For a number of reasons, I'm running Red Hat Enterprise 4 on my desktop. Yes, it's not Windows or Mac, so that makes me an outlier. On the other hand, Adobe advertises that Flash is available for "Linux". If I want Flash, I have to dump RHEL 4 and load RHEL 5. One of the reasons I use the Enterprise editions is to *not* have to update my primary system every six months or so -- indeed, I'm waiting for RHEL 6 before I go through the process.
And I do have a RHEL 5 system I use on occasion...and Flash will mysteriously die on that platform. The failure occurs most often when there are many Flash objects in a Web window. Everything just goes blank, and my CPU loading shoots to the moon. That includes ads.
RHEL 4? The lack of flash prevents all sorts of problems. Not to mention being free of obnoxious ads when browsing the World Wide Web.
Before InfoWorld's truce proposal can be seriously considered, I think Adobe has to clean house first. When Flash runs well where it says it runs, then they have a better position in the peace talks.
Add to that the fact the M$ published the spec under a recognized standards body and that was the point at which the zealot's heads began to swell until the point of explosion.
That "publishing under a recognized standards body" didn't stop RAMBUS from trying to pull a fast one. It all depends how Microsoft presented the specification that determines what they can and cannot do in court to cripple use of Mono. Most of the recognized standards bodies have required that the contributor(s) grant licenses to use the ideas described by the standard. The licenses do not have to be royalty-free, just "reasonable" and uniform. Most licenses have a fixed payment per use, which means that "free" is not an option.
you do realise that America was founded on immigration? probably your great grandfather. and that immigration (a lot of it through H visas to Green Card to Citizenship) is what continues to help build the country?
s/realise/realize/
How does immigration currently help to "build the country"? The population has expanded from sea to shining sea already; there isn't an excess of arable land on which to settle. If you studied history, the "boomers" settling what is now Oklahoma were able to claim 160 acres of land simply by going there. (Then, of course, there were the "Sooners", but that's just a story of people abusing the system.) That's "building the country".
One of the issues that is rearing its ugly head is that drinkable water is, in some places, in short supply. Water is being pumped in large quantities from sparsely-populated areas to densely-populated areas. The United States of America doesn't need growth, it needs to manage the population it has. Immigration just adds to the problem.
Look up the history of Ellis Island to see the true path for immigration when immigration was at its peak. Or do you prefer being an ignorant git?
As contrasted to what I did to hard drives in 1985 for Infoworld, when we had the CMS hard drive in the IBM PC AT? My editor *still* talks about the torture test I did with a number of drives. Dropping drives on the floor while they are running...
I read Cram's column with quite a bit of amusement. What's interesting is I ran across exactly the same deal but over a different medium: telephone service over cable. A cable equipment company called me in when their customers reported they were unable to send faxes over the telephony-over-cable product. When I visted the test lab of this company (named withheld to protect the guilty) they demonstrated the failure. Interestingly, the faxing worked when they first started up the testbed, and then it got steadily worse. In order to measure the quality of the channel, I built a POTS tester that would measure line performance over a long timebase -- 24 hours, in fact. (To do a longer test, I would have needed more disk space.)
When I analyzed the data, I found that the cable system had a bad case of vibrato, which translated in modems terms to phase jitter. The oscillators the cable vendor used in it circuits weren't up to the task. In some cases, the 1-Khz phase jitter exceeded 360 degrees. No modem using phase modulation can stand that. Period.
Do the VoIP terminals, not to mention the implementations using PCs, have better oscillators? Can they do a better job of maintaining time coherence? That's even before you look at the effects of routing and propagation over the Internet.
Bob, I think you are asking too much of Vonage.
And reading way too much into what you are seeing. It isn't just the network. It's the end terminals, too. The oscillators in the cable system were supposed to be good to 20 parts per million. The cheap crystals used in the VoIP terminals and in PCs are more than an order of magnitude worse in stability. Did I mention that the cable-system oscillators used phase-locked loop technology to maintain even better accuracy with each other?
It's unfortunate that Mr. Dvorak didn't talk with the proponents of and contributors to the One Laptop Per Child (OLPC) project. He admits he depended on information on a Web site. Normally, this isn't a problem, but...unlike an organization with a for-profit motive with which Mr. Dvorak is used to dealing, there are no PR flacks in this small group of people doing the work. There isn't an army of copywriters keeping the OLPC web site up-to-the-minute. The focus of the OLPC army, about a platoon in strength, is getting the laptop built and distributed, to a price, to a performance level, to a quality level. There are no information officers here.
As a consequence, Mr. Dvorak's factual basis for his opinions appear to be flawed. That's the problem with fast-moving, lean projects that don't have a profit motive: the worker-bees don't budget time to spoon-feed journalists.
I base this critique on the facts shown at a presentation I attended last week on the project and its current status. During that presentation, many of Mr. Dvorak's criticisms were answered in full. I'll run down the factual points, based on the information I gleaned from that presentation. I don't vouch for absolute accuracy, as I wasn't taking notes, and I'm not part of the project. Keeping those caveats in mind:
* Justification: Mr. Dvorak doesn't touch on this issue at all, except in the negative and through the words of another person. He missed the one reason this project is interesting to the governments of the developing nations: it saves money in education.
Mr. Dvorak, have you looked at the price of school textbooks these days? How much does your local school spend, per year, on books for their kids? In developing countries, the textbook cost may be lower than here but it's still high compared to, say, food.
(N.B.: The situation in college is even worse. I leave research on that issue as an exercise to the reader, as most of the hits on Google about textbook pricing focus on higher education.)
You say, Mr. Dvorak, "with $100 you could almost feed a village for a year" but that same $100 doesn't cover educating ONE child for ONE year. You want to fill their stomachs, but starve their brains?
The OLPC project got the facts from the horse's mouth, the governments who have to somehow educate their children in order to raise the standard of living in their country. The cost of the laptop, roughly $20/year for the five-year life of the laptop, is less than the cost of the books needed to teach the kids. Throw in the infrastructure costs (development of electronic textbooks, "libraries", access points and their connections to a country-wide network) and the country still sees a savings.
Interestingly, like most "problems", it comes down to money.
* Manufacturing cost: While the presenter didn't provide a complete bill of materials for the laptop, the cost projections for building the laptop in million quantities falls well below $100 at the current time. Further cost reduction is possible as the laptop matures. The cost projection shown by the presenter was verified by members of the audience who have been on the front lines of manufacturing products like this laptop.
How much lower can the price go? You know as well as anyone the cost curve over lifetime of a computer product. Is $50 possible?
* Maintenance: Photos of the prototypes shown at the presentation show a modular approach very similar to that used by IBM in making the PS/2 Model 50 personal computer (and *not* used in virtually every PC made today). The only tool required to service the machines is a single screwdriver. Kids in the US, UK, Canada, and other developed countries have no problems servicing computers *not* designed to be serviced easily by untrained personnel. So the only infrastructure required is a way to get spare parts to those who need them.
* Networking: The laptops use mesh networking to communicate with each other, and to access points provided as part of the
Sorry, BULLSHIT. I like ISPConfig because it interfaces with PostFix, the MTA of choice around my shop. CPanel uses EXIM, Plesk uses QMail (hi, Dr. Ex-Lax!), with its many unaudited patches and workarounds to modern mail problems. (I have front-ended QMail with PostFix edge servers so that I don't have to deal with the many holes in QMail.) But it's an immature interface, and lacking some features that customers want.
As you might guess, I work for a web house that uses Ensim, Plesk for Windows, Plesk for Linux, and CPanel.
CPanel is the more polished package of all that I run.
i have over 3 million users and 1.5 million hits a day, they going to cover my costs for storage?
Two words: DVD ROM
Seriously, the devil is in the details. My initial reading of the press release and other "public" information is that there isn't enough detail to make an intelligent analysis to the effect of this proposed act.
If we hired smart security people, overall we'd be more secure.
I have my Washoe County, Nevada, work card for security guard work in my wallet. When are you going to step up to the plate and be one of those smarter security guards?
I once worked on a Standards-writing subcommittee, and ended up being the editor of a proposed standard. I was new to the process at the time. I took the work that was done and completely re-wrote it, from the ground up, according to the published guidelines of the Telecommunications Industry Association (TIA). I then presented my work to the subcommittee.
"It's too clear. People might actually understand it." I argued that because it was a specification for testing, it should be clear. Yes, I won the argument, but at what cost?
Over the next few years I watched as more standards were created, edited, published, submitted to the ITU, and eventually turned into Recommendations. When I asked, "what does this section REALLY try to say?" I was told that in order to understand that section I needed to know another piece of the puzzle that wasn't spelled out but was "understood" by "practitioners of the art." In other words, the specification was incomplete...but not according to the rules. I asked why. The answer I got boiled down to one thing: you can't implement the specification without the "stories around the campfire" behind them.
Put starkly, you can't play unless you join the club.
Now, in reality, people have taken these less-than-complete specifications and actually made products with them, products that successfully interoperated with those implemented by members of the club. The development time, however, was extended by the need to discover the missing pieces on one's own, or to buy the missing pieces.
Then there was the story of what eventually became V.80, which I discussed in a Slashdot interview. That particular standard proposal was so bad that I had to vote "no". Again, I ended up rewriting the entire thing so it made sense, and in addition covered not only the corner cases but also future extensions and vendor extensions. It took DAYS to prove that the two versions said technically the same thing (within limits). You could code to mine; the other was almost impossible and "open to interpretation."
Most specifications (or Standards) are written by partisan participants. It's to their best interests to write these things so that outsiders can't understand them -- be it commercial gain or personal ego. Good spec writing is HARD, and not for beginners. It takes work. It rarely pays anything to write a good specification, especially if the writer views it as a pro-forma task. Just as programmers from several decades ago viewed flow-charting as a useless task.
Just as people are starting to view Open Source not as a way to lose money but as a way to gain money, perhaps the partisans will see that writing clear, understandable, WORKABLE specifications is in their better interest....or not.
Given the current state of the art, though, I would tend to agree completely with Linus that specifications, and Standards, that don't provably track with reality deserve not "no", but "HELL NO!"
I read the entire 24 page report, with I hope some thought and consideration. What I found very, very interesting is this "fact-finding" body did nothing to examine the current structure of "Internet Control" and the role of the Internet Society and its divisions. They mention the IETF *once*, and neglect to mention that IETF RFCs are now accepted in the Standards community as Standards. International standards -- the ITU says so. Instead, the report concentrates exclusively on the role of the United States Department of Commerce and *one* US corporation, ICANN.
What about the role of the technical committees that have kept the Tier One routers running all these years without too many hiccups? How would they fit into a UN-based "oversight"? Either the routers work, or they don't. Does Grand Fenwick have anything to contribute to that process? Oh, let's not forget that NANOG is not a US-centric organization now...
A previous contributor showed the country breakdown of the participants. For my part, I looked through all the names of the people on this commission and didn't recognize a single name as part of the original Internet Construction Crew (ICC).
The report, if I were grading it on completeness, would get a D+. The report concentrates on those few things that bring certain peoples to a slow boil. I'm sure that one of the most important questions will be how to handle right-to-left writing systems in the current structure. It completely neglects those portions of oversight and control that mean the life and death of the Internet, either as we know it or as people have envisioned it in the future.
My great fear? "Regulation." As in putting together a list of conflicting requirements on users of the Internet that will spawn a whole new industry that generates not one cent of revenue. Oh, and someone has to pay for all this work and effort to make my life as an admin miserable. Can you say "Internet Tax"? I knew you could!
As a system administrator, I will continue to run my network. my routers, and my servers as I see fit. If the UN wants to play power games and screw it all up, then I as an operator and administrator will do everything technically possible to be sure that UN screwups don't affect my customers.
When $DAYJOB had a present from Sierra Pacific Power of a two-hour blackout, and we discovered there were major problems with our generator, the poor APC UPS batteries weren't able to hold up the 150 servers I run.
When the power came back on, we had 143 servers back on-line in ten minutes. We had 149 on line in fifteen minutes. We had two servers (leased dedicateds) that requires some file system repairs before they would come back on-line, but that task was finished 30 minutes after power restoration.
What's so hard about that?
(With the addition of a three-phase power transformer, our generator is working properly.)
Customers kept calling asking us why it took us so long!
People tend to forget what launched Mr. Stallman on this road toward software freedom: he wanted to use a laser printer he had on hand with his word processing program. The software didn't have drivers, and as I recall the printer didn't have documentation, either.
OK, for those of you who read NANAE, this is old news, but for the rest of you...
I'm a sysadmin who worked very hard to get a/24 listed in SPEWS delisted. The netblock was in the list because a customer of ours decided to provide DNS service to a known and notorious spammer. We earned the listing, period. I killed the bastard, reported the fact, and got the listing lowered to a zero, historical. In the process of doing that job, I learned a lot about the whole blocklist thing and realized that even the operators didn't see what they are really doing. They think it's about spam. Wrong.
It's not about spam. It's about TRUST
A listing in a recognized blocking list is a vote of "no confidence" in the IP owner's ability to run its network, to make its users -- ALL its users -- conform to the Internet society's accepted code of conduct.
Follow along with me a moment, and you'll see why I think this way. First, the Internet is, by definition, a "network of networks", a large anarchy run by a very large number of system administrators (greater than 10,000) who make private decisions about who and how they allow to access their bandwidth, systems, and services. The Internet Society and its sub-units provide a forum to publish community notes, the Requests for Comments, which are nothing more and nothing less than agreements for how to play nice in this employee-owned swimming pool.
The Internet community has decided on standards of behavior, and each system operator trusts every other system operator in the pool to conform to the rules of society, and to ensure that the users conform to the community rules -- not unlike CC&Rs in a neighborhood development that form part of the purchase contract of many homes and condominiums. Some operators have become lax in their expected enforcement of the rules on particularly not-nice people, the ones who break the rules in order to win money, or some other benefit. There are enough of these Internet con men out there that the community coined a word to describe them: "spammers."
Back in the NSF days, a lapse in administration resulted in disconnection, quick and swift, so the system adminstrators, up and down the line, toed the line to avoid being banished. In the Commercial Internet that replaced the NSF Internet, personal greed gets in the way of this remedy, and so the disdain of social customs is left largely unpunished by the society.
Just about every system operator who runs a mail service with more than three users has been yammered at by those users: "WE WANT LESS SPAM -- DO SOMETHING." Complaints to ISPs who take spammer money go largely ignored, and appeals "upstream" -- to the connection providers and to the Tier One networks -- have also gone largely ignored. So the small administrators started to implement mail filters and blocks on "spammy" IP addresses in the hopes that they can block the crap and thus appease their users.
Spammers countered by having their providers move them around in IP space, and by using techniques to "get around" the content filters. It's become a war, frankly. First there were keyword filters, and so spammers started to "do things" to their messages, like replace the letter 'o' with the digit '0' -- you've all seen the tricks. Hash identification of bulk messages were thwarted by inserting random nonsense text. Learning filters are poisoned by spammers injecting random words. And so on and so on. In addition to these content-based counters, spammers also steal resources of innocent people: open mail relays, open proxies, and hijacked Web scripts like formmail.pl, so that the wrong person gets blames for their flood of commercial feces.
What the block-list people decided is that having each of the 10,000 to 100,000 system administrators deal with this individually was eating up too much time, and there was this nifty thing already in place that could be used to reduce the system overhead of id
P.S.
Don't think the new "CAN-SPAM" law is going to fix the spam problem. The Direct Marketing Association considers the law a victory for the spam business. Check this C-NET [story]...
After reviewing and crafting a new AUP document for my boss at a Web hosting company, I'm beginning to appreciate how the CAN-SPAM law will get a handle on spam. Let me explain.
There is a lot more to the CAN-SPAM law than just spam. The thing that caught my eye when I read the actual Act of Congress and the law referenced by the Act is that the issue of unauthorized access to computers "used in interstate commerce" (and Web hosting falls into that class) has been clarified, and the penalties for unauthorized access have been strengthened. That means when some twerp uses one of my customer's Web sites to commit FormMail spam, it becomes a Federal matter. Falsified credentials isn't just a civil matter any more -- the Feds have criminalized the offense.
One of the biggest problems that the anti-spam crowd has had is with forgery. I have a domain, fluent (dash) access (dot) com, which has been used as a forged return address in someone's spew for a long time -- I found out when I turned the entire domain into a spam-trap and started getting bounces from all kinds of networks. (Some of those networks are now blocked because of that, but that's life on the Internet these days. Maybe I'll implement a sunset script for my automated access blocking...or maybe I won't.) With the new law, a forged From: line is criminalized. While the law is silent on the matter, forged Received From: lines may also be roped into the "deceit" intent of the law -- but that will have to be tested in a Court.
As a commercial operator of Internet services, it lets me write things into my AUP that now stand a chance of being enforced, and not necessarily on my nickel. Before CAN-SPAM, it just didn't pay to go after a guy who is probing my network looking for broken formmail.pl and formmail.cgi scripts. What CAN-SPAM really does is put the black hats on alert that their activities may attract more attention than they would be comfortable getting.
It also unshackles the programmer in me to write scripts to enforce some of these things using technological means instead of investigative means. For example, if I enforce the From: line restriction, the script person can't complain because it's an enforcement of an AUP requirement -- I'm just making sure that my customers toe the legal line.
For the DMA, it is a win, because it makes more people play by the rules. Consider that CAN-SPAM is also a win for the block-lists like SpamCop and SPEWS, because it reduces the detective work required to recognize the spew just starting from yet another IP address, and it also limits the methods available to the abusers to avoid being caught and blocked.
After writing my representatives via email a few months back, one of my Senators (Lamar Alexander (TN)) saw fit to add me to his mailing list.
I guess if you choose to express your opinion on the issues, you're opting in to be spammed.
What you did was present the beginning of a "business relationship" with your Congressman. If my understanding of the definition of UBE is accurate (I make no such claim) anything he sends you until you say "Stop" is not UBE/Spam.
As a SysAdmin I've been studying the DNS-based blocking lists in general and SPEWS in particular. Seeing how they say they operate, how long do you think it will take for the US Government to "win" an escalated listing in the SPEWS database?
"I'm sorry, Congressman, but the reason all your mail is being bounced is that our server IP address is listed in SPEWS. What is SPEWS? 'Spam Prevention Early Warning System.' Because we have been unable to answer complaints to abuse@house.net to their satisfaction, they have put together a 'crimes file' showing that The House of Representatives is a spam-lovin enterprise, have listed our entire netblock, and we've run out contractors to superserve our mail servers -- every time we hire one, it ends up listed in SPEWS, too."
Will the blocking lists work as they are supposed to, or are they going to take the smart path and NOT piss off the one organization who makes the "Laws of the Land?"? I can see it now: it becomes illegal for any operator of a mail server with more than 100 commercial clients to use any DNS- or domain-based blocking list.
Not exactly the death of the Internet, but possibly a case of felony if you do, damned if you don't.
For me, the uptime on my servers tell me how long it has been since the power company shut off the power for non-payment, or the last time some semi took out the main power feeder here in the mountains. My UPS is good for five hours, so minor outages (and four hours IS considered minor in the Sierras) don't cause things to roll back to zero.
Can you picture Jar-Jar complaining about the diodes down his left side? No, I don't think so. Too loose-jointed.
What I'm afriad of is that Marvin will be played by some muscle-bound beach-bum, so that the Plastic Pal Who Is Fun To Be With advertising sequence will "play better" to the US audience. (Remember in the BBC version, the robot playing beach-ball with the winsome girl in the not-much bikini?) Not unlike the title character in Rocky Horror, perhaps.
I totally agree, let me have the IP range of your boxes and I'll let you know when the next exploit comes out
I doubt you are on the IP ranges from which we allow SSH connections, so even though the exploit is there the nice folks across both ponds are going to have an, er, interesting time gaining access. IPTABLES is your friend...
Hey, security is so tight that I have to make three hops to get onto one of our systems from the "outside" -- which makes remote administration interesting.
Ahh, the joys of another afternoon spent patching boxes. I guess it is better than waiting for a vendor to come up with a patched binary package.
When I heard there was a second patched version last week, I said to myself that these things come in threes, and that I would wait for "the next round." So much for updating 50 boxes more than once.
Will the third time be the charm, or should I avoid being on the bleeding edge and wait for next week's discoveries?
(At least it isn't like the Microsoft patches, which come at less frequent intervals and usually do more damage to my apps than the protection is worth. -- Obligatory Microsoft Bash)
Don't need a FAQ, just fix the prompts and docs
on
Where is the Any Key?
·
· Score: 4, Insightful
When I had a product out in the field, I had seen the "any" key problem enough that I decided not to do that. In every prompt, I specify exactly which key to depress. The documentation made it clear that a specific key was to be depressed (pressed firmly and then released quickly).
We never got a service call on that point. EVER.
We DID get one call from a technician who said that he pressed the wrong key and it seemed to work anyway. I assured him that he was just fine and to continue working with the product.
As you can guess, the CODE didn't require a specific key, but the documentation and prompt specified one. I also didn't allow type-ahead on those prompts so that there wasn't a problem with astonishing results to clueless users who didn't release the key fast enough. I also had code to wait until the keyboard stopped sending characters before putting up the prompt "Depress the C key to continue."
Yep. You forget that the RIAA is working in the civil arena, where the rules are completely different. First, the burden of proof is the "proponderance of the evidence", you don't have to have 12 out of 12 jurors -- hell, you don't have to have 12 jurors, some cases have been decided by a panel of 6 -- but 9 out of 12, and there is no such thing as the right to remain silent.
On the other hand, the only thing the RIAA can get is money.
Slashdot Mirror
For a number of reasons, I'm running Red Hat Enterprise 4 on my desktop. Yes, it's not Windows or Mac, so that makes me an outlier. On the other hand, Adobe advertises that Flash is available for "Linux". If I want Flash, I have to dump RHEL 4 and load RHEL 5. One of the reasons I use the Enterprise editions is to *not* have to update my primary system every six months or so -- indeed, I'm waiting for RHEL 6 before I go through the process.
And I do have a RHEL 5 system I use on occasion...and Flash will mysteriously die on that platform. The failure occurs most often when there are many Flash objects in a Web window. Everything just goes blank, and my CPU loading shoots to the moon. That includes ads.
RHEL 4? The lack of flash prevents all sorts of problems. Not to mention being free of obnoxious ads when browsing the World Wide Web.
Before InfoWorld's truce proposal can be seriously considered, I think Adobe has to clean house first. When Flash runs well where it says it runs, then they have a better position in the peace talks.
You've never worked in a company with nepotism disease. I can tell. The OP said that getting the kid to do it the right way isn't going to happen.
Add to that the fact the M$ published the spec under a recognized standards body and that was the point at which the zealot's heads began to swell until the point of explosion.
That "publishing under a recognized standards body" didn't stop RAMBUS from trying to pull a fast one. It all depends how Microsoft presented the specification that determines what they can and cannot do in court to cripple use of Mono. Most of the recognized standards bodies have required that the contributor(s) grant licenses to use the ideas described by the standard. The licenses do not have to be royalty-free, just "reasonable" and uniform. Most licenses have a fixed payment per use, which means that "free" is not an option.
you do realise that America was founded on immigration? probably your great grandfather. and that immigration (a lot of it through H visas to Green Card to Citizenship) is what continues to help build the country?
s/realise/realize/
How does immigration currently help to "build the country"? The population has expanded from sea to shining sea already; there isn't an excess of arable land on which to settle. If you studied history, the "boomers" settling what is now Oklahoma were able to claim 160 acres of land simply by going there. (Then, of course, there were the "Sooners", but that's just a story of people abusing the system.) That's "building the country".
One of the issues that is rearing its ugly head is that drinkable water is, in some places, in short supply. Water is being pumped in large quantities from sparsely-populated areas to densely-populated areas. The United States of America doesn't need growth, it needs to manage the population it has. Immigration just adds to the problem.
Look up the history of Ellis Island to see the true path for immigration when immigration was at its peak. Or do you prefer being an ignorant git?
Didn't people at Microsoft, like, go to school for whatever it is that they do there, or something?
Look at Bill Gates, and the space on his wall set aside for a diploma. Then re-think your comment.
As contrasted to what I did to hard drives in 1985 for Infoworld, when we had the CMS hard drive in the IBM PC AT? My editor *still* talks about the torture test I did with a number of drives. Dropping drives on the floor while they are running...
I read Cram's column with quite a bit of amusement. What's interesting is I ran across exactly the same deal but over a different medium: telephone service over cable. A cable equipment company called me in when their customers reported they were unable to send faxes over the telephony-over-cable product. When I visted the test lab of this company (named withheld to protect the guilty) they demonstrated the failure. Interestingly, the faxing worked when they first started up the testbed, and then it got steadily worse. In order to measure the quality of the channel, I built a POTS tester that would measure line performance over a long timebase -- 24 hours, in fact. (To do a longer test, I would have needed more disk space.)
When I analyzed the data, I found that the cable system had a bad case of vibrato, which translated in modems terms to phase jitter. The oscillators the cable vendor used in it circuits weren't up to the task. In some cases, the 1-Khz phase jitter exceeded 360 degrees. No modem using phase modulation can stand that. Period.
Do the VoIP terminals, not to mention the implementations using PCs, have better oscillators? Can they do a better job of maintaining time coherence? That's even before you look at the effects of routing and propagation over the Internet.
Bob, I think you are asking too much of Vonage.
And reading way too much into what you are seeing. It isn't just the network. It's the end terminals, too. The oscillators in the cable system were supposed to be good to 20 parts per million. The cheap crystals used in the VoIP terminals and in PCs are more than an order of magnitude worse in stability. Did I mention that the cable-system oscillators used phase-locked loop technology to maintain even better accuracy with each other?
Oops.
It's unfortunate that Mr. Dvorak didn't talk with the proponents of and contributors to the One Laptop Per Child (OLPC) project. He admits he depended on information on a Web site. Normally, this isn't a problem, but...unlike an organization with a for-profit motive with which Mr. Dvorak is used to dealing, there are no PR flacks in this small group of people doing the work. There isn't an army of copywriters keeping the OLPC web site up-to-the-minute. The focus of the OLPC army, about a platoon in strength, is getting the laptop built and distributed, to a price, to a performance level, to a quality level. There are no information officers here.
As a consequence, Mr. Dvorak's factual basis for his opinions appear to be flawed. That's the problem with fast-moving, lean projects that don't have a profit motive: the worker-bees don't budget time to spoon-feed journalists.
I base this critique on the facts shown at a presentation I attended last week on the project and its current status. During that presentation, many of Mr. Dvorak's criticisms were answered in full. I'll run down the factual points, based on the information I gleaned from that presentation. I don't vouch for absolute accuracy, as I wasn't taking notes, and I'm not part of the project. Keeping those caveats in mind:
* Justification: Mr. Dvorak doesn't touch on this issue at all, except in the negative and through the words of another person. He missed the one reason this project is interesting to the governments of the developing nations: it saves money in education.
Mr. Dvorak, have you looked at the price of school textbooks these days? How much does your local school spend, per year, on books for their kids? In developing countries, the textbook cost may be lower than here but it's still high compared to, say, food.
(N.B.: The situation in college is even worse. I leave research on that issue as an exercise to the reader, as most of the hits on Google about textbook pricing focus on higher education.)
You say, Mr. Dvorak, "with $100 you could almost feed a village for a year" but that same $100 doesn't cover educating ONE child for ONE year. You want to fill their stomachs, but starve their brains?
The OLPC project got the facts from the horse's mouth, the governments who have to somehow educate their children in order to raise the standard of living in their country. The cost of the laptop, roughly $20/year for the five-year life of the laptop, is less than the cost of the books needed to teach the kids. Throw in the infrastructure costs (development of electronic textbooks, "libraries", access points and their connections to a country-wide network) and the country still sees a savings.
Interestingly, like most "problems", it comes down to money.
* Manufacturing cost: While the presenter didn't provide a complete bill of materials for the laptop, the cost projections for building the laptop in million quantities falls well below $100 at the current time. Further cost reduction is possible as the laptop matures. The cost projection shown by the presenter was verified by members of the audience who have been on the front lines of manufacturing products like this laptop.
How much lower can the price go? You know as well as anyone the cost curve over lifetime of a computer product. Is $50 possible?
* Maintenance: Photos of the prototypes shown at the presentation show a modular approach very similar to that used by IBM in making the PS/2 Model 50 personal computer (and *not* used in virtually every PC made today). The only tool required to service the machines is a single screwdriver. Kids in the US, UK, Canada, and other developed countries have no problems servicing computers *not* designed to be serviced easily by untrained personnel. So the only infrastructure required is a way to get spare parts to those who need them.
* Networking: The laptops use mesh networking to communicate with each other, and to access points provided as part of the
Sorry, BULLSHIT. I like ISPConfig because it interfaces with PostFix, the MTA of choice around my shop. CPanel uses EXIM, Plesk uses QMail (hi, Dr. Ex-Lax!), with its many unaudited patches and workarounds to modern mail problems. (I have front-ended QMail with PostFix edge servers so that I don't have to deal with the many holes in QMail.) But it's an immature interface, and lacking some features that customers want.
As you might guess, I work for a web house that uses Ensim, Plesk for Windows, Plesk for Linux, and CPanel.
CPanel is the more polished package of all that I run.
Yes, I'm quite afraid.
Two words: DVD ROM
Seriously, the devil is in the details. My initial reading of the press release and other "public" information is that there isn't enough detail to make an intelligent analysis to the effect of this proposed act.
So write and ask questions. I did.
If we hired smart security people, overall we'd be more secure.
I have my Washoe County, Nevada, work card for security guard work in my wallet. When are you going to step up to the plate and be one of those smarter security guards?
Fill the void!
I once worked on a Standards-writing subcommittee, and ended up being the editor of a proposed standard. I was new to the process at the time. I took the work that was done and completely re-wrote it, from the ground up, according to the published guidelines of the Telecommunications Industry Association (TIA). I then presented my work to the subcommittee.
"It's too clear. People might actually understand it." I argued that because it was a specification for testing, it should be clear. Yes, I won the argument, but at what cost?
Over the next few years I watched as more standards were created, edited, published, submitted to the ITU, and eventually turned into Recommendations. When I asked, "what does this section REALLY try to say?" I was told that in order to understand that section I needed to know another piece of the puzzle that wasn't spelled out but was "understood" by "practitioners of the art." In other words, the specification was incomplete...but not according to the rules. I asked why. The answer I got boiled down to one thing: you can't implement the specification without the "stories around the campfire" behind them.
Put starkly, you can't play unless you join the club.
Now, in reality, people have taken these less-than-complete specifications and actually made products with them, products that successfully interoperated with those implemented by members of the club. The development time, however, was extended by the need to discover the missing pieces on one's own, or to buy the missing pieces.
Then there was the story of what eventually became V.80, which I discussed in a Slashdot interview. That particular standard proposal was so bad that I had to vote "no". Again, I ended up rewriting the entire thing so it made sense, and in addition covered not only the corner cases but also future extensions and vendor extensions. It took DAYS to prove that the two versions said technically the same thing (within limits). You could code to mine; the other was almost impossible and "open to interpretation."
Most specifications (or Standards) are written by partisan participants. It's to their best interests to write these things so that outsiders can't understand them -- be it commercial gain or personal ego. Good spec writing is HARD, and not for beginners. It takes work. It rarely pays anything to write a good specification, especially if the writer views it as a pro-forma task. Just as programmers from several decades ago viewed flow-charting as a useless task.
Just as people are starting to view Open Source not as a way to lose money but as a way to gain money, perhaps the partisans will see that writing clear, understandable, WORKABLE specifications is in their better interest....or not.
Given the current state of the art, though, I would tend to agree completely with Linus that specifications, and Standards, that don't provably track with reality deserve not "no", but "HELL NO!"
I read the entire 24 page report, with I hope some thought and consideration. What I found very, very interesting is this "fact-finding" body did nothing to examine the current structure of "Internet Control" and the role of the Internet Society and its divisions. They mention the IETF *once*, and neglect to mention that IETF RFCs are now accepted in the Standards community as Standards. International standards -- the ITU says so. Instead, the report concentrates exclusively on the role of the United States Department of Commerce and *one* US corporation, ICANN.
What about the role of the technical committees that have kept the Tier One routers running all these years without too many hiccups? How would they fit into a UN-based "oversight"? Either the routers work, or they don't. Does Grand Fenwick have anything to contribute to that process? Oh, let's not forget that NANOG is not a US-centric organization now...
A previous contributor showed the country breakdown of the participants. For my part, I looked through all the names of the people on this commission and didn't recognize a single name as part of the original Internet Construction Crew (ICC).
The report, if I were grading it on completeness, would get a D+. The report concentrates on those few things that bring certain peoples to a slow boil. I'm sure that one of the most important questions will be how to handle right-to-left writing systems in the current structure. It completely neglects those portions of oversight and control that mean the life and death of the Internet, either as we know it or as people have envisioned it in the future.
My great fear? "Regulation." As in putting together a list of conflicting requirements on users of the Internet that will spawn a whole new industry that generates not one cent of revenue. Oh, and someone has to pay for all this work and effort to make my life as an admin miserable. Can you say "Internet Tax"? I knew you could!
As a system administrator, I will continue to run my network. my routers, and my servers as I see fit. If the UN wants to play power games and screw it all up, then I as an operator and administrator will do everything technically possible to be sure that UN screwups don't affect my customers.
My network, my rules.
When $DAYJOB had a present from Sierra Pacific Power of a two-hour blackout, and we discovered there were major problems with our generator, the poor APC UPS batteries weren't able to hold up the 150 servers I run.
When the power came back on, we had 143 servers back on-line in ten minutes. We had 149 on line in fifteen minutes. We had two servers (leased dedicateds) that requires some file system repairs before they would come back on-line, but that task was finished 30 minutes after power restoration.
What's so hard about that?
(With the addition of a three-phase power transformer, our generator is working properly.)
Customers kept calling asking us why it took us so long!
People tend to forget what launched Mr. Stallman on this road toward software freedom: he wanted to use a laser printer he had on hand with his word processing program. The software didn't have drivers, and as I recall the printer didn't have documentation, either.
Big trees from little acorns grow.
OK, for those of you who read NANAE, this is old news, but for the rest of you...
I'm a sysadmin who worked very hard to get a /24 listed in SPEWS delisted. The netblock was in the list because a customer of ours decided to provide DNS service to a known and notorious spammer. We earned the listing, period. I killed the bastard, reported the fact, and got the listing lowered to a zero, historical. In the process of doing that job, I learned a lot about the whole blocklist thing and realized that even the operators didn't see what they are really doing. They think it's about spam. Wrong.
Follow along with me a moment, and you'll see why I think this way. First, the Internet is, by definition, a "network of networks", a large anarchy run by a very large number of system administrators (greater than 10,000) who make private decisions about who and how they allow to access their bandwidth, systems, and services. The Internet Society and its sub-units provide a forum to publish community notes, the Requests for Comments, which are nothing more and nothing less than agreements for how to play nice in this employee-owned swimming pool.
The Internet community has decided on standards of behavior, and each system operator trusts every other system operator in the pool to conform to the rules of society, and to ensure that the users conform to the community rules -- not unlike CC&Rs in a neighborhood development that form part of the purchase contract of many homes and condominiums. Some operators have become lax in their expected enforcement of the rules on particularly not-nice people, the ones who break the rules in order to win money, or some other benefit. There are enough of these Internet con men out there that the community coined a word to describe them: "spammers."
Back in the NSF days, a lapse in administration resulted in disconnection, quick and swift, so the system adminstrators, up and down the line, toed the line to avoid being banished. In the Commercial Internet that replaced the NSF Internet, personal greed gets in the way of this remedy, and so the disdain of social customs is left largely unpunished by the society.
Just about every system operator who runs a mail service with more than three users has been yammered at by those users: "WE WANT LESS SPAM -- DO SOMETHING." Complaints to ISPs who take spammer money go largely ignored, and appeals "upstream" -- to the connection providers and to the Tier One networks -- have also gone largely ignored. So the small administrators started to implement mail filters and blocks on "spammy" IP addresses in the hopes that they can block the crap and thus appease their users.
Spammers countered by having their providers move them around in IP space, and by using techniques to "get around" the content filters. It's become a war, frankly. First there were keyword filters, and so spammers started to "do things" to their messages, like replace the letter 'o' with the digit '0' -- you've all seen the tricks. Hash identification of bulk messages were thwarted by inserting random nonsense text. Learning filters are poisoned by spammers injecting random words. And so on and so on. In addition to these content-based counters, spammers also steal resources of innocent people: open mail relays, open proxies, and hijacked Web scripts like formmail.pl, so that the wrong person gets blames for their flood of commercial feces.
What the block-list people decided is that having each of the 10,000 to 100,000 system administrators deal with this individually was eating up too much time, and there was this nifty thing already in place that could be used to reduce the system overhead of id
After reviewing and crafting a new AUP document for my boss at a Web hosting company, I'm beginning to appreciate how the CAN-SPAM law will get a handle on spam. Let me explain.
There is a lot more to the CAN-SPAM law than just spam. The thing that caught my eye when I read the actual Act of Congress and the law referenced by the Act is that the issue of unauthorized access to computers "used in interstate commerce" (and Web hosting falls into that class) has been clarified, and the penalties for unauthorized access have been strengthened. That means when some twerp uses one of my customer's Web sites to commit FormMail spam, it becomes a Federal matter. Falsified credentials isn't just a civil matter any more -- the Feds have criminalized the offense.
One of the biggest problems that the anti-spam crowd has had is with forgery. I have a domain, fluent (dash) access (dot) com, which has been used as a forged return address in someone's spew for a long time -- I found out when I turned the entire domain into a spam-trap and started getting bounces from all kinds of networks. (Some of those networks are now blocked because of that, but that's life on the Internet these days. Maybe I'll implement a sunset script for my automated access blocking...or maybe I won't.) With the new law, a forged From: line is criminalized. While the law is silent on the matter, forged Received From: lines may also be roped into the "deceit" intent of the law -- but that will have to be tested in a Court.
As a commercial operator of Internet services, it lets me write things into my AUP that now stand a chance of being enforced, and not necessarily on my nickel. Before CAN-SPAM, it just didn't pay to go after a guy who is probing my network looking for broken formmail.pl and formmail.cgi scripts. What CAN-SPAM really does is put the black hats on alert that their activities may attract more attention than they would be comfortable getting.
It also unshackles the programmer in me to write scripts to enforce some of these things using technological means instead of investigative means. For example, if I enforce the From: line restriction, the script person can't complain because it's an enforcement of an AUP requirement -- I'm just making sure that my customers toe the legal line.
For the DMA, it is a win, because it makes more people play by the rules. Consider that CAN-SPAM is also a win for the block-lists like SpamCop and SPEWS, because it reduces the detective work required to recognize the spew just starting from yet another IP address, and it also limits the methods available to the abusers to avoid being caught and blocked.
What you did was present the beginning of a "business relationship" with your Congressman. If my understanding of the definition of UBE is accurate (I make no such claim) anything he sends you until you say "Stop" is not UBE/Spam.
You started it, you have to end it.
As a SysAdmin I've been studying the DNS-based blocking lists in general and SPEWS in particular. Seeing how they say they operate, how long do you think it will take for the US Government to "win" an escalated listing in the SPEWS database?
"I'm sorry, Congressman, but the reason all your mail is being bounced is that our server IP address is listed in SPEWS. What is SPEWS? 'Spam Prevention Early Warning System.' Because we have been unable to answer complaints to abuse@house.net to their satisfaction, they have put together a 'crimes file' showing that The House of Representatives is a spam-lovin enterprise, have listed our entire netblock, and we've run out contractors to superserve our mail servers -- every time we hire one, it ends up listed in SPEWS, too."
Will the blocking lists work as they are supposed to, or are they going to take the smart path and NOT piss off the one organization who makes the "Laws of the Land?"? I can see it now: it becomes illegal for any operator of a mail server with more than 100 commercial clients to use any DNS- or domain-based blocking list.
Not exactly the death of the Internet, but possibly a case of felony if you do, damned if you don't.
For me, the uptime on my servers tell me how long it has been since the power company shut off the power for non-payment, or the last time some semi took out the main power feeder here in the mountains. My UPS is good for five hours, so minor outages (and four hours IS considered minor in the Sierras) don't cause things to roll back to zero.
Can you picture Jar-Jar complaining about the diodes down his left side? No, I don't think so. Too loose-jointed.
What I'm afriad of is that Marvin will be played by some muscle-bound beach-bum, so that the Plastic Pal Who Is Fun To Be With advertising sequence will "play better" to the US audience. (Remember in the BBC version, the robot playing beach-ball with the winsome girl in the not-much bikini?) Not unlike the title character in Rocky Horror, perhaps.
I doubt you are on the IP ranges from which we allow SSH connections, so even though the exploit is there the nice folks across both ponds are going to have an, er, interesting time gaining access. IPTABLES is your friend...
Hey, security is so tight that I have to make three hops to get onto one of our systems from the "outside" -- which makes remote administration interesting.
When I heard there was a second patched version last week, I said to myself that these things come in threes, and that I would wait for "the next round." So much for updating 50 boxes more than once.
Will the third time be the charm, or should I avoid being on the bleeding edge and wait for next week's discoveries?
(At least it isn't like the Microsoft patches, which come at less frequent intervals and usually do more damage to my apps than the protection is worth. -- Obligatory Microsoft Bash)
When I had a product out in the field, I had seen the "any" key problem enough that I decided not to do that. In every prompt, I specify exactly which key to depress. The documentation made it clear that a specific key was to be depressed (pressed firmly and then released quickly).
We never got a service call on that point. EVER.
We DID get one call from a technician who said that he pressed the wrong key and it seemed to work anyway. I assured him that he was just fine and to continue working with the product.
As you can guess, the CODE didn't require a specific key, but the documentation and prompt specified one. I also didn't allow type-ahead on those prompts so that there wasn't a problem with astonishing results to clueless users who didn't release the key fast enough. I also had code to wait until the keyboard stopped sending characters before putting up the prompt "Depress the C key to continue."
Yep. You forget that the RIAA is working in the civil arena, where the rules are completely different. First, the burden of proof is the "proponderance of the evidence", you don't have to have 12 out of 12 jurors -- hell, you don't have to have 12 jurors, some cases have been decided by a panel of 6 -- but 9 out of 12, and there is no such thing as the right to remain silent.
On the other hand, the only thing the RIAA can get is money.