Since it's not the politically correct point to make at/., this wasn't mentioned in michael's writeup [although it was hinted at]:
Iceland's net traffic plummets, following P2P raids
Police in Iceland raided the homes of 12 people and confiscated computer equipment and CDs this week as the global war on file sharing reached the volcanic homeland of elves and trolls. Police targeted individuals using the popular DC++ file sharing application to share movie files. One suspect was found with approximately 2.5TB of allegedly illicit material.
Within hours of the raids, net traffic in Iceland fell 40 per cent, according to SMAIS (Iceland's association of film right holders), which filed the complaints which prompted police action. Its take on the raids (in Icelandic, unfortunately) can be found here.
So 12 file-sharers were accounting for 40% of all internet traffic for an entire nation.
That's a heckuva lotta file sharing.
And within that 2.5TB of data, I wouldn't be surprised if there were some pirated software [MSDN Universal, Autocad, Acrobat/Photoshop] that might interest the BSA [or whatever they call it in Iceland].
NDS scales far, far better than OpenLDAP, has multi-master replication to provide high availability. These aren't trivial features, and have taken significant development time to get right, with thousands of hours of coding and test case development. Moreover, NDS ships in mission-critical systems as part of HP-UX... People don't generally know this, but CMS is THE Certificate authority run by the Department of Defense. That's right, DoD has many CA's installed within their organization, and every one is CMS. That's over 10 million certs issued in the last 4 years for one single deployment... Geotrust is also a huge deployment of CMS - issuing more certs than Verisign, these days... CMS supports FIPS approved hardware crypto devices... CMS is Common Criteria certified (http://niap.nist.gov/cc-scheme/vpl/vpl_type.html# cimc) to evaluation level 4 (the highest level possible). You can say what you like about Common Criteria, but the fact is that it takes considerable effort, adds value, and, moreover, is required to sell into the federal government space. CMS has huge amount of auditing capability. Not to mention that CMS is just more secure, scalable, performant, and highly-reliable than any other CA out there.
There so much more in upcoming releases.
Sounds like a lot of nice code.
So do you think Red Hat will stick to their stated principles and give it all away for free?
I didn't even realize there still was a standalone Netscape offerring. We migrated from Netscape to iPlanet to Sun Web to Sun Java One (or something like that). Anybody out there stick with the Netscape product?
This is a direct challenge to Novell/SuSE and Novell Directory Services [or eDirectory, or whatever they're calling it this week].
Red Hat must have realized that they needed a directory offering to compete in the enterprise.
That gives us four major directory vendors:
1) Novell/SuSE/Ximian/Novell Directory Services
2) Microsoft/Active Directory
3) Sun/Sun One [iPlanet]
4) RedHat/Netscape Directory Server
PS: Now that the Netscape browser has devolved into Firefox, and the enterprise stuff has been sold to Red Hat, does Netscape still exist as an independent company [other than some "portal" site on the web]?
PPS: And are there any/. CPAs who'd care to calculate AOL's return on investment from the Netscape purchase?
Last weekend, I was messing around with writing my own WAV files [in conjunction with a LabVIEW project], and, oddly enough, M$FT's wmplayer.exe was the ONLY media player that checked the file for integrity.
Real Player and that piece of crap spyware that Dell calls a media player just blithely tried to open the file without performing any integrity checks whatsoever, and damn near crashed the system.
I bet this sort of thing is a helluva lot more endemic than people realize.
Eric Schmidt named CEO at Google
Former Novell Inc. CEO Eric Schmidt has been appointed CEO of search engine company Google Inc., five months after joining the company as chairman of the board. http://www.computerworld.com/printthis/2001/0,4814,62835,00.html
Elton John helps raise money for Gore
Flamboyant rock star Elton John, making his first foray into American politics after three decades of performing in the United States, endorsed Vice President Al Gore at a ritzy Silicon Valley fund-raiser... The fund-raiser, at the home of Novell Corp. Chief Executive Eric Schmidt, raised $3.25 million for the Democratic National Committee. http://www.cnn.com/2000/ALLPOLITICS/stories/09/20/ campaign.gore.john.reut/
Buddhist Nuns Admit Destroying Documents
Two Buddhist nuns who helped coordinate an April 1996 temple fund-raiser attended by Vice President Al Gore admitted today they destroyed a list of donors and other documents because they thought the information would embarrass the temple... Gore's appearance at the fund-raiser has proven a major embarrassment for the vice president, but he also faces new Justice Department scrutiny on another front: his 46 fund-raising calls from the White House. http://www.cnn.com/ALLPOLITICS/1997/09/04/thompson/hearings.main/
Buddhist nuns indicted for failure to testify in trial of Democratic fund-raiser
Two Buddhist nuns were indicted Wednesday on contempt of court charges for failing to appear as witnesses in the trial of Maria Hsia, who was convicted in March of campaign law violations in connection with a 1996 Democratic fund-raising event at the Hsi Lai Buddhist temple in California http://www.cnn.com/2000/ALLPOLITICS/stories/04/05/ nuns.cnn/
Hsia Convicted in Campaign Finance Scandal
The jury convicted Hsia, a friend and political supporter of Vice President Al Gore, for arranging more than $100,000 in illegal donations during the 1996 presidential campaign. http://www.asianweek.com/2000_03_09/news_hsia_fina nce.html
Clinton's greatest peril isn't Monica
James Riady and his Lippo Group latched on to a young Bill Clinton and constructed a web of Asian influence that funneled millions of dollars into various Clinton campaigns and causes (such as silencing Webster Hubbell). For this, Riady enjoyed not only access to Clinton, but Riady's chief stooge, John Huang, got top-secret security clearance and continued to see classified information even after he became a big fund-raiser for the Democratic National Committee...
"Three remarkable women,'' as the authors describe them -- Democratic Party activist Maria Hsia, Pauline Kanchanalak of Thailand and Hong Kong billionaire Nina Wang -- all have money ties to Bill Clinton and Al Gore and all have connections to Chinese intelligence or the military arm of the Chinese Communist Party, the People's Liberation Army (PLA).
"Beijing did not hesitate to exploit this connection, even face-to-face with Bill Clinton,'' the authors say. Hsia is a known agent for the Chinese government who has been indicted for immigration and campaign-fund-raising scams and, say the authors, probably helped Chinese spies enter the United States.
Kanchanalak, who has been indicted on charges of violating election laws, brought leaders of a Thai conglomerate that is in business with Middle East terrorists and with China's biggest arms smugglers to t
Do we as employees have any rights (and can they dock salaried employees so easily)?
...that all men are created equal, that they are endowed by their Creator with certain unalienable rights, that among these are the RIGHTS TO THINK FOR THEMSELVES, TO STOP LOOKING TO THE GOVERNMENT TO SOLVE ALL THEIR PROBLEMS, TO QUIT WHINING, AND TO TELL THEIR EMPLOYERS TO GO fsck THEMSELVES.
Okay, exactly how hard is it to write a stable driver for an internal PCI card -vs- how hard is it to write a stable driver for an external USB device?
Or: How many times have you gotten a BSOD from an internal PCI card -vs- how many times have you gotten a BSOD from an external USB device?
And if you don't get BSODs because you use Linux, then: How easy is it to port a driver for an internal PCI card -vs- how easy is it to port a driver for an external USB device?
And if you've got a halfway decent Northbridge chipset, your USB bus traffic won't necessarily pollute your PCI bus.
Yeah, I'd like everything internal myself, but I'm also sick and tired of lousy video drivers and lousy sound card drivers crashing my system.
Linux uses *very* dumb routing algorithms at the moment. There is a fair amount of public research on much smarter mechanisms (one of which I guess FreeBSD would have used).
Thanks.
Listen - do you know of any good texts or treatises that describe these [new] mechanisms?
claims that FreeBSD can now route 1Mpps on a 2.8GHz Xeon whilst Linux can't do much more than 100kpps
Because FreeBSD has a grown-up license [i.e. a license written by grown-ups, as opposed to a license written by Peter Pan Syndrome utopian-fantasist fifty-year-old teenagers], the Linux crowd is free to "borrow" code from the BSDs [which has been pretty common in the past, e.g. I think that's how the Adaptec SCSI drivers made their way to the Linux kernel].
Anyway, I was wondering about the rules on this playground: How would you pull a stunt like that without losing face?
Research suggests that people with IBS seem to have a colon that is more sensitive and reactive than usual to a variety of things, including certain foods and stress...
The colon responds strongly to stimuli (for example, foods or stress) that would not bother most people...
In people with IBS, stress and emotions can strongly affect the colon...
The following have been associated with a worsening of IBS symptoms... drinks with caffeine, such as coffee, tea, or colas; stress, conflict, or emotional upsets
PS: As most of you know, you can't get any real work done in an atmosphere of "hypertasking," i.e. if you're trying to do physics, math, or symbolic logic, then get yourself a cabin in the mountains.
Look at the entropy of the data stream, and look at the frequency spectrum of the signal.
I ought to know enough about discrete ergodic theory to know the answer to this, but is there a simple definition of "entropy" in this situation? And how does it depend on the sampling interval? I.e. does the value of "entropy" change as the sampling interval increases, from 1 second, to 2 seconds, to..., to infinity?
Similar problems with "the frequency spectrum" - just trying to decide on a time domain sampling interval [prior to applying a Fourier Transform] is a whale of a problem in and of itself.
As an aside, I'd ask you the same questions I asked another poster: Doesn't a Pulsating Radio Source [PULSAR] have a nice, stable, utterly utterly predictable spectrum? Similarly, doesn't 120V/60Hz [cf 240V/50Hz] wall current have a nice, stable, utterly predictable spectrum? Now which is interesting, and which is uninteresting? And does either one of them transmit a "message"? Or are they more akin to noise? By the way, we get this very sort of problem in our lab - 60Hz wall current pollutes our signals, but, mathematically, the pollution is very similar to e.g. the kinds of pollution we get from mammalian respiration motions, so we have to decide things like "respiration, or wall current"? And until very recently we what I had presumed to be noise generated by bad leads or metal touching metal, until I realized that it was emanating from a heating lamp that we use to keep mammalian temperatures at 37C.
Anyway, I've been thinking about the possibility of using Wavelets and Mother Functions that are specifically tailored for each situation, but that brings me to the next point:
Most noise sources have very high entropy, and spectra that fall into a few well-defined shapes.
What are these "few well-defined shapes" [again, presumably relative to which "well-defined sampling intervals"]? The presumption with "well-defined shapes" seems to be that you've studied a plethora of known noise generators, and their known spectra [relative to some agreed-upon sampling interval], and have catalogued them. [And, for the record, if anyone knows of some standard text or treatise which presents this catalogue, I'd love to know about it. Thanks!] But the assumption with a communication from some random, hitherto unknown [but intelligent] alien species must be that you do NOT know what form their communications protocols will take and what "shapes" the associated spectra will comprise.
Signals (or rather, symbol streams in a decoded signal) that are intended to convey language have their own very-recognizable statistical patterns as well.
Again, could you clue me in as to these "very-recognizable statistical patterns"? For instance, I'm familiar with the Shakespeare authorship question, and the use of word pattern analysis to try to determine whether Edward de Vere, Earl of Oxford, was actually writing under the pseudonym "William Shakespeare" [turns out he doesn't fare very well in the pattern-recognition tests, whereas Queen Elizabeth, of all people, fares quite well], but again, in that instance, you're taking known phenomena [the surviving works of "Shakespeare" and the surviving letters of de Vere, or of Elizabeth] and analyzing them for patterns, whereas in the case of transmissions from outerspace, you're trying to "analyze" almost infinitely many EM point sources [there are what, 100 billion stars in the Milky Way alone?] and trying to both discard the ones that are noise and, at the same time, zero in on the ones that contain an UNKNOWN but intelligent language.
Again, if it were so easy to decode these things, why is it that some of the brightest minds in the American NSA and the British MI5/MI6 still can't decode the overwhelming majority of the Venona traffic, even though the agents being monitored were speaking in KNOWN languages, such as English and Russian?
For that matter, how easy would it be for us to decode ASCII
Finally, with all that checked, someone might try to see if the radio waves are transmitting an actual message
I've been spending a lot of time lately with signal analysis - high speed sampling of analog signals [i.e. high bit-rate digitization of analog phenomena] and the subsequent software analysis of the digitized sequences - and it strikes me as a really, really difficult [maybe impossible?] problem trying to decide if a stream of 0's and 1's is just noise, or if it represents some phenomenon worth paying attention to.
Consider something as simple as "Hello World!"
In a word processor [or a web browser] that is capable of recognizing ASCII, it looks like
Pretty soon, if you don't know what you're looking for, this stuff becomes indistinguishable from gibberish. ["Ewww - the high bits are all zero!!!" Well, congratulations Einstein - now tell me what the other seven bits mean.]
History's Verdict
Victor Davis Hanson
July 16, 2004
About this time 60 years ago, six weeks after the Normandy beach landings, Americans were dying in droves in France. We think of the 76-day Normandy campaign of summer and autumn 1944 as an astounding American success -- and indeed it was, as Anglo-American forces cleared much of France of its Nazi occupiers in less than three months. But the outcome was not at all preordained, and more often was the stuff of great tragedy. Blunders were daily occurrences -- resulting in 2,500 Allied casualties a day. In any average three-day period, more were killed, wounded, or missing than there have been in over a year in Iraq...
The army soon learned that their light Sherman tanks were no match for Nazi Panthers and Tigers. Hundreds of their "Ronson-lighters" -- crews and all -- went up in smoke. Indeed, 60 percent of all lost Shermans were torched by single shots from enemy Panzers. In contrast, only one in three of the Americans' salvos even penetrated German armor...
So -- what OS is better suited to this kind of application?
Hate to play the role of the troll here [better say goodbye to those Karma points], but Linux has pretty lousy numbers for an embedded OS. Heck - "pretty lousy" is being generous - I've never seen Linux come in anywhere but dead last in an RTOS review.
In fact, the dirty little secret of the embedded OS marketplace is that WinCE is a rather solid, stable, and flexible platform:
What's more, full-blown x86 Linux ain't exactly the cat's meow, either. People laugh at M$FT for their problems with WinFS & Longhorn, but that rusting, archaic, monolithic kernel, called "Linux," is a disaster waiting to happen.
Of course, some of the problem here may be semantics - people seem to think that if you build a SBC/PICMG platform, load it with "Linux," and call it a "firewall," then you're doing realtime work. Well, guess what - you're not. Realtime is an OS in a USAF jet, flying at Mach 3, reacting to a gazillion interrupts per second, trying to keep the pilot from both killing himself and from being killed by that SAM missile on his tail.
That's when you call in the grown-ups and the grown-up RTOSes.
Anyway, you can argue about what the words "embedded" and "realtime" mean, but, when the rubber hits the road, Linux is a very poor substitute for the real thing.
I thought about bidding on some VA units on Ebay [I'm sure the dude still has them up for sale].
Again, no online documentation. Fortunately, the guy was kind enough to go open a case and give me the motherboard part number [kudos to him for doing that].
Turns out VA just took a basic Intel boxed motherboard, with six PCI slots, slapped it into a two unit rackmount, and put in a riser card to give you [at least theoretical] access to two ["risen"] PCI slots.
Anyway, to make a long story short, in my decision to purchase or not purchase, I just used the Intel site for documentation, since VA was, in essence, just serving as one big Intel reseller.
We think the closest thing to this is ZenWorks, which we have in the longterm plans of implementing when upgrading of our servers are done.
Sounds like you answered your own question.
Of course, ZenWorks ain't free, and it'll require training, and NDS tree rights will introduce political turf fights, and eventually some pointy-headed bean counter will object to the cost, but hey, welcome to life in IT.
If you're writing code that must not possibly be contaminated by the GPL, and if you're writing that code for a Linux platform, then I guess you use either of the Intel or Metrowerks [Motorola] compilers [& their respective libraries].
But if you're writing for one of the BSDs [e.g. FreeBSD], what compiler [& libraries] do you use? Obviously you can't touch the GCC with a ten foot pole.
There's plenty of technology that already does that sort of thing (relational databases and OLTP being some examples); now we're just talking about bringing that down to a level that individuals can exploit better.
Do you have any idea how much these things cost? There's a reason that Oracle begins at about $100,000 per site - because it's damned difficult to get these things to be halfway stable, and it requires a small army of CompSci PhDs to write the code, debug it, and regression test the patches.
Not to mention the DBAs that have to try to keep the thing up and running. And if suddenly you've got secretaries and clerks who are contributing to ongoing, collaborative documents, then they're gonna need training that will begin to resemble the training required to become a DBA.
These things are hard, dude, and very, very expensive. Obviously the Holy Grail of subscription-based computing services is to make these things easy for the end-user, but remember: The "end-user" is the kind of clown who spends an hour looking for the "Any Key".
Many programs need to work on operating in a collaborative environment.
Do you have any idea how very nearly impossibly difficult this sort of thing is? It makes The Theory of Relativity look like a stroll on the beach.
Indeed, the sorts of problems encountered [when concepts like "TRUE" and "FALSE" cease to have meanings independent of their times and places] bear more than a passing resemblance to The Theory of Relativity.
Think I'm kidding? Try reading the RFC for the Network Time Protocol:
All that NTP seeks to do is get two computers to engage in the most fundamental task of computing: Come to some reasonable agreement as to the time. And yet, the RFC requires just about a PhD in mathematics and about 1000 pages of background reading from old AT&T switching standards just to begin to get an idea of what the heck is going on.
Slashdot Mirror
Since it's not the politically correct point to make at
That's a heckuva lotta file sharing.
And within that 2.5TB of data, I wouldn't be surprised if there were some pirated software [MSDN Universal, Autocad, Acrobat/Photoshop] that might interest the BSA [or whatever they call it in Iceland].
NDS scales far, far better than OpenLDAP, has multi-master replication to provide high availability. These aren't trivial features, and have taken significant development time to get right, with thousands of hours of coding and test case development. Moreover, NDS ships in mission-critical systems as part of HP-UX... People don't generally know this, but CMS is THE Certificate authority run by the Department of Defense. That's right, DoD has many CA's installed within their organization, and every one is CMS. That's over 10 million certs issued in the last 4 years for one single deployment... Geotrust is also a huge deployment of CMS - issuing more certs than Verisign, these days... CMS supports FIPS approved hardware crypto devices... CMS is Common Criteria certified (http://niap.nist.gov/cc-scheme/vpl/vpl_type.html
Sounds like a lot of nice code.
So do you think Red Hat will stick to their stated principles and give it all away for free?
I didn't even realize there still was a standalone Netscape offerring. We migrated from Netscape to iPlanet to Sun Web to Sun Java One (or something like that). Anybody out there stick with the Netscape product?
This is a direct challenge to Novell/SuSE and Novell Directory Services [or eDirectory, or whatever they're calling it this week].
Red Hat must have realized that they needed a directory offering to compete in the enterprise.
That gives us four major directory vendors:
PS: Now that the Netscape browser has devolved into Firefox, and the enterprise stuff has been sold to Red Hat, does Netscape still exist as an independent company [other than some "portal" site on the web]?PPS: And are there any /. CPAs who'd care to calculate AOL's return on investment from the Netscape purchase?
Real Player and that piece of crap spyware that Dell calls a media player just blithely tried to open the file without performing any integrity checks whatsoever, and damn near crashed the system.
I bet this sort of thing is a helluva lot more endemic than people realize.
Eric Schmidt named CEO at Google
Former Novell Inc. CEO Eric Schmidt has been appointed CEO of search engine company Google Inc., five months after joining the company as chairman of the board.
http://www.computerworld.com/printthis/2001/0,4814
Elton John helps raise money for Gore
Flamboyant rock star Elton John, making his first foray into American politics after three decades of performing in the United States, endorsed Vice President Al Gore at a ritzy Silicon Valley fund-raiser... The fund-raiser, at the home of Novell Corp. Chief Executive Eric Schmidt, raised $3.25 million for the Democratic National Committee.
http://www.cnn.com/2000/ALLPOLITICS/stories/09/20/ campaign.gore.john.reut/
Buddhist Nuns Admit Destroying Documents /hearings.main/
Two Buddhist nuns who helped coordinate an April 1996 temple fund-raiser attended by Vice President Al Gore admitted today they destroyed a list of donors and other documents because they thought the information would embarrass the temple... Gore's appearance at the fund-raiser has proven a major embarrassment for the vice president, but he also faces new Justice Department scrutiny on another front: his 46 fund-raising calls from the White House.
http://www.cnn.com/ALLPOLITICS/1997/09/04/thompson
Buddhist nuns indicted for failure to testify in trial of Democratic fund-raiser
Two Buddhist nuns were indicted Wednesday on contempt of court charges for failing to appear as witnesses in the trial of Maria Hsia, who was convicted in March of campaign law violations in connection with a 1996 Democratic fund-raising event at the Hsi Lai Buddhist temple in California
http://www.cnn.com/2000/ALLPOLITICS/stories/04/05/ nuns.cnn/
Hsia Convicted in Campaign Finance Scandal
The jury convicted Hsia, a friend and political supporter of Vice President Al Gore, for arranging more than $100,000 in illegal donations during the 1996 presidential campaign.
http://www.asianweek.com/2000_03_09/news_hsia_fina nce.html
Clinton's greatest peril isn't Monica
James Riady and his Lippo Group latched on to a young Bill Clinton and constructed a web of Asian influence that funneled millions of dollars into various Clinton campaigns and causes (such as silencing Webster Hubbell). For this, Riady enjoyed not only access to Clinton, but Riady's chief stooge, John Huang, got top-secret security clearance and continued to see classified information even after he became a big fund-raiser for the Democratic National Committee...
"Three remarkable women,'' as the authors describe them -- Democratic Party activist Maria Hsia, Pauline Kanchanalak of Thailand and Hong Kong billionaire Nina Wang -- all have money ties to Bill Clinton and Al Gore and all have connections to Chinese intelligence or the military arm of the Chinese Communist Party, the People's Liberation Army (PLA).
"Beijing did not hesitate to exploit this connection, even face-to-face with Bill Clinton,'' the authors say. Hsia is a known agent for the Chinese government who has been indicted for immigration and campaign-fund-raising scams and, say the authors, probably helped Chinese spies enter the United States.
Kanchanalak, who has been indicted on charges of violating election laws, brought leaders of a Thai conglomerate that is in business with Middle East terrorists and with China's biggest arms smugglers to t
Slashdot Goes Political: Announcing politics.slashdot.org
Ok, exactly _how_ hard is it to open your case.
Okay, exactly how hard is it to write a stable driver for an internal PCI card -vs- how hard is it to write a stable driver for an external USB device?
Or: How many times have you gotten a BSOD from an internal PCI card -vs- how many times have you gotten a BSOD from an external USB device?
And if you don't get BSODs because you use Linux, then: How easy is it to port a driver for an internal PCI card -vs- how easy is it to port a driver for an external USB device?
And if you've got a halfway decent Northbridge chipset, your USB bus traffic won't necessarily pollute your PCI bus.
Yeah, I'd like everything internal myself, but I'm also sick and tired of lousy video drivers and lousy sound card drivers crashing my system.
Linux uses *very* dumb routing algorithms at the moment. There is a fair amount of public research on much smarter mechanisms (one of which I guess FreeBSD would have used).
Thanks.
Listen - do you know of any good texts or treatises that describe these [new] mechanisms?
claims that FreeBSD can now route 1Mpps on a 2.8GHz Xeon whilst Linux can't do much more than 100kpps
Because FreeBSD has a grown-up license [i.e. a license written by grown-ups, as opposed to a license written by Peter Pan Syndrome utopian-fantasist fifty-year-old teenagers], the Linux crowd is free to "borrow" code from the BSDs [which has been pretty common in the past, e.g. I think that's how the Adaptec SCSI drivers made their way to the Linux kernel].
Anyway, I was wondering about the rules on this playground: How would you pull a stunt like that without losing face?
Research suggests that people with IBS seem to have a colon that is more sensitive and reactive than usual to a variety of things, including certain foods and stress...
The colon responds strongly to stimuli (for example, foods or stress) that would not bother most people...
In people with IBS, stress and emotions can strongly affect the colon...
The following have been associated with a worsening of IBS symptoms... drinks with caffeine, such as coffee, tea, or colas; stress, conflict, or emotional upsets
http://digestive.niddk.nih.gov/ddiseases/pubs/ibs/ #whatcauses
PS: As most of you know, you can't get any real work done in an atmosphere of "hypertasking," i.e. if you're trying to do physics, math, or symbolic logic, then get yourself a cabin in the mountains.
Look at the entropy of the data stream, and look at the frequency spectrum of the signal.
I ought to know enough about discrete ergodic theory to know the answer to this, but is there a simple definition of "entropy" in this situation? And how does it depend on the sampling interval? I.e. does the value of "entropy" change as the sampling interval increases, from 1 second, to 2 seconds, to ..., to infinity?
Similar problems with "the frequency spectrum" - just trying to decide on a time domain sampling interval [prior to applying a Fourier Transform] is a whale of a problem in and of itself.
As an aside, I'd ask you the same questions I asked another poster: Doesn't a Pulsating Radio Source [PULSAR] have a nice, stable, utterly utterly predictable spectrum? Similarly, doesn't 120V/60Hz [cf 240V/50Hz] wall current have a nice, stable, utterly predictable spectrum? Now which is interesting, and which is uninteresting? And does either one of them transmit a "message"? Or are they more akin to noise? By the way, we get this very sort of problem in our lab - 60Hz wall current pollutes our signals, but, mathematically, the pollution is very similar to e.g. the kinds of pollution we get from mammalian respiration motions, so we have to decide things like "respiration, or wall current"? And until very recently we what I had presumed to be noise generated by bad leads or metal touching metal, until I realized that it was emanating from a heating lamp that we use to keep mammalian temperatures at 37C.
Anyway, I've been thinking about the possibility of using Wavelets and Mother Functions that are specifically tailored for each situation, but that brings me to the next point:
Most noise sources have very high entropy, and spectra that fall into a few well-defined shapes.
What are these "few well-defined shapes" [again, presumably relative to which "well-defined sampling intervals"]? The presumption with "well-defined shapes" seems to be that you've studied a plethora of known noise generators, and their known spectra [relative to some agreed-upon sampling interval], and have catalogued them. [And, for the record, if anyone knows of some standard text or treatise which presents this catalogue, I'd love to know about it. Thanks!] But the assumption with a communication from some random, hitherto unknown [but intelligent] alien species must be that you do NOT know what form their communications protocols will take and what "shapes" the associated spectra will comprise.
Signals (or rather, symbol streams in a decoded signal) that are intended to convey language have their own very-recognizable statistical patterns as well.
Again, could you clue me in as to these "very-recognizable statistical patterns"? For instance, I'm familiar with the Shakespeare authorship question, and the use of word pattern analysis to try to determine whether Edward de Vere, Earl of Oxford, was actually writing under the pseudonym "William Shakespeare" [turns out he doesn't fare very well in the pattern-recognition tests, whereas Queen Elizabeth, of all people, fares quite well], but again, in that instance, you're taking known phenomena [the surviving works of "Shakespeare" and the surviving letters of de Vere, or of Elizabeth] and analyzing them for patterns, whereas in the case of transmissions from outerspace, you're trying to "analyze" almost infinitely many EM point sources [there are what, 100 billion stars in the Milky Way alone?] and trying to both discard the ones that are noise and, at the same time, zero in on the ones that contain an UNKNOWN but intelligent language.
Again, if it were so easy to decode these things, why is it that some of the brightest minds in the American NSA and the British MI5/MI6 still can't decode the overwhelming majority of the Venona traffic, even though the agents being monitored were speaking in KNOWN languages, such as English and Russian?
For that matter, how easy would it be for us to decode ASCII
becomes dumbly obvious as a message when it loops over and over and over again
Doesn't a Pulsating Radio Source [PULSAR] loop over and over again?
By contrast, doesn't 120V/60Hz [cf 240V/50Hz] wall current loop over and over again?
Now tell me which is interesting and which is uninteresting. And does either one of them transmit a "message"? Or are they both more akin to noise?
Finally, with all that checked, someone might try to see if the radio waves are transmitting an actual message
I've been spending a lot of time lately with signal analysis - high speed sampling of analog signals [i.e. high bit-rate digitization of analog phenomena] and the subsequent software analysis of the digitized sequences - and it strikes me as a really, really difficult [maybe impossible?] problem trying to decide if a stream of 0's and 1's is just noise, or if it represents some phenomenon worth paying attention to.
Consider something as simple as "Hello World!"
In a word processor [or a web browser] that is capable of recognizing ASCII, it looks like
When we switch to base ten, we get and, in base 2, Pretty soon, if you don't know what you're looking for, this stuff becomes indistinguishable from gibberish. ["Ewww - the high bits are all zero!!!" Well, congratulations Einstein - now tell me what the other seven bits mean.]It's my understanding that, to this day [i.e. some fifty years later], the vast, overwhelming majority of Venona traffic still hasn't been decoded: Out of some hundreds of thousands of intercepted cyphertexts, it is claimed that under 3000 have been partially or wholly decrypted.
So my question: Are there any standard texts or treatises on the theory of how to distinguish interesting signals from large amplitude noise?
History's Verdict
Victor Davis Hanson
July 16, 2004
About this time 60 years ago, six weeks after the Normandy beach landings, Americans were dying in droves in France. We think of the 76-day Normandy campaign of summer and autumn 1944 as an astounding American success -- and indeed it was, as Anglo-American forces cleared much of France of its Nazi occupiers in less than three months. But the outcome was not at all preordained, and more often was the stuff of great tragedy. Blunders were daily occurrences -- resulting in 2,500 Allied casualties a day. In any average three-day period, more were killed, wounded, or missing than there have been in over a year in Iraq...
The army soon learned that their light Sherman tanks were no match for Nazi Panthers and Tigers. Hundreds of their "Ronson-lighters" -- crews and all -- went up in smoke. Indeed, 60 percent of all lost Shermans were torched by single shots from enemy Panzers. In contrast, only one in three of the Americans' salvos even penetrated German armor...
http://victorhanson.com/articles/hanson071604.htm
http://www.nationalreview.com/script/printpage.asp ?ref=/hanson/hanson200407160827.asp
So -- what OS is better suited to this kind of application?
Hate to play the role of the troll here [better say goodbye to those Karma points], but Linux has pretty lousy numbers for an embedded OS. Heck - "pretty lousy" is being generous - I've never seen Linux come in anywhere but dead last in an RTOS review.
In fact, the dirty little secret of the embedded OS marketplace is that WinCE is a rather solid, stable, and flexible platform:
What's more, full-blown x86 Linux ain't exactly the cat's meow, either. People laugh at M$FT for their problems with WinFS & Longhorn, but that rusting, archaic, monolithic kernel, called "Linux," is a disaster waiting to happen.Of course, some of the problem here may be semantics - people seem to think that if you build a SBC/PICMG platform, load it with "Linux," and call it a "firewall," then you're doing realtime work. Well, guess what - you're not. Realtime is an OS in a USAF jet, flying at Mach 3, reacting to a gazillion interrupts per second, trying to keep the pilot from both killing himself and from being killed by that SAM missile on his tail.
That's when you call in the grown-ups and the grown-up RTOSes.
Anyway, you can argue about what the words "embedded" and "realtime" mean, but, when the rubber hits the road, Linux is a very poor substitute for the real thing.
I thought about bidding on some VA units on Ebay [I'm sure the dude still has them up for sale].
Again, no online documentation. Fortunately, the guy was kind enough to go open a case and give me the motherboard part number [kudos to him for doing that].
Turns out VA just took a basic Intel boxed motherboard, with six PCI slots, slapped it into a two unit rackmount, and put in a riser card to give you [at least theoretical] access to two ["risen"] PCI slots.
Anyway, to make a long story short, in my decision to purchase or not purchase, I just used the Intel site for documentation, since VA was, in essence, just serving as one big Intel reseller.
We think the closest thing to this is ZenWorks, which we have in the longterm plans of implementing when upgrading of our servers are done.
Sounds like you answered your own question.
Of course, ZenWorks ain't free, and it'll require training, and NDS tree rights will introduce political turf fights, and eventually some pointy-headed bean counter will object to the cost, but hey, welcome to life in IT.
Or maybe a Swedish meatball?
Sweden - hmmm... Hey, you don't work on that Nokia platform do ya' - whadda they call it, the Sybian?
All of the BSDs use GCC as the default compiler.
Obviously. But what does it link against???
Is "Hello World" contamination-free if you compile it with GCC on a FreeBSD platform?
It's a serious question; I'd appreciate serious answers.
[where BSDs != BSoDs]
If you're writing code that must not possibly be contaminated by the GPL, and if you're writing that code for a Linux platform, then I guess you use either of the Intel or Metrowerks [Motorola] compilers [& their respective libraries].
But if you're writing for one of the BSDs [e.g. FreeBSD], what compiler [& libraries] do you use? Obviously you can't touch the GCC with a ten foot pole.
I was watching the movie with my nephew [he's four].
I don't think he quite understood what happened, but if he had, I think he would have been devastated.
The URL you just referenced lives at nonstandard port 81: Does the Google spider look for nonstandard ports?
There's plenty of technology that already does that sort of thing (relational databases and OLTP being some examples); now we're just talking about bringing that down to a level that individuals can exploit better.
Do you have any idea how much these things cost? There's a reason that Oracle begins at about $100,000 per site - because it's damned difficult to get these things to be halfway stable, and it requires a small army of CompSci PhDs to write the code, debug it, and regression test the patches.
Not to mention the DBAs that have to try to keep the thing up and running. And if suddenly you've got secretaries and clerks who are contributing to ongoing, collaborative documents, then they're gonna need training that will begin to resemble the training required to become a DBA.
These things are hard, dude, and very, very expensive. Obviously the Holy Grail of subscription-based computing services is to make these things easy for the end-user, but remember: The "end-user" is the kind of clown who spends an hour looking for the "Any Key".
Many programs need to work on operating in a collaborative environment.
Do you have any idea how very nearly impossibly difficult this sort of thing is? It makes The Theory of Relativity look like a stroll on the beach.
Indeed, the sorts of problems encountered [when concepts like "TRUE" and "FALSE" cease to have meanings independent of their times and places] bear more than a passing resemblance to The Theory of Relativity.
Think I'm kidding? Try reading the RFC for the Network Time Protocol:
All that NTP seeks to do is get two computers to engage in the most fundamental task of computing: Come to some reasonable agreement as to the time. And yet, the RFC requires just about a PhD in mathematics and about 1000 pages of background reading from old AT&T switching standards just to begin to get an idea of what the heck is going on.