.
You make good points about the office formats and the influence of IE. I was so carried away with my point that I overlooked those.
But pointing out that linux runs 30% of web servers (for some definition of web servers) does not invalidate my point - it shows that a given Linux box is much more likely to be a web server than a given Windows box. Assumption: there are far more Windows boxes than Linux boxes.
As for BSD vs Linux, the distinction is pretty tiny, isn't it? But whenever I read a current paper by someone researching encryption/tempest/any new scientific thing, to the extent that their platform is mentioned it's almost always Linux. BSD had a greater contribution in the past before Linux matured and gained acceptance. But nowadays, companies like SGI are trying to add cool file systems to Linux. If they did this with BSD, they'd be inviting competitors to simply swallow the code into their proprietary OS's.
Hate to break up the banter with newslist-newbie-type question, but what exactly are pipes used for?
Pipes are one of several IPC (interprocess communication) techniques used in Unix. Type 'perldoc perlipc' for a (perl-centric) overview. Example of where I used pipes: a program that forks, with one process handling the GUI and several handling web transactions. How can the webgetters share data with the GUI? Some kind of IPC. Why did I fork? Because I didn't want the GUI stuck while downloading data from the web.
Named pipes are a bit special, because they sit there in the file system looking much like ordinary files, even when no process is using them.
Never once have I ever heard of ANYTHING taught except for "Elegance, elegance, elegance."
That is not the same as optimizing for speed. Professors generally don't want speed at the expense of elegance. For example, recursion is frequently the most elegant approach to a problem. However a simple loop can be faster for some things.
Beta was "better" than VHS, but VHS won out because it won the mass market. Unfortunately for you, MASS MARKET has more sway than the elites ever will.
It's not that simple. Computer technology frequently originates with researchers and hackers, and continues to be marked by its origin as it trickles down. The web brought a ton of unixisms to Windows: the '/' directory separator, which is now intermittently supported in various parts of Windows/Office, MIME types, gzip, and of course TCP/IP. Look at it another way. People who work with electricity and electronics use volts, ohms and amperes to measure electricity. This happens to be the same set of units used by electrical engineers. You could devise a different system of units that still fulfills V=IR (call it MSelectricity). If that were popular, I think it would be pushed out gradually by the 'scientific' units, because technology flows from engineers to technicians, not the other way around.
When I was an audio/video engineer, I spec'd tons of VHS decks for surveillance, playback and other applications. If I had spec'd Beta, it would have been a very hard decision to defend because Beta is nonstandard.
Now I work with computers, and when we have to implement a new system I push to do it on Linux. And it always is on Linux unless the software just doesn't exist.
The difference between Windows/Linux and VHS/Beta is that the network effect simply benefited the majority in the VCR instance. In the computer instance, however, the vast majority of Windows PCs offer nothing to the world. Therefore they are a nonfactor. A linux box is more likely to be involved in software development or in serving data over a network.
That bullshit about hardware invalidating the need for fast efficient code, is the bullshit rhetoric taught in college classes that brought us the blue screen of death in the first place.
There may be some linkage, but I'd draw a distinction between the bullshit of academia and bullshit of Microsoft. Both can produce slow code, but for different reasons: the academic produces robust, elegant code. It's slow because it's full of error checking and levels of abstraction. Microsoft produces buggy, badly designed code. I remember reading an article describing how a relatively small Microsoft application had been made by pasting code from a larger one, resulting in insane bloat. Someone reverse engineered it and found tons of code in there that was not called by the app.
There is a tradeoff between safety and speed. An OS with intrinsic guards against BSOD would be slower. It would probably be a microkernel with paranoid checking of inbound messages.
There will always be a role for fast, efficient code, but for a lot of business programming robustness is more important.
You make a good point about Linux seizing the strategic position - the people who advance computing. Two counterpoints, though: Microsoft is working hard to inject their OS into academia. Their technique consists of free and discounted stuff in conjunction with the 'elitism' ploy - they try to make people feel like special members of the MS tribe who are allowed exclusive access to source code and inside info. Not all academics are proof against such ploys. Secondly, through laws like SSSCA and crypto restrictions, Linux could become illegal or practically illegal (won't run on newer hardware).
Microsoft isn't going to just sit there and wait for us to run them over.
How's the hardware going to tell if you're viewing restricted content when the viewing operation isn't even in the same machine code?
The content will be encrypted. Only the trusted hardware (video card) can decrypt it. The virtual machine doesn't help any if the content is encrypted and you don't have the decryption keys.
All of these technologies will be based upon security-by-obscurity because that is the only avenue available. Look at DVD's if you want to see how these technologies are implemented.
DVD CSS did not inherently rely on STO. It was broken due to a bad, homemade cipher and the greedy decision to allow software implementations of a DVD player.
The proof is simple: the computer ultimately has to have access to the raw data for display purposes.
No it doesn't. The video hardware can do the MPEG decoding. It can do decryption too. The general purpose computer simply feeds data to the trusted subsystem. There is no need for the computer to have access to the decrypted data.
This law is insanely broad and could well ban Linux. However it is quite possible to have strong, highly restrictive content control that works just fine with Linux. It's based on trusting hardware, not software.
If Jack buys a copy of a song via the internet, it will be somehow linked to him or his hardware. If he mails it to Jill, it will not play on her hardware.
When Jack plays the song, the decrypted version exists briefly in computer hardware. The intent of this law and associated technologies is to stop Jack from intercepting this decrypted information.
Essentially the computer is divided into two compartments, Red and Black. Black can connect to the internet, and the user has complete control over Black. Red is secured. Black can stick protected material into Red, and delete material in Red. Black can list, play, pause, stop material in Red. Email is a function of Black. It has no effect on the security of Red.
FYI, here's how to start a business: Go down to city hall and fill out a DBA ("doing-business-as") form. Now start doing business under that name. Congratulations on your new business!
Really? How about resale permits? Occupational licensing? Insurance? Negotiating a decent lease with a landlord? Negotiating revolving credit with a bank? They want a balance sheet? What's that again? What if a client refuses to pay? How do you decide a client's credit-worthiness, anyhow? How to you establish credit for your business? What rights do your employees have? What if they try to organize? How exactly did you decide that a sole proprietorship was the best form for this business?
I ran my own business for one year. If you start a business, I think you have some surprises coming.
I disagree. Where I work, the managers insist on HTML working on all popular browsers. I've frequently seen the GUI guy get a tongue-lashing for forgetting to test on Mac. And we need the pages to make at least half-assed sense in lynx, because this makes it easier to troubleshoot connectivity and proxy issues from machines all over the internet.
Mac users are few, but they have a voice out of proportion to their numbers. Browser specific HTML is only for very small, ignorant projects or for those who have a captive audience. If you're trying to compete in the big world, you have to support everything.
Imagine demoing your system to a prospective client or investor, and falling on your face because they use Macs and you never tested with Macs. You create a permanent perception of your product as shoddy and unfinished.
Think about it -- should someone who has no education in Computer Science be a sysadmin or an IT?
What makes you think CS has anything to do with system administration? If I were hiring a sysadmin, I wouldn't look for a CS graduate. I've seen some truly top-notch sysadmins, and none of them were CS majors. You're more likely to find great sysadmins from chemistry or physics backgrounds. Their story always begins, "The department bought a Sun, and somebody had to unpack it and plug it in..."
Actually I think they are frequently paying for safety. I've worked a lot in construction, and the typical construction worker has never had his "ideas" seriously challenged and picked apart. Therefore he is not a suitable contributor to the corporate decision-making process. When someone has graduated college, you can be assured that he's been exposed to debate and understands that ideas are fair game and his fists won't rescue him from faulty thinking. As a greater proportion of people get degrees, the presumption increases that the remaining folks are too dumb or too violent to do so.
I feel your pain, but look at the other side of the coin. The University has a lot invested in its brand. When you graduate, you're entitled to use their brand. This helps you in getting jobs, and potentially hurts them by diluting their brand.
Like all societies, America has class divisions. However, ours are much more subtle. Clothing and speech are no longer the cues they once were. Countless times in your carreer your class will be assessed. Mythology and American History both happen to be part of the common language of the American ruling class. If you do not understand an allusion to Cerberus or to the Boston Tea Party, you contribute to an impression of yourself as a lower class individual, not suitable for high level policy discussions. And by extension, you cast disrepute on the University whose brand you bear.
The trouble is, liberal arts education is nearly dead. Marxism, deconstructionism and a thousand other fads have rotted the once-proud liberal arts curriculum till all that remains is a grinning skull. Intelligent people who perhaps were destined for liberal arts become engineers and programmers instead. The people remaining in liberal arts are mostly those who should not be in a University at all. They will party for four years, sleep in class while the professor deconstructs the imperialist tropes of Shakespear, and emerge on the job market as mental children in the guise of adults.
They'd be much better off if they just knew Latin, Greek, arithmetic and the classics. Then at least they'd equal schoolboys of 100 years ago. Four years of mental laziness at that critical age, however, permanently harms the mind.
I don't have a degree either, and I haven't found it much of an obstacle. Then again, I've never sent my resume to an HR department. Realize that lots of jobs exist in places that don't even have HR departments.
Your real problem is probably lack of salesmanship. Geeks generally hate to sell, and I can't say I like it. But I have learned to pound the pavement and offer my services. I don't respond to ads and I don't email resumes. I have managed to get enjoyable jobs.
Please go to the library and read some books on selling. Realize that you have a right to make your pitch, and if anyone fences you out with receptionists or other tactics, you have the right to do anything necessary to bypass that fence. It may be that you will need to be starving before you understand that right. If you had kids to support, you would probably understand it now.
Slashdot Mirror
Some thoughts on relating this data to your post: The RSS acts directly, by blocking spam from hopelessly open relays. The RBL acts indirectly by placing pressure on the owner of a netblock. RBL nominations can be based on spam support as well as spam origination. Therefore, the fact that the RBL directly stops very little spam does not necessarily mean it's ineffective. The RBL is a deterrent to harboring spam. Maybe this is what you mean when you say "the only war they are winning is political."
The RBL has made it difficult to brazenly maintain a spam support website.
Many spammers have paying clients. Why do you think that this legitimizes their activity? As for a "right to send email," you are talking about making a TCP connection to a box owned by someone else. You do not have an absolute right to connect to someone else's computer against his will. Even if you have "paying clients".
It sure looks strange, when you put it that way. How about this: first they built the infrastructure, then they kept it clear of junk.
Great. So while you're working on the right solution, please don't be offended that more practical-minded people just don't accept traffic from malicious companies.
Really? Let's just abolish the government then, and let businesses do whatever they want. But wait a minute. Suppose a business B1 wants to null route another business B2 because it's flooding the network with junk. Does anyone have the right to tell B1 how to run things? What if B1 shares the info with B3, B4, and B5, all of whom object to floods of junk? Don't they have the right to blackhole the offender? You can't complain because you don't think that anyone has the right to tell a business how to run things. Your position is self-contradictory.
Your company is a spam house, pure and simple. Even if the log information you offer is legit, which is open to question, using it would require the cooperation of the ISP that owns that address. It also involves substantial work on the part of the injured party, all to end up at a dialup account that was either stolen or disconnected for abuse eight hours after it was activated.
That's why the pressure will rightfully be placed on your company, because you are the people running a "single opt-in" (which means no opt-in) list. Eventually you will cause enough harm that you will be RBL'd.
I definitely agree with you that the war on spam is making the net more unfriendly, and imposing a bigger cost on technical individuals than on corporations. The situation you describe is unpleasant, and I understand why you may not want to relay through your ISP's mailserver. But here's why your anger may still be misplaced: MAPS is merely acting as a proxy for many mail admins who do not want to accept mail from dialup and similar addresses. Unfortunately, for most people who make decisions about mail handling, your IP address represents more risk than reward.
I think you are confusing two different issues. Many sites reject mail from IP's that don't have reverse DNS, or which don't resolve to the domain they're claiming to be from. This has nothing to do with MAPS. There is also the possibility that your IP range is on the Dial-up User List (DUL). Pac Bell, in particular, tends to place networks on this list even if they're not dialup. I suggest you query the RBL to find out.
The MAPS RBL is a system which accepts an IP address and returns a result if the IP is associated with spam. It does not work with domain names at all.
Netblock != domain. The fact that MAPS blacklists spamming netblocks does not imply that they target domains. Domains exist within the context of name resolution. They have nothing to do with network architecture. It's possible to host a thousand domains at one IP address, or have one domain that includes thousands of IP addresses.
The first sentence does not support the second. They probably listed some or all of the netblocks which your ISP controls. The subnet of which you speak was probably contained within one of those netblocks. Your conclusion that they "filtered" your ISP's domain name does not make sense.
Before slinging insults, you might consider learning a bit more about the internet. It does not work the way you think it does.
I think your concern might be excessive. On their homepage, MAPS lists three 'truces': Exactis/Experian, Media3 and Harris. The net effect is that in each case the parties came to an agreement. MAPS claims they got some concessions from Experian, though not as much as they hoped for.
And the MAPS press release ends with:
which seems like a pretty strong hint.
And I agree with you about local, ad-hoc blocklists. This is the ultimate reality behind all the bickering over MAPS - it is just a proxy for the judgement of individual sysadmins. Ban MAPS, and everyone must spend time maintaining blocklists. And it will be easier to get on than to get off.
That link is worthless. It's basically an infomercial for a commercial spam filtering service. Reading between the lines, they probably have a rule-based or fuzzy logic approach that judges messages on the fly. MAPS obviously does not do that - they list IP addresses of bad guys. Based on this difference, the article claims that MAPS only blocks "2% of spam". Sounds like they generated their own spam-like messages, and tested different software's efficiency in stopping them. Of course MAPS is not going to filter a message based on spam-like content. Likewise, the article claims that MAPS wrongly rejected non-spam mail as spam. Again, MAPS does not claim that an individual message is spam. Rather, it says that it comes from a spam-friendly IP address.
As for your complaint about listing large blocks, do you think any other approach is really practical? Spam-friendly netblock owners will move the spammers around to evade the ban. I don't see how MAPS can get into the business of determining "innocence". If someone harbors spam, he is dishonest, and his statements about who uses IP addresses within his range should be seen as ruses, rather than honest reporting.
First of all, "double opt-in" is spammer jargon for opt-in. What spammers call "single opt-in" is not opt-in at all, but merely arbitrary addtion of a mail address to a list without permission of the address's owner. By using this phrase, you paint yourself as an ally or dupe of spammers.
System administrators are responsible for keeping their machines and domains running smoothly. That includes blackholing IP ranges that cause massive harm. MAPS is just a way for a bunch of sysadmins to pool their knowledge.
Actually, MAPS is winning. SPAM has a very bad name on the internet and large corporations shun it. Spammers are driven to more extreme and more criminal tactics, which show their stripes more clearly. Spamming operations frequently involve stolen credit card numbers. If we ever transition to secure credit cards, I guess spammers will break into buildings at night to use their connectivity. Or else they might finally give up.
You could be missing the big picture. Every mail admin for a large site has to maintain some kind of blocklist to avoid being inundated with spam. Almost everyone complaining about such lists is already benefiting from them and doesn't realize it.
MAPS is a step past these ad-hoc lists because it has some semblance of due process and some method for removal. MAPS understands that most domains harboring spammers don't realize what they're doing and will change their ways.
Take away MAPS and the spamming domain or netblock will end up permanently blocked by thousands of large sites. And nobody has time to review these blocklists.
You make good points about the office formats and the influence of IE. I was so carried away with my point that I overlooked those.
But pointing out that linux runs 30% of web servers (for some definition of web servers) does not invalidate my point - it shows that a given Linux box is much more likely to be a web server than a given Windows box. Assumption: there are far more Windows boxes than Linux boxes.
As for BSD vs Linux, the distinction is pretty tiny, isn't it? But whenever I read a current paper by someone researching encryption/tempest/any new scientific thing, to the extent that their platform is mentioned it's almost always Linux. BSD had a greater contribution in the past before Linux matured and gained acceptance. But nowadays, companies like SGI are trying to add cool file systems to Linux. If they did this with BSD, they'd be inviting competitors to simply swallow the code into their proprietary OS's.
He did test with a 4k buffer on NT. It was slow, so he tried the huge buffer.
Pipes are one of several IPC (interprocess communication) techniques used in Unix. Type 'perldoc perlipc' for a (perl-centric) overview. Example of where I used pipes: a program that forks, with one process handling the GUI and several handling web transactions. How can the webgetters share data with the GUI? Some kind of IPC. Why did I fork? Because I didn't want the GUI stuck while downloading data from the web.
Named pipes are a bit special, because they sit there in the file system looking much like ordinary files, even when no process is using them.
That is not the same as optimizing for speed. Professors generally don't want speed at the expense of elegance. For example, recursion is frequently the most elegant approach to a problem. However a simple loop can be faster for some things.
It's not that simple. Computer technology frequently originates with researchers and hackers, and continues to be marked by its origin as it trickles down. The web brought a ton of unixisms to Windows: the '/' directory separator, which is now intermittently supported in various parts of Windows/Office, MIME types, gzip, and of course TCP/IP. Look at it another way. People who work with electricity and electronics use volts, ohms and amperes to measure electricity. This happens to be the same set of units used by electrical engineers. You could devise a different system of units that still fulfills V=IR (call it MSelectricity). If that were popular, I think it would be pushed out gradually by the 'scientific' units, because technology flows from engineers to technicians, not the other way around.
When I was an audio/video engineer, I spec'd tons of VHS decks for surveillance, playback and other applications. If I had spec'd Beta, it would have been a very hard decision to defend because Beta is nonstandard.
Now I work with computers, and when we have to implement a new system I push to do it on Linux. And it always is on Linux unless the software just doesn't exist.
The difference between Windows/Linux and VHS/Beta is that the network effect simply benefited the majority in the VCR instance. In the computer instance, however, the vast majority of Windows PCs offer nothing to the world. Therefore they are a nonfactor. A linux box is more likely to be involved in software development or in serving data over a network.
There may be some linkage, but I'd draw a distinction between the bullshit of academia and bullshit of Microsoft. Both can produce slow code, but for different reasons: the academic produces robust, elegant code. It's slow because it's full of error checking and levels of abstraction. Microsoft produces buggy, badly designed code. I remember reading an article describing how a relatively small Microsoft application had been made by pasting code from a larger one, resulting in insane bloat. Someone reverse engineered it and found tons of code in there that was not called by the app.
There is a tradeoff between safety and speed. An OS with intrinsic guards against BSOD would be slower. It would probably be a microkernel with paranoid checking of inbound messages.
There will always be a role for fast, efficient code, but for a lot of business programming robustness is more important.
You make a good point about Linux seizing the strategic position - the people who advance computing. Two counterpoints, though: Microsoft is working hard to inject their OS into academia. Their technique consists of free and discounted stuff in conjunction with the 'elitism' ploy - they try to make people feel like special members of the MS tribe who are allowed exclusive access to source code and inside info. Not all academics are proof against such ploys. Secondly, through laws like SSSCA and crypto restrictions, Linux could become illegal or practically illegal (won't run on newer hardware).
Microsoft isn't going to just sit there and wait for us to run them over.
The content will be encrypted. Only the trusted hardware (video card) can decrypt it. The virtual machine doesn't help any if the content is encrypted and you don't have the decryption keys.
DVD CSS did not inherently rely on STO. It was broken due to a bad, homemade cipher and the greedy decision to allow software implementations of a DVD player.
No it doesn't. The video hardware can do the MPEG decoding. It can do decryption too. The general purpose computer simply feeds data to the trusted subsystem. There is no need for the computer to have access to the decrypted data.
This law is insanely broad and could well ban Linux. However it is quite possible to have strong, highly restrictive content control that works just fine with Linux. It's based on trusting hardware, not software.
If Jack buys a copy of a song via the internet, it will be somehow linked to him or his hardware. If he mails it to Jill, it will not play on her hardware.
When Jack plays the song, the decrypted version exists briefly in computer hardware. The intent of this law and associated technologies is to stop Jack from intercepting this decrypted information.
Essentially the computer is divided into two compartments, Red and Black. Black can connect to the internet, and the user has complete control over Black. Red is secured. Black can stick protected material into Red, and delete material in Red. Black can list, play, pause, stop material in Red. Email is a function of Black. It has no effect on the security of Red.
Really? How about resale permits? Occupational licensing? Insurance? Negotiating a decent lease with a landlord? Negotiating revolving credit with a bank? They want a balance sheet? What's that again? What if a client refuses to pay? How do you decide a client's credit-worthiness, anyhow? How to you establish credit for your business? What rights do your employees have? What if they try to organize? How exactly did you decide that a sole proprietorship was the best form for this business?
I ran my own business for one year. If you start a business, I think you have some surprises coming.
I disagree. Where I work, the managers insist on HTML working on all popular browsers. I've frequently seen the GUI guy get a tongue-lashing for forgetting to test on Mac. And we need the pages to make at least half-assed sense in lynx, because this makes it easier to troubleshoot connectivity and proxy issues from machines all over the internet.
Mac users are few, but they have a voice out of proportion to their numbers. Browser specific HTML is only for very small, ignorant projects or for those who have a captive audience. If you're trying to compete in the big world, you have to support everything.
Imagine demoing your system to a prospective client or investor, and falling on your face because they use Macs and you never tested with Macs. You create a permanent perception of your product as shoddy and unfinished.
What makes you think CS has anything to do with system administration? If I were hiring a sysadmin, I wouldn't look for a CS graduate. I've seen some truly top-notch sysadmins, and none of them were CS majors. You're more likely to find great sysadmins from chemistry or physics backgrounds. Their story always begins, "The department bought a Sun, and somebody had to unpack it and plug it in..."
Actually I think they are frequently paying for safety. I've worked a lot in construction, and the typical construction worker has never had his "ideas" seriously challenged and picked apart. Therefore he is not a suitable contributor to the corporate decision-making process. When someone has graduated college, you can be assured that he's been exposed to debate and understands that ideas are fair game and his fists won't rescue him from faulty thinking. As a greater proportion of people get degrees, the presumption increases that the remaining folks are too dumb or too violent to do so.
I feel your pain, but look at the other side of the coin. The University has a lot invested in its brand. When you graduate, you're entitled to use their brand. This helps you in getting jobs, and potentially hurts them by diluting their brand.
Like all societies, America has class divisions. However, ours are much more subtle. Clothing and speech are no longer the cues they once were. Countless times in your carreer your class will be assessed. Mythology and American History both happen to be part of the common language of the American ruling class. If you do not understand an allusion to Cerberus or to the Boston Tea Party, you contribute to an impression of yourself as a lower class individual, not suitable for high level policy discussions. And by extension, you cast disrepute on the University whose brand you bear.
The trouble is, liberal arts education is nearly dead. Marxism, deconstructionism and a thousand other fads have rotted the once-proud liberal arts curriculum till all that remains is a grinning skull. Intelligent people who perhaps were destined for liberal arts become engineers and programmers instead. The people remaining in liberal arts are mostly those who should not be in a University at all. They will party for four years, sleep in class while the professor deconstructs the imperialist tropes of Shakespear, and emerge on the job market as mental children in the guise of adults.
They'd be much better off if they just knew Latin, Greek, arithmetic and the classics. Then at least they'd equal schoolboys of 100 years ago. Four years of mental laziness at that critical age, however, permanently harms the mind.
I don't have a degree either, and I haven't found it much of an obstacle. Then again, I've never sent my resume to an HR department. Realize that lots of jobs exist in places that don't even have HR departments.
Your real problem is probably lack of salesmanship. Geeks generally hate to sell, and I can't say I like it. But I have learned to pound the pavement and offer my services. I don't respond to ads and I don't email resumes. I have managed to get enjoyable jobs.
Please go to the library and read some books on selling. Realize that you have a right to make your pitch, and if anyone fences you out with receptionists or other tactics, you have the right to do anything necessary to bypass that fence. It may be that you will need to be starving before you understand that right. If you had kids to support, you would probably understand it now.
Anyway, good luck!