You have an excellent point. My first drafts, first versions, are not intended for external view. Note, however, that when I do release something to "the world" (usually co-workers), I've made sure to dot the I's and cross the T's. This includes making sure that documents contain what I intended. The author of the Word document in question did not finish the job.
The embarrassment here is that someone with serious responsibility failed to finish the job correctly.
Back then, we could only hope that others with greater responsibility were doing their jobs. As far as I can tell from the news, there are too many cases where people did a slapdash job, with "life changing effects" for too many. A nice euphemism there: "Life changing"... like "life ending", or "no more legs", or "no more father", things like that.
This stuff is very very important to large numbers of people. This is one of those examples where we expect people to get the bloody details right.
A better idea is to provide the utility with a list of suspect numbers and receive notification of change of state in those accounts. Such state changes include received calls, dialed calls, forwarded calls, et cetera.
The point is to provide law enforcement entities with all information relevant to suspects that have received judicial review of probable cause.
If we're going to track things, the least we can do is filter them for relevancy. In this case, my disagreement with Verizon (and AT&T, who has also been entirely too cooperative with this exceptional monitoring) is that they are not filtering the content for relevancy to the actual suspects.
I remain committed to the idea that "they" (government) should required to submit requests to invade "our" privacy for a theoretically disinterested judicial review. Upon receiving that permission, the technical means is available for providing all relevant information to the monitoring law enforcement entities. All I would ask from our service providers is the filtering for relevance to those whose activities have passed the judicial test of probable cause.
I assume that as a commenter on Slashdot, you are aware of the basic technologies involved in this discussion, and therefore are aware that event filtering is technically feasible. (I will provide a resume to you, including the 8 years that I worked at AT&T BL developing SS7 switching infrastructure, if you really need to get into the ability of modern systems to provide basic filtering.)
I see no place in my comment where the concept of "currently committing a crime" appears in the discussion. That statement is a simple red herring and is irrelevant to the argument. The issue is simply judicial review of probable cause, and while there are likely subtleties to the legal arguments, we still have the fourth amendment. Showing probable cause to justify looking into what are otherwise private communications is a central part of it.
...It's monitoring *everyone*. The point to the tracking program was to note the originating and dialed numbers for *all* conversations, not merely those between suspects and the rest of the world. Furthermore, the whole argument from the beginning is that FISA provides for getting permission to monitor up to 72 hours from the start of the monitoring process.
FISA is intended to provide *exactly* the flexibility required to enable surveillance responsive to changing conditions (the genesis of the 72-hour provision), while still requiring the judicial review that is part of the fourth amendment's requirement of showing probable cause.
And I agree with other commenters that customer transaction records (be they phone calls, or reporting on who bought what groceries for how much) is by no stretch of the imagination "protected free speech".
To the best of my knowledge PJ was a paralegal, has never been a lawyer, and therefore is not an officer of the court. Furthermore, whatever she was before 2003, she is now unambiguously a blogger. Neither more nor less.
PJ has been flawlessly consistent in her trust in the U.S. legal system. She says that it can produce imperfect results, as can any human-created system, but it has a strong tendency to work in the right direction.
In the SCO case, PJ has in her own inimitable way, contributed to this case going the right way: She brings the actinic glare of illumination onto a process that SCO, and others, have tried to accomplish.
PJ started her blog several months before the SCO case became public knowledge, in order to connect geeks with legal concepts. She believed then and still believes that we (the technologists) should be aware of the rules of the game in order avoid being steamrolled by it.
I've been reading the blog since early 2003, and PJ sounds like a gentle spoken (she insists on decorum) but very very intelligent paralegal. The truth justice and the American way language is backed up by consistent and finely honed research and argument. Plus one additional ingredient, a very angry hornet's nest: 10,000+ pissed off developers, some of whom have been around from the inception of UNIX and derived technologies. PJ, more than anything else, has brought us together to face the threat from SCO and folks like them, and for that alone her contribution is inestimable.
The original Eclipse was a highly glorified Nova, both being 16-bit machines. The new machine that Tracy Kidder wrote about was actually the MV/8000 Eagle 32 bit system.
We (meaning the management at Warner-Amex Cable Communccations, Inc, A.K.A. WACCI) hired several D.G. systems guys around 1980/1981 to help grow the network, which was growing to include Cincinnati, Pittsburgh and a few other cities that I don't remember. We decided to run AOS on a few of the Eclipse C300's (upgraded with COBOL-supporting character instructions through microcode). We were introduced to AOS by some *very* cool D.G. guys from Massachusetts. At the ripe old age of 23, I was *enthralled* with AOS. We received some of the earliest MV/8000 Eagle systems that D.G. produced, and ran AOS/VS (AOS "virtual system") on them. In our view, these machines kicked Major Ass.
AOS and AOS/VS were architected to present a full (virtual) R/DOS machine environment to each of their processes, so all the organizations using INFOS, IDEA, etc. (D.G.'s "database" and online multiterminal system) could move forward into an AOS (16-bit) and AOS/VS (32-bit) environment with minimal hassle. So, for example, the tasks in a classical foreground/background Nova/Eclipse RDOS system were treated as tasks contained with the many processes running in the new AOS and AOS/VS systems.
The result was that AOS/VS gave us multithreaded technology in 1981. And we used it to great benefit to keep our online systems running as quickly as possible. (I was able to write an INFOS-table caching subsystem that doubled the online capacity of each MV/8000 system, in about 1000 lines of PL/I).
The architecture of AOS and AOS/VS were derived from MULTICS, and in my opinion were very successful in achieving their goals. It would be 10 years before the UNIX world would catch up, when Sun Microsystems brought in their lightweight process technology. (I don't know whether other developers had threads going, it's merely that my experience was with Sun.)
Those machines were comparative monsters for us, with 4 Mbyte RAM, 700 Mbyte disk, and processors based on AMD bit-slice technology. (Anyone remember the AMD 29xx bit-slice processor building blocks? 4 bits of ALU per chip, supported by carry lookahead; with creative support from the diskette-loaded microcode, these machines were *wonderful* to work with.)
All this stuff reminds me that this is the same AMD that eventually built the Athlon X64 in my 3-year-old inexpensive workstation. It has 2 Gbytes of RAM, 500 gbytes of disk, running Linux. Not dual-core (yet:-)... give me a few months to afford it, I've been laid-off for several years) but when I compare my little Linux workstation to those *immense beasts*, I get back some of that old feeling that I used to have in my youth.
My first job out of school. Very cool place. Their polling system consisted of a stack of Data General Nova single board computers, each responsible for polling one supertrunk. They were supervised by a Data General Eclipse (the polling system), which had aggregate responsibility for the entire system.
There was a separate Eclipse, the "Studio System", which used a high speed interprocessor bus to move polling data to and from the polling system.
I wrote several of the studio system's technical scripts, which needed to be synchronized with the TV shows.
QUBE flopped as a technology due mostly to the fact that people are (and were in the late 70's) in the habit of being couch potatoes, rather than interacting through a rather stilted 2-way system.
QUBE gave two-way cable communications hardware people some pretty good practice in how to run signals both ways through a hierarchical network. Eventually, (with huge improvements, etc.) it led to today's cable modems.
A cute cultural story: The two-way boxes were designed by Pioneer Electronics (the stereo folks) in Japan. The Japanese engineers had absolutely no idea how quickly Americans would learn to hack the boxes to watch pay-per-view premium content without the box reporting that they had selected premium channels. It turns out that the box was designed to detect channel change events and track the changes, rather than reporting the channel that was currently selected for viewing. The result was that as soon as someone discovered how to disable the change detection logic (with a paper clip), they started watching premium content for free.
The business management folks had me write a program that statistically analyzed premium purchasing habits, noting (for example) when a given customer transitioned from several months of reasonable amount of premium content, to absolutely zero premium viewing. The program was called "zerobill". Naturally, its capabilities grew in various ways to track a whole range of statistics about viewing habits during the next few years. Eventually, zerobill became *the report* that every manager wanted to see, every morning without fail. I had some *exceptionally early* mornings caused by various bugs and vicissitudes in the database.
Phone rings... Me: (knowing damn well what was coming next) Hello? Night operator: "Daily batch died." Me: "and..." Night operator: "Not sure, it looks like an error." Me: "Did it leave a suicide note, or was it just shot in the head?" et cetera...
My best friend and I were not scheduled the evening of the Rundgren concert, and we had a *kickass* time at the concert, including a little while backstage. It was a great time and place to be a young software geek, mixing television and technology.
Pushing some patch 30 minutes later for an OSS package that 2000 rag tag home users use.. just isn't the same.
2000 ragtag home users? You are smarter than that, I can tell by the quality of your writing and sentence structure alone. While some OSS packages serve small communities, there are lots of packages that serve large and diverse communities. (PostgreSQL, Apache, the Linux kernel, Firefox, the list goes on). Those packages have, on occasion introduced vulnerabilities due to the natural vicissitudes of software development. And when their vulnerabilities are discovered, they get fixed quickly. (And this one hit me this morning: I don't need Linux Genuine Advantage for permission to receive updates to my damn software!!!)
It is worth noting, however, that such vulnerabilities are nearly always limited in scope due the inherently modular nature of the OSS world. Microsoft built a highly integrated system to support its business model. They are welcome to their high integration approach. And those of use who do not appreciate the effects of that way of doing business are welcome to complain when it wacks the shit out of our families' productivity when we are trying to get some proprietary fix.
Because prostate problems do not affect reproduction, and do not even affect a male's ability to support his progeny until they are essentially ready for breeding themselves.
After writing 10k+ lines of code on the PDP-8, I can tell you that there's no argument that could convince me that it was a better architecture than the 8088/8086.
One could almost sustain an argument that the PDP-11 was a better architecture than the 8088/8086 as far as it went, which at the time could be extended to a Mbyte of physical RAM (later instances of the -11 could get all the way to 4 Mbytes physical, as in the 11/73). Manipulating the APRs (active page registers), one could do some mild hackery to make a single process address more than 64kbytes.
Through it would have changed computing history fairly dramatically, I believe the MC68008, an 8-bit-bus variant of the MC68000 would have been a much better choice for IBM's new play-toy in the long run. The reason is the 68k began its design life with 32-bit registers. (IBM was still king of the mainframe hill with their mainframes back then, and they might be forgiven for not taking this new toy-sized computer seriously.)
But I will not second-guess IBM, since their inspiration was the Apple][ and they would have to be very forward-looking indeed to apply Moore's law 10+ years into the future.
The 8080's registers can be mapped directly from the 8008's registers: A, B, C, D, E, H, L. HL taken as a pair could be used to address anywhere in the 8008's 14-bit address space.
Intel's promise that the 8086 had upgrade compatibility with 8080 (or the 8085), came straight from the success of the 8008->8080 transition.
All this talk about ancient processors... I wrote a debugger for an 8008 system in 1976. A collection of geeks in college borrowed an 8086 development kit from the local Bell Labs office, and we did it with an ASR-33 Teletype. I forgot to return the 8008 manual, and still have it. The *fast version* ran at 200 KHz, and each instruction executed in a multiple of 5 clocks (single-byte instructions in 5 clock cycles, two-byte instructions in 10 clock cycles...)
I remember marveling a few years ago that one could trace the 8008's registers and capabilities right up through the latest Pentium4.
Your argument basically assumes that there will be no checking done.
Earlier this year, the state of Ohio did exactly that. House Bill 3 as amended by the Ohio Senate removed *all* automatic audits of electronic voting systems. This means that unless recounts are requested and paid (a slick new higher price) for, there is no examination of the paper trail.
Whether they are secure or not, voting machines DO NOT SCALE. They introduce the possibility of "unintentional" misallocation. This translates to under-allocation of machines to precincts in which the current powers-that-be do not want high turnout.
Voting should be performed with two characteristics:
Scalable (Lots of people can fill out paper ballots in parallel)
Countable, in full public view, by UNAUGMENTED humans.
Even without the possibility of denial of voting-service attacks, there is the issue of any real security being too expensive to justify. Anyone who has done flight-critical software development knows that the cost of that development is waay higher than the ordinary system design and implementation practices. Voting systems are far closer to hard-realtime systems than most people expect.
So... doing it properly is very likely to be more expensive than it's worth, and counting in public view lets everyone know what's going on.
And the processes for maintaining chain of custody for ballot boxes are well known. Competent election operators count the ballots *in the precinct*, in public view, immediately after closing the polls.
For a good example, see the classical IBM 360/370/etc./Z-series "SAVE" macro. Also the PowerPC, which has no distinct instruction-level stack. It does, however, have some nice instructions and addressing modes that manipulate registers in a stack-friendly way.
In-town, mostly local freeway fillup-to-fillup average 28.7 MPG summer, 26.9 MPG winter (occasional) long haul trips: 31.6 fillup-to-fillup
I am 6 feet 4 inches tall (38+ inch in-seam), and the car is *perfect* for my height, the best I've ever had. Fold down the rear seats, and it will hold my custom-made extra-large bicycle and my (5 feet 10 inch) wife's bicycle and plenty of additional stowage. No mechanical or electrical problems (it's only 4 years old, ask me in 5 years; we drive our cars >12 years). My wife decided which cars were acceptable from a safety rating perspective, and I got the one I wanted.
The car has plenty of turbo-juice to get onto the freeway from an on-ramp safely.
This is not rocket science and it is possible to have a vehicle that carries you, your friends and furniture without blowing our environmental budget. Our other car is a small sedan that also gets OK mileage (a 1996 Integra), which in a year or two will be replaced with the most efficient vehicle we can find that meets required safety criteria. Maybe hybrid, maybe low-emission diesel.
I don't feel much like an eco-nazi, but to be perfectly straight, my wife and I much prefer walking to the local supermarket: Quality time together is in short supply these days.
Anyone wanting a good buzz should consider aerobic exercise (I like riding a bicycle). With a bit of training, the aerobic rush is awesome, the biochemical effects begin immediately and last several days. Longer term, the exercise helps with stamina and delayed onset of diseases of age. I am 49, eat anything I want any time I want and am not overweight (198 lb).
The human body is built to be worked. Reasonable nutrition and good consistent workouts are just about the best way to maintain the only body you will ever have.
Mod Parent up please!! For once, I wish I had mod points.
In the final analysis, "the market is always right". The trouble, however, is twofold:
1) It tends to take a while, during which time the extortionists and market-manipulators make money hand-over-fist, while the rest of us pay.
2) The adjustment back to equilibrium tends to be unnecessarily violent (think "French Revolution" in politics, or the workers for formerly strong companies getting the shaft when their managements can no longer maintain the manipulation).
That's all *good quality* regulation should do: Recognize the long-term values of good behavior and consequences of bad behavior, with the goal of less resource-sucking and whiplash.
In the '70's, MicroData built a (get this) User Microprogrammable machine. Naturally, not too many people other than E.E. and C.S. researchers wanted to write their own microcode, so MicroData needed to find another line of business. So, they developed the Reality system.
I worked on a MicroData Reality system in during a summer job in 1977. I know that Pick and Reality use the same underlying database approach, but I don't know who derived from whom.
I returned to my parents' place for a weekend visit one year, and next to my sister... Behold! A 6' 2" real live woman! (I was a 21 year old geek and all too aware that are too few girls my size, this could be good...) So I asked my mother who the babysitter was, and was she single? Replied my dear mum, "Don't get any ideas. Period. You will make no remarks, no comments, nothing. You got it?" Oh why, says I...
"She is 12 years old, is a friend of your sister's and she already has enough trouble."
I saw a beautiful woman. She was no such thing: A 12-year old girl, and had already been the target of multiple, completely inappropriate advances by men and boys. She and my sister remained friends for several years, and by the time she was in college, she had already been hurt in too many ways.
These studies of early puberty are not stupid, they are needed in the context of a society that fails to protect its children.
The issue is concentration of influence.
on
The Patent Epidemic
·
· Score: 1
Consequent to the merger-mania of the past 30 years, the number of human minds affecting huge swathes of the intellectual landscape has dropped dramatically. The smaller number of voices has increased the relative power of each of those voices. Furthermore, as they get higher in the food chain, their interests correlate more closely: They are all "big money people", instead of being "newspaper people", "car people", "steel people", "corn people', et cetera.
So we have a much much smaller set of people who have the ability to set our national agenda.
Unsurprisingly, that's just what they are doing.
You are somewhat correct though, we the people have not been watching the store. Instead, we've been caught busting ass trying to survive the onslaught of exceptionally high inflation in the areas where our middle-class finances are affected the most: housing, health and energy. Lazy? Indifferent? only by comparison to the rather more basic issues of survival.
The big money folks are not affected by the same forces as the rest of us, so they set the national agenda. That is why some of us are frustrated. Our middle class voices are part of a general din, but the voices of the powerful are amplified by the megaphone of their big money.
Corporations have many of the "rights" of individuals, due to what I think is a misinterpretation of the 14th amendment. Unlike people who have lives, children, and such, corporations are designed and built for exactly one purpose: Make money. They build stuff and sell stuff only because those behaviors serve the ultimate goal: Generate cash.
Our choices in elections are simple: Choose the representative of the corporate rights party versus the representative of the radical corporate rights party. They have the support (money) of the corporations.
You want us to be less indifferent? Start advocating for clean elections. But watch over your shoulder for the big money folks that are not interested in losing the influence their slush funds have bought them in the recent past, as they will be gunning for you.
I meant to say that in 1980, the end-to-end concepts for delivering what is now commonly known as email was not nearly so complete as it was in 1997. In 1980 (or perhaps earlier), many of the inventions involved in routing, switching and the other tools in the now well-known Internet toolkit were not yet well-known, and establishing competent networking would have been a real invention, to say nothing of wireless delivery.
As of 1997, *everybody* with any experience in communications understood both the conceptual framework and specific techniques needed to deliver email or any other message-like information just about anywhere.
So, I agree with you: By 1997, when the NTP patent was filed, wireless delivery was obvious and the patent should have been summarily rejected.
First of all, I absolutely reject this criteria. One must be entitled to enjoy her property, even if he does not do anything (perceived by others as useful) with it.
Define "enjoy": I assume that you mean either "making money" from the idea, having bought it from an inventer, or simply sitting on it. For now, let's assume you mean "making money"...
This is simply, as they say, factually wrong. NTP first contacted RIM in 2000, a year after the service was introduced. RIM chose to fight, and lost...
"Factually wrong"... As in, RIM had *not* invested their own research and development of infrastructure prior to and during that first year in business? Or alternatively, that they knew that someone else had patented wireless email before starting their business, and started their business anyway? Which is it? I stand by my description. RIM had already been in an increasingly promising business for a year when NTP went after them.
This is exactly what I mean by "broken patent system". By what measure does someone justify starting a business, whatever business that may be, when there are 500,000 nearly secret possible ways for that business to be subject to financial attack by unknown entities?
The current structure of the patent system can only work well when:
Patents are published in a way that allows their claims to be readable by non-lawyers;
It is possible for every developer to recognize, essentially unambiguously, whether every idea in her implementation is patented or free of encumbrance.
(Being a smartaleck here, I realize) That is *exactly* why I refer to my already slim margin on development costs: The way the patent system is currently organized, the required search is a substantial overhead. Enough that it calls into question the basic viability of small development businesses. Intentional or not, the current process is a potentially significant barrier to entry in the technical marketplace.
I agree absolutely that people should be able to benefit from their inventions. I agree absolutely that people should be able to buy and sell their inventions. But I continue ask one central question:
How can I know which of my ideas is already someone else's property? Remember one of the rules of patentability: That an idea is patentable only if it's not obvious to practitioners of the art. That's why I disagree with NTP: Their patents are obvious.
I think there are excellent software/process patents, by the way. Diffie-Hellman key exchange, RSA public key cryptography, quicksort (not patented by C.A.R. Hoare, but as a non-obvious and extremely creative algorithm, it deserved patentability), Feistel networks for simplifying encrypt/decrypt symmetry in cryptography). The common element of these processes is that they were real inventions, they took real research to build, and represented genuine work, which is what I mean by "value add".
I've looked at the patents, and being a mere developer and not an attorney, they *look* all nonobvious, but nothing in them is "new" for anyone who had ever seen a real network in operation. Delivering messages wirelessly to addressable remote devices... Yeah, real creative. I assume you have read them too. Do you think that they represent any development? (In 1980, maybe; in 1997, it was nothing new.)
Retailers, grocery stores provide a clear value add (a.k.a. service): They *make the products accessible* to a broader marketplace of customers with shipping, inventory, marketing, establishment of quality name, et cetera et cetera.
Now compare that to NTP. They provide *no value add*. No work, no service, no accessibility, no publishing. If on the other hand, they make the ideas accessible to those who would like to license them, *that* would be a value add.
As far as I can tell, NTP simply held some patents (silently) until they saw a company that had done its own research and actually did the work to build a profitable business. *Then* they jumped on RIM's "infringement" of "their intellectual property".
I consider this to be the equivalent of a company like NTP staking out a legal but private claim for a piece of land in the middle of a public place, unmarked. Someone comes along and sets up a fruit stand in what they think, incorrectly, is legal open place. After investing effort in building their business, NTP comes along and says that the fruit-stand builders owe them 3 years in back rent.
The issues here are twins: 1) NTP didn't say anything to anybody about "their" ideas. 2) They waited until RIM had invested *big money* in their infrastructure, not knowing about the virtual landmine.
Classically, patents existed to enable the patent-holders to receive a return on their research investment and to get the ideas out into the world to serve as bases for conteinued economic development. NTP's behavior is exemplary of an economically abuse of the patent system.
It's worth noting that patent language is so impenetrable, and the numbers of patents so massive, that it (the patent system as it stands today) probably can no longer serve its original purpose. As a developer, how do verify that
a) my code doesn't infringe one of hundreds of thousands of software patents
b) If I discover that some element of my work happens to be patented by someone else, can I license it for a price that doesn't eliminate the remaining shreds of margin that I still have?
Nobody want MS' IP. They want open standards.
on
The Demise of IP?
·
· Score: 1
Nobody wants Microsoft to "give up their IP for the better good of the state." The Massachusetts IT department wants its document standards to be open.
If Microsoft wants to participate in massachusetts' open document process, they are welcome to support the open document standard that they helped to establish.
Microsoft is welcome to keep their IP proprietary as long as they wish. Massachusetts has simply stated that for their public documents, any vendor willing to support the open standard may participate.
This shit is fucking unbelievable. It's yet another case of Microsoft waving their fists in the air complaining that the whole world wants to take away their ability to compete. Total bullshit. They simply cannot compete with the only amorphous group that they haven't been able to beat to shit with their bloody monopoly.
One was the punchcard system. That's the source of the "hanging chads" stories. (Actually, the source of these stories was that poll workers did not empty the chad collection boxes frequently enough and the punches worked inefficiently with jammed little bits of card). See this FAQ
Another was optical-scan ballots, on paper. According to at least one reporter, the optical scan systems were programmed to kick-out undervote/overvote ballots in some prectincts but not others (leaving "incorrect" ballots in the system to remain un-recountable, instead of corrected at the time the votes were cast). I believe this may have been reported by Greg Palast.
The question of paper vs. electronic voting is fairly simple: Which system is more likely to be hacked "invisibly"? I am old enough to be cynical. Developers know damn well that nontrivial software is essentially never perfect, even when everyone involved wants it to be perfect. Voting systems are nontrivial software systems where someone does *not* want it to be perfect. As far as I am concerned, the most perfect definition of "conflict of interest" is when county or state boards of elections are members of political parties and entrusted with choosing electronic voting systems.
The model of using lock-in as nearly the only way to force customers into staying with the product family may be over. It has been frustrating too many paying customers for too long.
Microsoft has lots (and lots and lots) of very very smart, motivated developers and marketers, and there is always the hope that they can begin to use those resources to build a product that really competes without resotring to bogus, short term ploys like lock-in.
Hmmm... Who turned on my "hopefulness neuron" today?:-)
Slashdot Mirror
You have an excellent point. My first drafts, first versions, are not intended for external view. Note, however, that when I do release something to "the world" (usually co-workers), I've made sure to dot the I's and cross the T's. This includes making sure that documents contain what I intended. The author of the Word document in question did not finish the job.
The embarrassment here is that someone with serious responsibility failed to finish the job correctly.
Back then, we could only hope that others with greater responsibility were doing their jobs. As far as I can tell from the news, there are too many cases where people did a slapdash job, with "life changing effects" for too many. A nice euphemism there: "Life changing"... like "life ending", or "no more legs", or "no more father", things like that.
This stuff is very very important to large numbers of people. This is one of those examples where we expect people to get the bloody details right.
A better idea is to provide the utility with a list of suspect numbers and receive notification of change of state in those accounts. Such state changes include received calls, dialed calls, forwarded calls, et cetera.
The point is to provide law enforcement entities with all information relevant to suspects that have received judicial review of probable cause.
If we're going to track things, the least we can do is filter them for relevancy. In this case, my disagreement with Verizon (and AT&T, who has also been entirely too cooperative with this exceptional monitoring) is that they are not filtering the content for relevancy to the actual suspects.
I remain committed to the idea that "they" (government) should required to submit requests to invade "our" privacy for a theoretically disinterested judicial review. Upon receiving that permission, the technical means is available for providing all relevant information to the monitoring law enforcement entities. All I would ask from our service providers is the filtering for relevance to those whose activities have passed the judicial test of probable cause.
I assume that as a commenter on Slashdot, you are aware of the basic technologies involved in this discussion, and therefore are aware that event filtering is technically feasible. (I will provide a resume to you, including the 8 years that I worked at AT&T BL developing SS7 switching infrastructure, if you really need to get into the ability of modern systems to provide basic filtering.)
I see no place in my comment where the concept of "currently committing a crime" appears in the discussion. That statement is a simple red herring and is irrelevant to the argument. The issue is simply judicial review of probable cause, and while there are likely subtleties to the legal arguments, we still have the fourth amendment. Showing probable cause to justify looking into what are otherwise private communications is a central part of it.
...It's monitoring *everyone*. The point to the tracking program was to note the originating and dialed numbers for *all* conversations, not merely those between suspects and the rest of the world. Furthermore, the whole argument from the beginning is that FISA provides for getting permission to monitor up to 72 hours from the start of the monitoring process.
FISA is intended to provide *exactly* the flexibility required to enable surveillance responsive to changing conditions (the genesis of the 72-hour provision), while still requiring the judicial review that is part of the fourth amendment's requirement of showing probable cause.
And I agree with other commenters that customer transaction records (be they phone calls, or reporting on who bought what groceries for how much) is by no stretch of the imagination "protected free speech".
To the best of my knowledge PJ was a paralegal, has never been a lawyer, and therefore is not an officer of the court. Furthermore, whatever she was before 2003, she is now unambiguously a blogger. Neither more nor less.
PJ has been flawlessly consistent in her trust in the U.S. legal system. She says that it can produce imperfect results, as can any human-created system, but it has a strong tendency to work in the right direction.
In the SCO case, PJ has in her own inimitable way, contributed to this case going the right way: She brings the actinic glare of illumination onto a process that SCO, and others, have tried to accomplish.
PJ started her blog several months before the SCO case became public knowledge, in order to connect geeks with legal concepts. She believed then and still believes that we (the technologists) should be aware of the rules of the game in order avoid being steamrolled by it.
I've been reading the blog since early 2003, and PJ sounds like a gentle spoken (she insists on decorum) but very very intelligent paralegal. The truth justice and the American way language is backed up by consistent and finely honed research and argument. Plus one additional ingredient, a very angry hornet's nest: 10,000+ pissed off developers, some of whom have been around from the inception of UNIX and derived technologies. PJ, more than anything else, has brought us together to face the threat from SCO and folks like them, and for that alone her contribution is inestimable.
The original Eclipse was a highly glorified Nova, both being 16-bit machines. The new machine that Tracy Kidder wrote about was actually the MV/8000 Eagle 32 bit system.
:-) ... give me a few months to afford it, I've been laid-off for several years) but when I compare my little Linux workstation to those *immense beasts*, I get back some of that old feeling that I used to have in my youth.
We (meaning the management at Warner-Amex Cable Communccations, Inc, A.K.A. WACCI) hired several D.G. systems guys around 1980/1981 to help grow the network, which was growing to include Cincinnati, Pittsburgh and a few other cities that I don't remember. We decided to run AOS on a few of the Eclipse C300's (upgraded with COBOL-supporting character instructions through microcode). We were introduced to AOS by some *very* cool D.G. guys from Massachusetts. At the ripe old age of 23, I was *enthralled* with AOS. We received some of the earliest MV/8000 Eagle systems that D.G. produced, and ran AOS/VS (AOS "virtual system") on them. In our view, these machines kicked Major Ass.
AOS and AOS/VS were architected to present a full (virtual) R/DOS machine environment to each of their processes, so all the organizations using INFOS, IDEA, etc. (D.G.'s "database" and online multiterminal system) could move forward into an AOS (16-bit) and AOS/VS (32-bit) environment with minimal hassle. So, for example, the tasks in a classical foreground/background Nova/Eclipse RDOS system were treated as tasks contained with the many processes running in the new AOS and AOS/VS systems.
The result was that AOS/VS gave us multithreaded technology in 1981. And we used it to great benefit to keep our online systems running as quickly as possible. (I was able to write an INFOS-table caching subsystem that doubled the online capacity of each MV/8000 system, in about 1000 lines of PL/I).
The architecture of AOS and AOS/VS were derived from MULTICS, and in my opinion were very successful in achieving their goals. It would be 10 years before the UNIX world would catch up, when Sun Microsystems brought in their lightweight process technology. (I don't know whether other developers had threads going, it's merely that my experience was with Sun.)
Those machines were comparative monsters for us, with 4 Mbyte RAM, 700 Mbyte disk, and processors based on AMD bit-slice technology. (Anyone remember the AMD 29xx bit-slice processor building blocks? 4 bits of ALU per chip, supported by carry lookahead; with creative support from the diskette-loaded microcode, these machines were *wonderful* to work with.)
All this stuff reminds me that this is the same AMD that eventually built the Athlon X64 in my 3-year-old inexpensive workstation. It has 2 Gbytes of RAM, 500 gbytes of disk, running Linux. Not dual-core (yet
My first job out of school. Very cool place. Their polling system consisted of a stack of Data General Nova single board computers, each responsible for polling one supertrunk. They were supervised by a Data General Eclipse (the polling system), which had aggregate responsibility for the entire system.
There was a separate Eclipse, the "Studio System", which used a high speed interprocessor bus to move polling data to and from the polling system.
I wrote several of the studio system's technical scripts, which needed to be synchronized with the TV shows.
QUBE flopped as a technology due mostly to the fact that people are (and were in the late 70's) in the habit of being couch potatoes, rather than interacting through a rather stilted 2-way system.
QUBE gave two-way cable communications hardware people some pretty good practice in how to run signals both ways through a hierarchical network. Eventually, (with huge improvements, etc.) it led to today's cable modems.
A cute cultural story: The two-way boxes were designed by Pioneer Electronics (the stereo folks) in Japan. The Japanese engineers had absolutely no idea how quickly Americans would learn to hack the boxes to watch pay-per-view premium content without the box reporting that they had selected premium channels. It turns out that the box was designed to detect channel change events and track the changes, rather than reporting the channel that was currently selected for viewing. The result was that as soon as someone discovered how to disable the change detection logic (with a paper clip), they started watching premium content for free.
The business management folks had me write a program that statistically analyzed premium purchasing habits, noting (for example) when a given customer transitioned from several months of reasonable amount of premium content, to absolutely zero premium viewing. The program was called "zerobill". Naturally, its capabilities grew in various ways to track a whole range of statistics about viewing habits during the next few years. Eventually, zerobill became *the report* that every manager wanted to see, every morning without fail. I had some *exceptionally early* mornings caused by various bugs and vicissitudes in the database.
Phone rings...
Me: (knowing damn well what was coming next) Hello?
Night operator: "Daily batch died."
Me: "and..."
Night operator: "Not sure, it looks like an error."
Me: "Did it leave a suicide note, or was it just shot in the head?"
et cetera...
My best friend and I were not scheduled the evening of the Rundgren concert, and we had a *kickass* time at the concert, including a little while backstage. It was a great time and place to be a young software geek, mixing television and technology.
2000 ragtag home users? You are smarter than that, I can tell by the quality of your writing and sentence structure alone. While some OSS packages serve small communities, there are lots of packages that serve large and diverse communities. (PostgreSQL, Apache, the Linux kernel, Firefox, the list goes on). Those packages have, on occasion introduced vulnerabilities due to the natural vicissitudes of software development. And when their vulnerabilities are discovered, they get fixed quickly. (And this one hit me this morning: I don't need Linux Genuine Advantage for permission to receive updates to my damn software!!!)
It is worth noting, however, that such vulnerabilities are nearly always limited in scope due the inherently modular nature of the OSS world. Microsoft built a highly integrated system to support its business model. They are welcome to their high integration approach. And those of use who do not appreciate the effects of that way of doing business are welcome to complain when it wacks the shit out of our families' productivity when we are trying to get some proprietary fix.
Because prostate problems do not affect reproduction, and do not even affect a male's ability to support his progeny until they are essentially ready for breeding themselves.
After writing 10k+ lines of code on the PDP-8, I can tell you that there's no argument that could convince me that it was a better architecture than the 8088/8086.
One could almost sustain an argument that the PDP-11 was a better architecture than the 8088/8086 as far as it went, which at the time could be extended to a Mbyte of physical RAM (later instances of the -11 could get all the way to 4 Mbytes physical, as in the 11/73). Manipulating the APRs (active page registers), one could do some mild hackery to make a single process address more than 64kbytes.
Through it would have changed computing history fairly dramatically, I believe the MC68008, an 8-bit-bus variant of the MC68000 would have been a much better choice for IBM's new play-toy in the long run. The reason is the 68k began its design life with 32-bit registers. (IBM was still king of the mainframe hill with their mainframes back then, and they might be forgiven for not taking this new toy-sized computer seriously.)
But I will not second-guess IBM, since their inspiration was the Apple][ and they would have to be very forward-looking indeed to apply Moore's law 10+ years into the future.
The 8080's registers can be mapped directly from the 8008's registers: A, B, C, D, E, H, L. HL taken as a pair could be used to address anywhere in the 8008's 14-bit address space.
Intel's promise that the 8086 had upgrade compatibility with 8080 (or the 8085), came straight from the success of the 8008->8080 transition.
All this talk about ancient processors... I wrote a debugger for an 8008 system in 1976. A collection of geeks in college borrowed an 8086 development kit from the local Bell Labs office, and we did it with an ASR-33 Teletype. I forgot to return the 8008 manual, and still have it. The *fast version* ran at 200 KHz, and each instruction executed in a multiple of 5 clocks (single-byte instructions in 5 clock cycles, two-byte instructions in 10 clock cycles...)
I remember marveling a few years ago that one could trace the 8008's registers and capabilities right up through the latest Pentium4.
Whether they are secure or not, voting machines DO NOT SCALE. They introduce the possibility of "unintentional" misallocation. This translates to under-allocation of machines to precincts in which the current powers-that-be do not want high turnout.
Voting should be performed with two characteristics:
- Scalable (Lots of people can fill out paper ballots in parallel)
- Countable, in full public view, by UNAUGMENTED humans.
Even without the possibility of denial of voting-service attacks, there is the issue of any real security being too expensive to justify. Anyone who has done flight-critical software development knows that the cost of that development is waay higher than the ordinary system design and implementation practices. Voting systems are far closer to hard-realtime systems than most people expect.So... doing it properly is very likely to be more expensive than it's worth, and counting in public view lets everyone know what's going on.
And the processes for maintaining chain of custody for ballot boxes are well known. Competent election operators count the ballots *in the precinct*, in public view, immediately after closing the polls.
For a good example, see the classical IBM 360/370/etc./Z-series "SAVE" macro. Also the PowerPC, which has no distinct instruction-level stack. It does, however, have some nice instructions and addressing modes that manipulate registers in a stack-friendly way.
Read Kingdom Coming by Michelle Goldberg to learn what she has discovered during her recent research.
I live In Columbus, Ohio and have seen some of the things Michelle described first hand.
2002 VW Passat, 1.8L turbo
In-town, mostly local freeway fillup-to-fillup average 28.7 MPG summer, 26.9 MPG winter
(occasional) long haul trips: 31.6 fillup-to-fillup
I am 6 feet 4 inches tall (38+ inch in-seam), and the car is *perfect* for my height, the best I've ever had. Fold down the rear seats, and it will hold my custom-made extra-large bicycle and my (5 feet 10 inch) wife's bicycle and plenty of additional stowage. No mechanical or electrical problems (it's only 4 years old, ask me in 5 years; we drive our cars >12 years). My wife decided which cars were acceptable from a safety rating perspective, and I got the one I wanted.
The car has plenty of turbo-juice to get onto the freeway from an on-ramp safely.
This is not rocket science and it is possible to have a vehicle that carries you, your friends and furniture without blowing our environmental budget. Our other car is a small sedan that also gets OK mileage (a 1996 Integra), which in a year or two will be replaced with the most efficient vehicle we can find that meets required safety criteria. Maybe hybrid, maybe low-emission diesel.
I don't feel much like an eco-nazi, but to be perfectly straight, my wife and I much prefer walking to the local supermarket: Quality time together is in short supply these days.
Anyone wanting a good buzz should consider aerobic exercise (I like riding a bicycle). With a bit of training, the aerobic rush is awesome, the biochemical effects begin immediately and last several days. Longer term, the exercise helps with stamina and delayed onset of diseases of age. I am 49, eat anything I want any time I want and am not overweight (198 lb).
The human body is built to be worked. Reasonable nutrition and good consistent workouts are just about the best way to maintain the only body you will ever have.
Mod Parent up please!! For once, I wish I had mod points.
In the final analysis, "the market is always right". The trouble, however, is twofold:
1) It tends to take a while, during which time the extortionists and market-manipulators make money hand-over-fist, while the rest of us pay.
2) The adjustment back to equilibrium tends to be unnecessarily violent (think "French Revolution" in politics, or the workers for formerly strong companies getting the shaft when their managements can no longer maintain the manipulation).
That's all *good quality* regulation should do: Recognize the long-term values of good behavior and consequences of bad behavior, with the goal of less resource-sucking and whiplash.
Unable to resist...
In the '70's, MicroData built a (get this) User Microprogrammable machine. Naturally, not too many people other than E.E. and C.S. researchers wanted to write their own microcode, so MicroData needed to find another line of business. So, they developed the Reality system.
I worked on a MicroData Reality system in during a summer job in 1977. I know that Pick and Reality use the same underlying database approach, but I don't know who derived from whom.
Where did UniData fit in?
I returned to my parents' place for a weekend visit one year, and next to my sister... Behold! A 6' 2" real live woman! (I was a 21 year old geek and all too aware that are too few girls my size, this could be good...) So I asked my mother who the babysitter was, and was she single? Replied my dear mum, "Don't get any ideas. Period. You will make no remarks, no comments, nothing. You got it?" Oh why, says I...
"She is 12 years old, is a friend of your sister's and she already has enough trouble."
I saw a beautiful woman. She was no such thing: A 12-year old girl, and had already been the target of multiple, completely inappropriate advances by men and boys. She and my sister remained friends for several years, and by the time she was in college, she had already been hurt in too many ways.
These studies of early puberty are not stupid, they are needed in the context of a society that fails to protect its children.
Consequent to the merger-mania of the past 30 years, the number of human minds affecting huge swathes of the intellectual landscape has dropped dramatically. The smaller number of voices has increased the relative power of each of those voices. Furthermore, as they get higher in the food chain, their interests correlate more closely: They are all "big money people", instead of being "newspaper people", "car people", "steel people", "corn people', et cetera.
So we have a much much smaller set of people who have the ability to set our national agenda.
Unsurprisingly, that's just what they are doing.
You are somewhat correct though, we the people have not been watching the store. Instead, we've been caught busting ass trying to survive the onslaught of exceptionally high inflation in the areas where our middle-class finances are affected the most: housing, health and energy. Lazy? Indifferent? only by comparison to the rather more basic issues of survival.
The big money folks are not affected by the same forces as the rest of us, so they set the national agenda. That is why some of us are frustrated. Our middle class voices are part of a general din, but the voices of the powerful are amplified by the megaphone of their big money.
Corporations have many of the "rights" of individuals, due to what I think is a misinterpretation of the 14th amendment. Unlike people who have lives, children, and such, corporations are designed and built for exactly one purpose: Make money. They build stuff and sell stuff only because those behaviors serve the ultimate goal: Generate cash.
Our choices in elections are simple: Choose the representative of the corporate rights party versus the representative of the radical corporate rights party. They have the support (money) of the corporations.
You want us to be less indifferent? Start advocating for clean elections. But watch over your shoulder for the big money folks that are not interested in losing the influence their slush funds have bought them in the recent past, as they will be gunning for you.
I meant to say that in 1980, the end-to-end concepts for delivering what is now commonly known as email was not nearly so complete as it was in 1997. In 1980 (or perhaps earlier), many of the inventions involved in routing, switching and the other tools in the now well-known Internet toolkit were not yet well-known, and establishing competent networking would have been a real invention, to say nothing of wireless delivery.
As of 1997, *everybody* with any experience in communications understood both the conceptual framework and specific techniques needed to deliver email or any other message-like information just about anywhere.
So, I agree with you: By 1997, when the NTP patent was filed, wireless delivery was obvious and the patent should have been summarily rejected.
This is exactly what I mean by "broken patent system". By what measure does someone justify starting a business, whatever business that may be, when there are 500,000 nearly secret possible ways for that business to be subject to financial attack by unknown entities?
The current structure of the patent system can only work well when:
(Being a smartaleck here, I realize) That is *exactly* why I refer to my already slim margin on development costs: The way the patent system is currently organized, the required search is a substantial overhead. Enough that it calls into question the basic viability of small development businesses. Intentional or not, the current process is a potentially significant barrier to entry in the technical marketplace.
I agree absolutely that people should be able to benefit from their inventions. I agree absolutely that people should be able to buy and sell their inventions. But I continue ask one central question:
How can I know which of my ideas is already someone else's property? Remember one of the rules of patentability: That an idea is patentable only if it's not obvious to practitioners of the art. That's why I disagree with NTP: Their patents are obvious.
I think there are excellent software/process patents, by the way. Diffie-Hellman key exchange, RSA public key cryptography, quicksort (not patented by C.A.R. Hoare, but as a non-obvious and extremely creative algorithm, it deserved patentability), Feistel networks for simplifying encrypt/decrypt symmetry in cryptography). The common element of these processes is that they were real inventions, they took real research to build, and represented genuine work, which is what I mean by "value add".
I've looked at the patents, and being a mere developer and not an attorney, they *look* all nonobvious, but nothing in them is "new" for anyone who had ever seen a real network in operation. Delivering messages wirelessly to addressable remote devices... Yeah, real creative. I assume you have read them too. Do you think that they represent any development? (In 1980, maybe; in 1997, it was nothing new.)
Retailers, grocery stores provide a clear value add (a.k.a. service): They *make the products accessible* to a broader marketplace of customers with shipping, inventory, marketing, establishment of quality name, et cetera et cetera.
Now compare that to NTP. They provide *no value add*. No work, no service, no accessibility, no publishing. If on the other hand, they make the ideas accessible to those who would like to license them, *that* would be a value add.
As far as I can tell, NTP simply held some patents (silently) until they saw a company that had done its own research and actually did the work to build a profitable business. *Then* they jumped on RIM's "infringement" of "their intellectual property".
I consider this to be the equivalent of a company like NTP staking out a legal but private claim for a piece of land in the middle of a public place, unmarked. Someone comes along and sets up a fruit stand in what they think, incorrectly, is legal open place. After investing effort in building their business, NTP comes along and says that the fruit-stand builders owe them 3 years in back rent.
The issues here are twins: 1) NTP didn't say anything to anybody about "their" ideas. 2) They waited until RIM had invested *big money* in their infrastructure, not knowing about the virtual landmine.
Classically, patents existed to enable the patent-holders to receive a return on their research investment and to get the ideas out into the world to serve as bases for conteinued economic development. NTP's behavior is exemplary of an economically abuse of the patent system.
It's worth noting that patent language is so impenetrable, and the numbers of patents so massive, that it (the patent system as it stands today) probably can no longer serve its original purpose. As a developer, how do verify that
a) my code doesn't infringe one of hundreds of thousands of software patents
b) If I discover that some element of my work happens to be patented by someone else, can I license it for a price that doesn't eliminate the remaining shreds of margin that I still have?
Nobody wants Microsoft to "give up their IP for the better good of the state." The Massachusetts IT department wants its document standards to be open.
If Microsoft wants to participate in massachusetts' open document process, they are welcome to support the open document standard that they helped to establish.
Microsoft is welcome to keep their IP proprietary as long as they wish. Massachusetts has simply stated that for their public documents, any vendor willing to support the open standard may participate.
This shit is fucking unbelievable. It's yet another case of Microsoft waving their fists in the air complaining that the whole world wants to take away their ability to compete. Total bullshit. They simply cannot compete with the only amorphous group that they haven't been able to beat to shit with their bloody monopoly.
There were a few systems in use in Florida.
One was the punchcard system. That's the source of the "hanging chads" stories. (Actually, the source of these stories was that poll workers did not empty the chad collection boxes frequently enough and the punches worked inefficiently with jammed little bits of card). See this FAQ
Another was optical-scan ballots, on paper. According to at least one reporter, the optical scan systems were programmed to kick-out undervote/overvote ballots in some prectincts but not others (leaving "incorrect" ballots in the system to remain un-recountable, instead of corrected at the time the votes were cast). I believe this may have been reported by Greg Palast.
The question of paper vs. electronic voting is fairly simple: Which system is more likely to be hacked "invisibly"? I am old enough to be cynical. Developers know damn well that nontrivial software is essentially never perfect, even when everyone involved wants it to be perfect. Voting systems are nontrivial software systems where someone does *not* want it to be perfect. As far as I am concerned, the most perfect definition of "conflict of interest" is when county or state boards of elections are members of political parties and entrusted with choosing electronic voting systems.
The model of using lock-in as nearly the only way to force customers into staying with the product family may be over. It has been frustrating too many paying customers for too long.
:-)
Microsoft has lots (and lots and lots) of very very smart, motivated developers and marketers, and there is always the hope that they can begin to use those resources to build a product that really competes without resotring to bogus, short term ploys like lock-in.
Hmmm... Who turned on my "hopefulness neuron" today?