Actually, it doesn't matter whether it's "boilerplate". The fact remains that commonly "boilerplated" terms of service are overly broad, and _could_ easily be used to subvert the rights of the individual, simply because the company chose to reserve overbearing rights in its favor by using the broad language.
Any time you have to depend on claims of "We wouldn't really do that, it's just for our protection", wonder why the language isn't equally broad in your own favor for _your_ own protection.
When I was taking an AI class in college I chose Perl for many of the smaller assignments, because it gave a great mix of the features of Scheme and C. You have pseudo-first-class procedures, good reference capabilities, plus all the looping and iterative constructs that one sometimes misses in a language like C. I found my assignments were really easy when Perl was used in this fashion.
I've always thought that Perl has a Lisp-spirit hiding in there somewhere, and I've programmed with it accordingly. Unfortunately the syntax is not meant to bring this forth, and it ends up as a mash of punctuation when one tries this programming style with it.
I've lately gotten very interested in Ruby, since it appears to have taken most of the good part of Perl and made it into a more logical language, with even more of the features of Scheme and Lisp. Unfortunately, even Ruby isn't meant to be used the way I like.
One question I have about the constant harping that Linux isn't there _yet_. What constitutes "good enough" for Linux?
Honest question. I agree with the assessment that BeOS is better for dethroning the giant, but everyone that supports Linux says "It's getting there but not _yet_. How many years, kernel revisions, and GUIs have to be constructed before someone hits on what _is_ good enough as a replacement?
Often the statement seems to come from someone who wants to show grudging respect to Linux but doesn't think it will _ever_ gain market share. To me this constant "not yet" harping just gets the public worried. Better to describe what it does and doesn't do, not to hold it up to a future ideal version of itself which it may or may not reach due to reasons entirely outside the concept of "better" or "worse".
Nope. Promise Keepers are considered anti-Christian by many Fundamentalist groups, because they believe it puts the importance of relations between men and discussion of sin before the Bible.
And there are lots of detractors on USENET. I'm surprised the whole of USENET isn't just blocked, but that _wouldn't_ seem evenhanded. The bible discussion groups draw lots more detractors than any forum dedicated to the detractors, so they might encourage honest discussion and thought about one's religious beliefs.
I'm finding that more in the miniseries matches the details and spirit of the novel than I was willing to believe at first. The most major _change_ I see, as opposed to ommission or reworking, is the interaction with Irulan.
Appreciating the Lynch version took me a few years, during which I came to see what an _economic_ use of language Lynch used. In the opening "talking head" of Irulan, and the first few minutes of free-association with the picture-book, volumes of background information were exposed. The miniseries does indeed portray many things differently, but in the end by referring back to the text I can figure out a kind of mapping between the series and the original text. Often descriptions of things that will later be _seen_ are not described; for example, the tooth. All the more surprise for the uninitiated, but already known to the well-read.
Perhaps after cooling off a bit and seeing the rest of it, some more sober analysis can be done. I have a feeling that this version is glossing over the beginning of the story to set up the minimum possible for the parts that focus on the Fremen. I also get the idea that many points, such as the nature of the Guild and navigators, will be brought up later, so some ommissions may still be included in some form.
Actually, it can be quite useful. Remember, not everyone has high bandwidth availability, and that latency does not adversely affect the total speed of transmission if some special techniques are used to tweak TCP parameters.
I should know, I've been working for a company that has provided two-way satellite internet since 1995. Several hundred sites, and a few thousand more in the works . . . So it's a misnomer that it's finally "here". But it's just now becoming consumer-targeted rather than company-targeted.
There are software packages that can be used to significantly improve response even with such a large latency. Mentat makes such a package. You can't overcome the latency but you can use a proxy that converts TCP to a protocol better suited to high latencies.
About the only thing "out of the question" on such a link is gaming. Other things might be slower but still possible. Webhosting would be a bad idea since you can be sure those figures are _shared_ among many clients, but you could use such a link to connect to a reverse proxy on a faster network.
As far as we _know_, their purpose wasn't to create amazingly high-speed processors. They probably expected to have the clock speed a bit higher than the competing x86 chips on the market at the time of release. I'm sure the power consumption was one of the design parameters, but I suspect they simply couldn't push the first model fast enough to make it worth competing on equal ground, and shifted their focus to move the product from vapor into market entity.
Not that I think that's bad, or that their chip isn't useful for the purposes they're advertising. They've chosen their strenghts to market on, and you can't tell me that a chip performing so much in software _isn't_ being constantly developed. I really think they have the benefit of being able to try lots of tricks with hardware and software for a quicker turnaround, should they choose to take advantage of the "virtual" nature of their chip.
When I first read this comment, it seemed impossible; the list of institutions, and the amount of routing power required, to block every single "public" IP would be resource-intensive.
But the more I think about it, the more it seems to be a decent solution. Like the RBL, except it lists sites that _want_ to block certain kinds of content. In fact, many sites would voluntarily use it if it would honestly reduce complaints or legal liability.
Would only work well for institutions with static IPs, individuals would be out of luck unless they have static IPs or some other authentication method is used.
Spoofing or tunneling through another site for access might be another problem.
Mostly agreee, but you need to add one more item to section 2.
(a) If general society doesn't like it, politicians won't support it UNLESS (i) there's something in it from them (campaign contributions, bribes) and (ii) the public is clueless enough to be convinced they'll like it, at the hands of a corporate entity with enough money to change public opinion almost at will.
My employer went with Cobalt because it _was_ running Intel Linux. We certainly like the functions it provides through the point-and-drool interface, but we're also loading specific applications onto it that the customer never sees.
We went with Cobalt because they offered incredible integration and preloading services crafted to our specifications. If the OS switches to Solaris, lots of our existing software won't run without convincing our suppliers to compile it for SolPliance OS or whatever they want to call it. If they even _let_ anyone do that.
Good thing we're investigating other suppliers of point-and-drool rackmount servers.
First of all, it isn't apple's very first attempt at a Unix-like system. Lisa (whose operating system was somewhat UNIX-like, and was massively simplified for the Mac), A UX, perhaps even the Next box if you count Jobs in the picture. I count it as their fourth attempt, something having been learned at each stage.
Actually the best part about Cobalt is that they will enter into agreements with one's company to do a lot of the software packaging and installation from images at their own factory, saving lots of time. They'll also contract out to expand their GUI to cover your applications. Not with single-unit purchases, but with a company-wide commitment they'll really work hard at making your life easy. That's been my experience, at least.
A good illustration of futility would be . . .
on
A New DeCSS
·
· Score: 1
to write a "metaprogram" that randomly writes valid programs, one after the other, and applies the programs to data. Use Perl or Scheme or something in which this is easy. Make sure to include system calls for any hardware on the system, or the ability to insert fragments of assembly language or raw random binary data called as code. Call it DeCSS.
There's a nonzero chance that it will be able to break into the Pentagon's networked computers, format your hard drive, or even decrypt a DVD and copy it to the hard drive ready for reproduction. Small, but nonzero, if the metaprogram can be proven to produce programs of enough complexity to handle the proven minimum complexity of these other tasks.
Better yet, have the program output many different programs called DeCSS that change over time, write another to do incremental modifications to this body of code, and start a "genetic code repository". Hey, it _could_ be used somehow to pirate DVD, why not ban code and programming?
it's all right to avoid TV, hours of pointless chimpanzee politicking among the alpha males, and general concern with that which lowers my intelligence?
Y'think? Maybe introversion isn't a disease to be cured, but another way of looking at things, perhaps superior at times? Hmmm? Mebbe?
I took Computer Science at a University because at the time it was considered the only "calling card" worthy of getting a job in the computer industry, among the programs offered where I was able to attend. Sure, I could have taken MIS, but I'm sure people here are aware of the horrible associations made with the worst of the MIS folks. (I didn't say they were _all_ bad, just that those that really don't know what they're doing seem to have greater influence on perception than their proportion.)
Fact is, I have used very little of a computer science education, since nobody really told me that computer _science_ is really a preparation for further academic work in the field--experimentation, research, invention--rather than business use of computer skills. MIS wasn't technical enough, didn't give enough programming experience. CS was way over my head in terms of the required mathematics classes and the general political structure.
I would have been better off with some tech school programs, along with some specific training courses in commercial UNIX systems, routing/switching, and other useful things. So are about 75% of the people who go into a Computer Science program--it just isn't an optimal way to enter the workforce as an administrator or programmer.
Now, I appreciate the academic angle I learned through completing my degree, once I realized my half-error. I _did_ learn many useful skills, but most have had to be twisted and modified in order to really apply.
Perhaps this is one reason attendance is dropping, among women or any other group? We're so bent on college educations, because the employers are as well. Employers ask for things like Bachelor's degree (Master's preferred) in CS plus 15 years of experience in Windows 2000 and 30 years of experience in PC hardware. This game really needs to stop somewhere, it's an endless triangle--businesses, educators, students--but any one party that stops playing the game stands the risk of being unemployable while everyone else continues to play.
Maybe not in the long run, but the short run is lots longer than you might realize. I think one great unconsidered market for wireless broadband is rural areas. It's difficult enough to get a high-speed data connection in these areas, but just try running a decent Internet service with the local phone cooperatives in some places.
Wireless allows an end-run around the traditional connectivity providers, increasing competition and perhaps supplying a long-term solution if the providers of land-based solutions don't get their acts together. Very very useful. In fact, it isn't news that broadband wireless is ready for deployment, many sites have already been deployed in the past few years.
that perhaps 2% of even the technically savy public really understands Transmeta's press release. Sigh. These issues were discussed in great detail, but they're enough different from the status quo that the technically inclined are attacking it from all sides at once. It's not just this isolated incident, I don't have digits enough to count the times when well-meaning (and not-so-well-meaning) members of the tech press have attacked Transmeta and its ideas immediately after having just heard the information that would refute their arguments.
Whether they can carry through with their ideas is another question, but the point stands.
Re:Why not just use the Crusoe as a G4?
on
Darwin on Crusoe?
·
· Score: 1
The real magic, it becomes apparent to me, is that they have die space to work with. I could see licensing agreements with Transmeta to put things like the AltiVec onto the die, and interface it to the rest of the software/hardware combination that runs the main processing functions.
The most interesting question asked at the initial press release was whether they would license the technology as a processor core, like ARM and other vendors, allowing production of custom architectures around the core. It appeared as if they hadn't really thought ahead that far yet.
Transmeta already made two entirely compatible but differently engineered chips, each with hardware optimizations for specific operating environments. Just because they chose to jump into a particular market right away (x86) doesn't mean that's where they're at now internal to the company, and doesn't mean that's all they'll do.
They might indeed make one, but don't count on it being much more than an inferior copy as long as it isn't running on Intel Windoze.
While true that they've bothered to port IE to some Unix platforms, the problem is that they don't consider a "platform" to be anything more than "a collection of hardware on which Windows does not run in some form". Thus they tooted their own horn about being "cross-platform" with IE for UNIX, while only creating the browser for those hardware platforms on which there was not ever/was no longer a Windows NT port. Forget about Linux on Intel.
There was an article some time ago on Slashdot about winning the temporary battle while losing the long-term browser war. The recent push towards third-generation display technologies (Apple's Aqua + OSX) and de-facto standardization of browser features against IE as a benchmark will extend farther into media formats and the players that are required to use them.
Does the fact that they're only now moving to have it sealed show that they weren't diligent to protect their trade secrets to begin with? I don't know, not being a lawyer and all, blah blah blah.
If so, it might be an interesting bit of weaponry for future legal action.
Of course we all know LinuxOne is a fully owned subsidiary of Microsoft, Inc. doing this to spread FUD. Couldn't be that anyone is out to just scam the public, or that anyone is plain incompetent.
Actually, I think Gilliam tends for overstatement on the surface, with very subtle sub-points underneath, using Brazil as the example. I'm sure for lots of the geek crowd the various warnings about the nature of information and consciousness are obviously tied in with the over-the-top nature of the movie, but I think the greater proportion of the population never sees this.
Brazil is such a deeply frightening movie for anyone who has worked with information technology. I think it should be required viewing/discussion for anyone studying computer science or information systems in school, or for anyone entering the corporate culture for the first time.
This is the real problem with the mainframes. They're wonderful for running stable applications 24/7, _if_ the site can make the appropriate investment in expertise to run a shop the way IBM intends.
Part of the benefit of UNIX shops is that the expertise is cheaper and that a small core of people can keep a shop running well. When I worked in a shop supporting a mainframe with "legacy" applications, getting anything done degenerated into a game of "find the scapegoat" because only one person with the appropriate knowledge was available, and that person liked to play mind games.
We really needed a team of analysts, engineers, managers, and operators to keep the machine running. Not because the _knowledge_ required was that great, but because the mainframe system enforces a specific _political_ system in its presence and operation.
Well, technically it is supposed to "ease migration". However, I found that
nobody at IBM was interested in helping us to get it working properly
the POSIX environment was terrible, even for a shop with mainframe _and_ unix experience in-house, and was more akin to NT's capacity to claim "POSIX" for itself
the develpoment tools were unable to deal with even the simplest of free software, and we were forced to use pre-compiled versions about 6 months behind the latest version
So, despite our desire to try this technology, we found that the easiest and cheapest way to "ease migration" was to purchase additional UNIX servers and upgrade to 100 meg Ethernet on the mainframe, at a cost of $100,000. (For just the TCP/Ethernet upgrade!)
Problem with lots of the mainframe stuff is that even when it offers certain capabilities, you need to have a corporate tech bureaucracy of the kind IBM is expecting (i.e. fifteen different levels of managers and engineers and analysts, all reporting through three overlapping chains of command to enhance any existing communications problems) to get anywhere with it.
Having Linux available on that hardware would have made it much easier to deal with their machine. The POSIX-environment provided by OMVS never cut it for my previous site, and having a UNIX partition on that machine would have helped greatly.
Slashdot Mirror
Any time you have to depend on claims of "We wouldn't really do that, it's just for our protection", wonder why the language isn't equally broad in your own favor for _your_ own protection.
I've always thought that Perl has a Lisp-spirit hiding in there somewhere, and I've programmed with it accordingly. Unfortunately the syntax is not meant to bring this forth, and it ends up as a mash of punctuation when one tries this programming style with it.
I've lately gotten very interested in Ruby, since it appears to have taken most of the good part of Perl and made it into a more logical language, with even more of the features of Scheme and Lisp. Unfortunately, even Ruby isn't meant to be used the way I like.
Honest question. I agree with the assessment that BeOS is better for dethroning the giant, but everyone that supports Linux says "It's getting there but not _yet_. How many years, kernel revisions, and GUIs have to be constructed before someone hits on what _is_ good enough as a replacement?
Often the statement seems to come from someone who wants to show grudging respect to Linux but doesn't think it will _ever_ gain market share. To me this constant "not yet" harping just gets the public worried. Better to describe what it does and doesn't do, not to hold it up to a future ideal version of itself which it may or may not reach due to reasons entirely outside the concept of "better" or "worse".
Nope. Promise Keepers are considered anti-Christian by many Fundamentalist groups, because they believe it puts the importance of relations between men and discussion of sin before the Bible. And there are lots of detractors on USENET. I'm surprised the whole of USENET isn't just blocked, but that _wouldn't_ seem evenhanded. The bible discussion groups draw lots more detractors than any forum dedicated to the detractors, so they might encourage honest discussion and thought about one's religious beliefs.
Appreciating the Lynch version took me a few years, during which I came to see what an _economic_ use of language Lynch used. In the opening "talking head" of Irulan, and the first few minutes of free-association with the picture-book, volumes of background information were exposed. The miniseries does indeed portray many things differently, but in the end by referring back to the text I can figure out a kind of mapping between the series and the original text. Often descriptions of things that will later be _seen_ are not described; for example, the tooth. All the more surprise for the uninitiated, but already known to the well-read.
Perhaps after cooling off a bit and seeing the rest of it, some more sober analysis can be done. I have a feeling that this version is glossing over the beginning of the story to set up the minimum possible for the parts that focus on the Fremen. I also get the idea that many points, such as the nature of the Guild and navigators, will be brought up later, so some ommissions may still be included in some form.
I should know, I've been working for a company that has provided two-way satellite internet since 1995. Several hundred sites, and a few thousand more in the works . . . So it's a misnomer that it's finally "here". But it's just now becoming consumer-targeted rather than company-targeted.
There are software packages that can be used to significantly improve response even with such a large latency. Mentat makes such a package. You can't overcome the latency but you can use a proxy that converts TCP to a protocol better suited to high latencies.
About the only thing "out of the question" on such a link is gaming. Other things might be slower but still possible. Webhosting would be a bad idea since you can be sure those figures are _shared_ among many clients, but you could use such a link to connect to a reverse proxy on a faster network.
Not that I think that's bad, or that their chip isn't useful for the purposes they're advertising. They've chosen their strenghts to market on, and you can't tell me that a chip performing so much in software _isn't_ being constantly developed. I really think they have the benefit of being able to try lots of tricks with hardware and software for a quicker turnaround, should they choose to take advantage of the "virtual" nature of their chip.
But the more I think about it, the more it seems to be a decent solution. Like the RBL, except it lists sites that _want_ to block certain kinds of content. In fact, many sites would voluntarily use it if it would honestly reduce complaints or legal liability.
Would only work well for institutions with static IPs, individuals would be out of luck unless they have static IPs or some other authentication method is used.
Spoofing or tunneling through another site for access might be another problem.
(a) If general society doesn't like it, politicians won't support it UNLESS (i) there's something in it from them (campaign contributions, bribes) and (ii) the public is clueless enough to be convinced they'll like it, at the hands of a corporate entity with enough money to change public opinion almost at will.
We went with Cobalt because they offered incredible integration and preloading services crafted to our specifications. If the OS switches to Solaris, lots of our existing software won't run without convincing our suppliers to compile it for SolPliance OS or whatever they want to call it. If they even _let_ anyone do that.
Good thing we're investigating other suppliers of point-and-drool rackmount servers.
First of all, it isn't apple's very first attempt at a Unix-like system. Lisa (whose operating system was somewhat UNIX-like, and was massively simplified for the Mac), A UX, perhaps even the Next box if you count Jobs in the picture. I count it as their fourth attempt, something having been learned at each stage.
Actually the best part about Cobalt is that they will enter into agreements with one's company to do a lot of the software packaging and installation from images at their own factory, saving lots of time. They'll also contract out to expand their GUI to cover your applications. Not with single-unit purchases, but with a company-wide commitment they'll really work hard at making your life easy. That's been my experience, at least.
There's a nonzero chance that it will be able to break into the Pentagon's networked computers, format your hard drive, or even decrypt a DVD and copy it to the hard drive ready for reproduction. Small, but nonzero, if the metaprogram can be proven to produce programs of enough complexity to handle the proven minimum complexity of these other tasks.
Better yet, have the program output many different programs called DeCSS that change over time, write another to do incremental modifications to this body of code, and start a "genetic code repository". Hey, it _could_ be used somehow to pirate DVD, why not ban code and programming?
Y'think? Maybe introversion isn't a disease to be cured, but another way of looking at things, perhaps superior at times? Hmmm? Mebbe?
Fact is, I have used very little of a computer science education, since nobody really told me that computer _science_ is really a preparation for further academic work in the field--experimentation, research, invention--rather than business use of computer skills. MIS wasn't technical enough, didn't give enough programming experience. CS was way over my head in terms of the required mathematics classes and the general political structure.
I would have been better off with some tech school programs, along with some specific training courses in commercial UNIX systems, routing/switching, and other useful things. So are about 75% of the people who go into a Computer Science program--it just isn't an optimal way to enter the workforce as an administrator or programmer.
Now, I appreciate the academic angle I learned through completing my degree, once I realized my half-error. I _did_ learn many useful skills, but most have had to be twisted and modified in order to really apply.
Perhaps this is one reason attendance is dropping, among women or any other group? We're so bent on college educations, because the employers are as well. Employers ask for things like Bachelor's degree (Master's preferred) in CS plus 15 years of experience in Windows 2000 and 30 years of experience in PC hardware. This game really needs to stop somewhere, it's an endless triangle--businesses, educators, students--but any one party that stops playing the game stands the risk of being unemployable while everyone else continues to play.
Wireless allows an end-run around the traditional connectivity providers, increasing competition and perhaps supplying a long-term solution if the providers of land-based solutions don't get their acts together. Very very useful. In fact, it isn't news that broadband wireless is ready for deployment, many sites have already been deployed in the past few years.
Whether they can carry through with their ideas is another question, but the point stands.
The most interesting question asked at the initial press release was whether they would license the technology as a processor core, like ARM and other vendors, allowing production of custom architectures around the core. It appeared as if they hadn't really thought ahead that far yet.
Transmeta already made two entirely compatible but differently engineered chips, each with hardware optimizations for specific operating environments. Just because they chose to jump into a particular market right away (x86) doesn't mean that's where they're at now internal to the company, and doesn't mean that's all they'll do.
While true that they've bothered to port IE to some Unix platforms, the problem is that they don't consider a "platform" to be anything more than "a collection of hardware on which Windows does not run in some form". Thus they tooted their own horn about being "cross-platform" with IE for UNIX, while only creating the browser for those hardware platforms on which there was not ever/was no longer a Windows NT port. Forget about Linux on Intel.
There was an article some time ago on Slashdot about winning the temporary battle while losing the long-term browser war. The recent push towards third-generation display technologies (Apple's Aqua + OSX) and de-facto standardization of browser features against IE as a benchmark will extend farther into media formats and the players that are required to use them.
If so, it might be an interesting bit of weaponry for future legal action.
Of course we all know LinuxOne is a fully owned subsidiary of Microsoft, Inc. doing this to spread FUD. Couldn't be that anyone is out to just scam the public, or that anyone is plain incompetent.
Brazil is such a deeply frightening movie for anyone who has worked with information technology. I think it should be required viewing/discussion for anyone studying computer science or information systems in school, or for anyone entering the corporate culture for the first time.
So I think he has a chance to pull it off.
Part of the benefit of UNIX shops is that the expertise is cheaper and that a small core of people can keep a shop running well. When I worked in a shop supporting a mainframe with "legacy" applications, getting anything done degenerated into a game of "find the scapegoat" because only one person with the appropriate knowledge was available, and that person liked to play mind games.
We really needed a team of analysts, engineers, managers, and operators to keep the machine running. Not because the _knowledge_ required was that great, but because the mainframe system enforces a specific _political_ system in its presence and operation.
nobody at IBM was interested in helping us to get it working properly
the POSIX environment was terrible, even for a shop with mainframe _and_ unix experience in-house, and was more akin to NT's capacity to claim "POSIX" for itself
the develpoment tools were unable to deal with even the simplest of free software, and we were forced to use pre-compiled versions about 6 months behind the latest version
So, despite our desire to try this technology, we found that the easiest and cheapest way to "ease migration" was to purchase additional UNIX servers and upgrade to 100 meg Ethernet on the mainframe, at a cost of $100,000. (For just the TCP/Ethernet upgrade!)
Problem with lots of the mainframe stuff is that even when it offers certain capabilities, you need to have a corporate tech bureaucracy of the kind IBM is expecting (i.e. fifteen different levels of managers and engineers and analysts, all reporting through three overlapping chains of command to enhance any existing communications problems) to get anywhere with it.
Having Linux available on that hardware would have made it much easier to deal with their machine. The POSIX-environment provided by OMVS never cut it for my previous site, and having a UNIX partition on that machine would have helped greatly.
"Unbreakable" is a bit of a difficult concept when dealing with the speeds involved in orbits and landings from orbit.