I had long held the same opinion of the paper, that patents are fundamentally just government sanctioned monopolies. But the theory the article presented made me rethink; there is one huge and important difference between a capitalistic monopoly and these artificial monopolies. With patents it is possible to obtain a monopoly with 0% market share!
Essentially patents allow you to get a monopoly on not producing something. You can't do that in the real market-driven world where monopolies must be gained by capturing near 100% of the market. This is also one of the main differences between copyright and patent: I can't copyright something before I write it, but I sure can patent something without ever having to produce anything.
Patents used legitimately as a means to allow an inventor to safely outlay a large initial economic investment are somewhat defensible, but patents which exist soley to prevent anybody from ever acting upon some idea are just evil. And that's what seems to be happening so much today, especially in the area of software patents.
This is such a refreshing article which finally attempts to put good economic theory to work rather than the extremes on both sides: artists will dies f starvation, or IP will make it illegal to think.
I particularly found interesting the concept of rent-seeking bahvior,
"...producers are likely to engage in what economists call "
rent-seeking behavior" -- efforts to protect or expand turf (and profits) by fighting for government-granted monopoly protection -- and that behavior is likely to stifle innovation. Expensive patent races, defensive patenting..., and costly infringement battles are common functions of corporate law departments."
I sure would like to have seen a deeper exploration of that theory, as I feel that is where the most problems lie with the whole IP issue. Consider for instance the cost of the patent itself. The article does a good job of analyzing the cost of R&D; the initial investment in technology. But it doesn't talk about patents in their own as an object of value. How many IP-hoarding companies exist soley to accumulate patents, and have never made any investment into research or innovation?
The patent has an inherent value separate from that of the technology that it may describe.
I would also like to see more thought about the issue of the interaction of patents with each other. The article seems to concentrate solely on one idea at a time...that each patent or IP instrument exists in isolation of all others. It does not deal with the deadlock of interdependent patents. When company A has a patent on idea X and company B has a patent on idea Y, it becomes an impediment to invent idea Z which is created from both. And in the real world, this deadlocking of ideas is out of control.
And finally, the article does not consider fully the costs imposed by the patent or IP system itself. Especially since patents are written in such a manner as to be maximally ambiguous and stealthy it becomes an extreme economic burden on any inventor to know if the ideas being developed have already been claimed. Thus why we are seeing such strange imbalances; you know, 10 lawyers for every engineer.
When you use cash you don't have to give out your identity, and short of using witnesses or video surveilence tapes you're pretty anonymous and untrackable. Be careful with gift certificates/cards though as they potentially can be tracked.
For the online world you'd pretty much need to develop some sort of zero-knowledge system; one that involves real-world entities, not just zero-knowledge cryptographic techniques. Something like an indentity-free payment system (there have been many of these experiements), combined with a zero-knowledge separation of the inventory system from the shipping system.
Of course that only covers you after the fact. If the government is able to set up things before your purchase (like wiretaps and other spy-stuff), then you will get caught.
But as soon as warning labels start showing up, some of 'em will start to wonder what they're being warned about
Especially when the warning reads: Secured for your protection.
The industry always likes to reverse the meaning of all important words to make something bad sound good. Remember SDMI or "Secure" media. All warnings that the product is so secure that you can't even use it!
That's why I'm still playing NetHack...well that and because it's still the best game out there.
Seriously though, can these super cards be used for anything other than the generation of display output? As they are doing so much 3D processing so much faster than any CPU can, I'd like to see the ability to use these GPU's as coprocessors for rendering images back to software/files rather than just to display output. Something like using it as a hardware accelerator for POV-Ray or Renderman. Does anybody have any insight into potential non-traditional uses of these super cards?
AltaVista used to be the best search engine; it's strength lies in basic text searching and it's incredible speed and scalability. Unfortunately it did not account much for the interlinked nature of the web and was easily subverted by web author tricks. These faults were mostly solved by Google.
However, just as Google offers a stand-alone embedded box, the Google Appliance, for use within corporate intranets, I suspect that is an area where AltaVista's technology could thrive much better.
Intranet searching and indexing is still a rather underexploited market. There's basically Microsoft's Index Server, flaws and all, the Google Appliance, and several good but not great minor choices such as ht://Dig. If we could get an AltaVista appliance that ran under Unix (or at least not bound to Microsoft) and underpriced the Google Appliance I would have to believe that a lot of companies would take notice.
C is a low level language that nearly nobody understands....Java is a good high-level language.
When will the myth ever die? Java is not a high-level language. Neither is C nor C++. They are all categorized as low-level languages. Being high or low-level has nothing to do with how "good" a language is or how much hype or popularity or evangelism it has. Go study the theory of programming languages.
The level of a language has to do with the expressiveness of the paradigms (concepts) you can use directly in the language. In this regard it can be easily argued that C++ is of a higher-level than Java because it supports the programing paradigm of generics, whereas Java in it's current form does not. But then look at something like Python and it's many higher-level features such as dictionaries (associative arrays), generators, or even built-in infinite-precision numbers and imaginary numbers. In those cases the language allows you to directly express those complex concepts that you have to "program" yourself (or use libraries) using lower level languages. And then you can progress up to languages like Haskell and so forth which are higher-level still.
As another example, it should be obvious that within it's intended problem domain even a language like SQL is of a higher-level than Java, and SQL is still just of some intermediate level. Even some generally unpopular old languages have some high-level features not found in Java/C/C++ like Scheme's continuations or COBOL's PIC formatting or Fortran's matrix arithmetic. I know you can program those in Java, but those high-level concepts are not directly provided by Java.
In the big picture of all the languages out there Java is decidedly pretty far over on the scale of being low-level. Again level is not a scale of goodness, so don't fall into that misconception and use that term as such.
The ability to electronically submit comments from a single website is new as well as the nice compilation of currently open comment periods, but otherwise this site is not really a big revolution or any indicator of a new form of representative government, nor is it providing information that wasn't already available online. If you check carefully, you'll find that this site is run by the EPA (Environmental Protection Agency) in cooperation with the GPO (Government Printing Office) and NARA (National Archives and Records Administation) among others. In particular the NARA has been the primary mechanism through which the public is informed of all these issues and comment periods.
Access to all this information has already been publically available and searchable mostly through the online Office of the Federal Register. They even run an open daily mailing list to inform the public about all the new documents. So this new website is really just an incremental improvement to their previous services and offerings.
As far as all the questions about what happens to your comments, the fine print says it all:
"The electronic comments you submit directly through the Regulations.gov website are temporarily maintained by EPA before being forwarded once per day to the proper agency. The agency receiving your comment is considered the official custodian of the comment. Your comment will not be considered until it has been properly received by that agency in accordance with the requirements described in the Federal Register notice. Users who want to verify that an agency has received their comment are urged to check directly with that agency."
So really the only change going on here is that the EPA will deliver your comments rather than the USPS. The processing and reading of comments is still the responsibility of each individual department or agency, just as it was before. Usually all comments received during an open comment period are collected and summarized in the Federal Register at the end of the comment period.
And as to the questions about why bills and other congresional items of interest seem to be missing, it should be noted that the Federal Register (the source of most of the information on the website) is primarily responsible for publishing documents from the various departments and agencies, as well as Public Law and Presidential Orders. A sizeable quantity of those documents are for proposed rules or notices, which by their nature invite public comment through a formalized process.
Bills as such are not handled the same way as department or agency proposals are. The Congress gets it's public feedback directly from its members and their offices (or lobbists), not through a mandated formal public comment periods. Access to bills is primarily available through the GPO Catalog of Congress, but you're on your own to get your comments back to them.
The software-portion that you mention is really the font-hinting code. It is written in a special-purpose language which runs on a VM in the font redering engine. Being a program, it is subject to all kinds of IP restritcions. Actually rather than copyright, it is protected by the much stronger patent law. A good reference for the background is the FreeType Patent Page.
As for copyrights, I was under the impression that a typeface could be protected under copyright as it is an artistic expression. Does anybody have a good reference page for copyright?
This is a VERY welcome bit of goodwill by Bitstream, especially considering how IP paranoid most font foundaries usually are. I do hope that they encode the fonts to allow embedding and subsetting (as many "free" fonts in the past have inadvertantly dissallowed that). Also I hope these fonts contain the full Unicode repertoire (as much as makes sense), and not just the Latin-1 subset.
But I am still anxiously awaiting Adobe to release free versions of their Base PDF fonts. Adobe always makes a big deal about the PDF format being "open" (albeit completely controlled by them). But the one MAJOR non-open component of PDF are the non-open base fonts! Sure the font metrics, aka AFM files, are free (but they hide them very well in the bowels of their ftp site), but not the font outlines.
Come on Adobe, please follow Bitstream's lead and release your base PDF fonts! You can't claim PDF is open until you release the fonts. (Perhaps the same goes for Postscript which has a larger set of Base/Mandatory fonts?)
Why does it seem that all these benchmarks are primarily concerned with CPU performance or network throughput or single-disk reading and writing? For a large category of enterprise applications (which this paper says it is trying to address), I/O performance can usually be the most important part.
The problem is that the typical PC hardware is just not designed for that. Large proprietary Unix or mainframe systems usually have multiple very high speed buses; a single 32-bit PCI bus is rather low-end in comparison. Now of course this is not Linux's fault; but then again Linux is not just a PC operating system! So I guess my question is if this is about benchmarking Linux for enterprise use, how about some information about Linux running on enterprise-class hardware rather than suped up PC's. I'm sure IBM must have a few resources there.
In particular I'm interested in how the Linux kernel is designed to handle multiple independent I/O buses. Are the I/O schedulers weighted down with locking issuesor interrupt contention. Or what about the allocation of memory buffers between faster and slower I/O devices. Or even it's support for advisory I/O operations (hinting) that some proprietary OS's provide? What about asynchronous I/O?
And of course Linux suffers from the general Unix philosophy when it come's to giving I/O the same level of attention as CPU. For instance there are lots of processor use controls, such as process nice levels, processor affinities, real-time schedulers, threading options galore, etc. But how do you say that a given process may only use 30% of the I/O bandwidth on a particular bus? And those are things that mainframes were good at, so how does Linux on mainframes compare?
I'm all for choice and openness, and if these front-ends bring more people to the wonderful world of nethack then all the better.
But if you have not weened yourself from the graphics to give the ASCII text mode a serious try then you are really depriving yourself of a lot of fun. And yes, after just a little period of adjustment the text only mode is actually a lot easier. After all there are over 320 different creature types in just the standard Nethack; using just letters and a few symbols you can readily recognize 58 categories, and then throw in color text and other symbols for objects and it's much better. Pretty graphics are going to almost always be ambiguous and similar looking and disappointing (the sense of scale is always an issue, you have everything from a spider to a dragon). Text is wonderfully expressive...that's why modern languages use small letter-like alphabets rather than artistic pictographs. And it's also why nethack is best played like you're reading a book rather than watching a movie. 3D graphics can never live up to what your imagination dreams up. I don't want to see a picture of a disenchanter thank you, that R will do just fine and lets my imagination do the rest. And yes my heart rate usually jumps when I see an L approaching me! Seriously, if you've been playing the graphical front ends and are starting to get bored, try the text version and use your imagination!
I've been playing nethack for about 15 years and still love it. And yes I am listed in the guidebook, but to be fair I only helped a little bit with the Amiga port a long time ago. I owe the real devteam members a lot of gratitude for all these wonderful years of play.
How does Intel's Hyperthreading Technology differ from the dual core? I realize the obvious, such as one in in the Pentium line and the other in the Itanium, and the physical differences of packaging.
But how how different will the architecture of a dual-die chip differ from hyperthreading, such as which CPU components will be shared (like cache, instruction decode/scheduler, etc.)?
Also would the Linux kernel's logical processor abstraction used to enable hyperthreading support (see IBM developerWorks Article) also continue to work effectively with a dual-die chip?
It's also cached "permamently" on the wonderful WayBack Machine with the latest cached copy on October 2001 here
It will at least be there unless Jim (or lawyers) contact them to have it removed.
Everybody should grab your copies now so this information remains public and useful. And by all means give lots of credit to the hardworking Jim Boemler and just remove the offensive PCI(tm) logo.
And how does it distinguish a DDoS of a nameserver from/. linking to a site, causing it's DNS entry to be looked up by bazillions of people in under 7.3 seconds? There's just not much state in typical DNS traffic for an IDS to analyze. And if an IDS says something look fishy what are you supposed to do, take the nameserver offline?
Not necessarily, it depends upon what you are protecting against. The advantage of ICMP or ICMPv6 (the equivalent layer in IPv6) is that they are very lightweight. There is no expensive crypto operations or other computation, so it is ideal to help protect against DoS floods.
IPv6 can though provide a very secure layer (IPsec) but it comes at an expense. It is not something that you would want to use for DNS queries, where the name of the game is speed and the number of hosts involved can be thousands or even millions.
But for the less voluminous DNS messages, such as zone transfers which occur between mirrors, authenticity is much more of a concern. IPsec could be very useful there, but it is probably unnecessary as DNS already has it's own security protocol built into it (DNSSEC).
In general though IPv6 does provide many benefits over IPv4 and in some ways does provide many new tools to address the DDoS and script kiddies; but like any single technology it is not a super pill that makes all the ills go away.
Any caching system must have a way to update itself or its data will decay and not keep up with changes. Companies change ISP or hosting services all the time, so there DNS entries must be able to change in a timely manner to reflect the IP address changes. Also when a domain name is not renewed it's DNS entries should likewise expire. There are many reasons why an out-of-date cache is bad.
Generally there are two ways to keep caches relatively fresh: expire records based on some precondition (such as time) or have the master source send out notifications when data was changed. And DNS can do BOTH.
First, there are three kinds of expirations in DNS, all time based where the periods are selected by the owner of the domain. The first is when you attempt to look up a name which doesn't exist; that's called negative caching and is typically set to just and hour or two. The next is the refresh time which indicates when an entry in a cache should be checked to see if it is still current and is typically about a half a day. And finally the time-to-live is the time after which the cache entry is forcibly thrown away, and is usually set to a couple weeks or more.
Finally DNS servers can coordinate notification messages, whereby the primary name server for a domain will send a message to any secondaries whenever the data has changed. This allows dirty cache entries to be flushed out almost immediately . But DNS notifications are usually used only between coordinated DNS servers, and not all the way to your home PC.
It should be noted though that most end users' operating systems do not really perform DNS caching very well if at all...usually it is your ISP that is doing the caching. Windows users are mostly out of luck unless you are running in a server or enterprise configuration. Linux can very easily run a caching nameserver if you install the package. I don't know what the Macs do by default.
How much do you want to bet that the MS vouchers will be full of those colorful MS holographs and other ultra-paranoid anti-counterfeit measures, and will probably require an activation code to use. I noticed they're even putting holographs on the bottom of mice, even though I've never heard of a pirated mouse. Watch out US Treasury Dept., MS is getting in the business of printing money now!
More seriously though concerning the distribution of these vouchers. How much of the potential claims were for all the silicon valley dot-coms which no longer exist? (hey they may have a new revenue model now!) Or what about all those PC's with Windows preloaded that were ordered from the many CA mail order vendors which were bought by out of state purchasers? Are they entitled to some refund too? Or businesses that bought Windows multiple times, perhaps without even knowing. I truly suspect that very little of the earmarked amount will actually be claimed. And the left-over pretty much goes back into MS's pockets.
I'd rather they keep their money and send me file format specs instead.
How does one actually qualify to make a claim for a voucher, especially if you double or triple purchased a Microsoft license. Most companies have a hard time just knowing what their current licensed products are, not to mention knowing how much they are eligible to claim going back to 1995. I can just imagine the helpline: "okay let the BSA do an audit and they'll tell you what we owe you (er, you owe us!)".
And did anybody else catch that of the unclaimed refunds schools get 1/3, and of that amount half will be for vouchers of MS products only.
Slashdot Mirror
I had long held the same opinion of the paper, that patents are fundamentally just government sanctioned monopolies. But the theory the article presented made me rethink; there is one huge and important difference between a capitalistic monopoly and these artificial monopolies. With patents it is possible to obtain a monopoly with 0% market share!
Essentially patents allow you to get a monopoly on not producing something. You can't do that in the real market-driven world where monopolies must be gained by capturing near 100% of the market. This is also one of the main differences between copyright and patent: I can't copyright something before I write it, but I sure can patent something without ever having to produce anything.
Patents used legitimately as a means to allow an inventor to safely outlay a large initial economic investment are somewhat defensible, but patents which exist soley to prevent anybody from ever acting upon some idea are just evil. And that's what seems to be happening so much today, especially in the area of software patents.
This is such a refreshing article which finally attempts to put good economic theory to work rather than the extremes on both sides: artists will dies f starvation, or IP will make it illegal to think.
I particularly found interesting the concept of rent-seeking bahvior,
I sure would like to have seen a deeper exploration of that theory, as I feel that is where the most problems lie with the whole IP issue. Consider for instance the cost of the patent itself. The article does a good job of analyzing the cost of R&D; the initial investment in technology. But it doesn't talk about patents in their own as an object of value. How many IP-hoarding companies exist soley to accumulate patents, and have never made any investment into research or innovation?
The patent has an inherent value separate from that of the technology that it may describe.
I would also like to see more thought about the issue of the interaction of patents with each other. The article seems to concentrate solely on one idea at a time...that each patent or IP instrument exists in isolation of all others. It does not deal with the deadlock of interdependent patents. When company A has a patent on idea X and company B has a patent on idea Y, it becomes an impediment to invent idea Z which is created from both. And in the real world, this deadlocking of ideas is out of control.
And finally, the article does not consider fully the costs imposed by the patent or IP system itself. Especially since patents are written in such a manner as to be maximally ambiguous and stealthy it becomes an extreme economic burden on any inventor to know if the ideas being developed have already been claimed. Thus why we are seeing such strange imbalances; you know, 10 lawyers for every engineer.
When you use cash you don't have to give out your identity, and short of using witnesses or video surveilence tapes you're pretty anonymous and untrackable. Be careful with gift certificates/cards though as they potentially can be tracked.
For the online world you'd pretty much need to develop some sort of zero-knowledge system; one that involves real-world entities, not just zero-knowledge cryptographic techniques. Something like an indentity-free payment system (there have been many of these experiements), combined with a zero-knowledge separation of the inventory system from the shipping system.
Of course that only covers you after the fact. If the government is able to set up things before your purchase (like wiretaps and other spy-stuff), then you will get caught.
Especially when the warning reads: Secured for your protection.
The industry always likes to reverse the meaning of all important words to make something bad sound good. Remember SDMI or "Secure" media. All warnings that the product is so secure that you can't even use it!
That's why I'm still playing NetHack...well that and because it's still the best game out there.
Seriously though, can these super cards be used for anything other than the generation of display output? As they are doing so much 3D processing so much faster than any CPU can, I'd like to see the ability to use these GPU's as coprocessors for rendering images back to software/files rather than just to display output. Something like using it as a hardware accelerator for POV-Ray or Renderman. Does anybody have any insight into potential non-traditional uses of these super cards?
AltaVista used to be the best search engine; it's strength lies in basic text searching and it's incredible speed and scalability. Unfortunately it did not account much for the interlinked nature of the web and was easily subverted by web author tricks. These faults were mostly solved by Google.
However, just as Google offers a stand-alone embedded box, the Google Appliance, for use within corporate intranets, I suspect that is an area where AltaVista's technology could thrive much better.
Intranet searching and indexing is still a rather underexploited market. There's basically Microsoft's Index Server, flaws and all, the Google Appliance, and several good but not great minor choices such as ht://Dig. If we could get an AltaVista appliance that ran under Unix (or at least not bound to Microsoft) and underpriced the Google Appliance I would have to believe that a lot of companies would take notice.
When will the myth ever die? Java is not a high-level language. Neither is C nor C++. They are all categorized as low-level languages. Being high or low-level has nothing to do with how "good" a language is or how much hype or popularity or evangelism it has. Go study the theory of programming languages.
The level of a language has to do with the expressiveness of the paradigms (concepts) you can use directly in the language. In this regard it can be easily argued that C++ is of a higher-level than Java because it supports the programing paradigm of generics, whereas Java in it's current form does not. But then look at something like Python and it's many higher-level features such as dictionaries (associative arrays), generators, or even built-in infinite-precision numbers and imaginary numbers. In those cases the language allows you to directly express those complex concepts that you have to "program" yourself (or use libraries) using lower level languages. And then you can progress up to languages like Haskell and so forth which are higher-level still.
As another example, it should be obvious that within it's intended problem domain even a language like SQL is of a higher-level than Java, and SQL is still just of some intermediate level. Even some generally unpopular old languages have some high-level features not found in Java/C/C++ like Scheme's continuations or COBOL's PIC formatting or Fortran's matrix arithmetic. I know you can program those in Java, but those high-level concepts are not directly provided by Java.
In the big picture of all the languages out there Java is decidedly pretty far over on the scale of being low-level. Again level is not a scale of goodness, so don't fall into that misconception and use that term as such.
Access to all this information has already been publically available and searchable mostly through the online Office of the Federal Register. They even run an open daily mailing list to inform the public about all the new documents. So this new website is really just an incremental improvement to their previous services and offerings.
As far as all the questions about what happens to your comments, the fine print says it all:
So really the only change going on here is that the EPA will deliver your comments rather than the USPS. The processing and reading of comments is still the responsibility of each individual department or agency, just as it was before. Usually all comments received during an open comment period are collected and summarized in the Federal Register at the end of the comment period.
And as to the questions about why bills and other congresional items of interest seem to be missing, it should be noted that the Federal Register (the source of most of the information on the website) is primarily responsible for publishing documents from the various departments and agencies, as well as Public Law and Presidential Orders. A sizeable quantity of those documents are for proposed rules or notices, which by their nature invite public comment through a formalized process.
Bills as such are not handled the same way as department or agency proposals are. The Congress gets it's public feedback directly from its members and their offices (or lobbists), not through a mandated formal public comment periods. Access to bills is primarily available through the GPO Catalog of Congress, but you're on your own to get your comments back to them.
The software-portion that you mention is really the font-hinting code. It is written in a special-purpose language which runs on a VM in the font redering engine. Being a program, it is subject to all kinds of IP restritcions. Actually rather than copyright, it is protected by the much stronger patent law. A good reference for the background is the FreeType Patent Page.
As for copyrights, I was under the impression that a typeface could be protected under copyright as it is an artistic expression. Does anybody have a good reference page for copyright?
This is a VERY welcome bit of goodwill by Bitstream, especially considering how IP paranoid most font foundaries usually are. I do hope that they encode the fonts to allow embedding and subsetting (as many "free" fonts in the past have inadvertantly dissallowed that). Also I hope these fonts contain the full Unicode repertoire (as much as makes sense), and not just the Latin-1 subset.
But I am still anxiously awaiting Adobe to release free versions of their Base PDF fonts. Adobe always makes a big deal about the PDF format being "open" (albeit completely controlled by them). But the one MAJOR non-open component of PDF are the non-open base fonts! Sure the font metrics, aka AFM files, are free (but they hide them very well in the bowels of their ftp site), but not the font outlines.
Come on Adobe, please follow Bitstream's lead and release your base PDF fonts! You can't claim PDF is open until you release the fonts. (Perhaps the same goes for Postscript which has a larger set of Base/Mandatory fonts?)
Why does it seem that all these benchmarks are primarily concerned with CPU performance or network throughput or single-disk reading and writing? For a large category of enterprise applications (which this paper says it is trying to address), I/O performance can usually be the most important part.
The problem is that the typical PC hardware is just not designed for that. Large proprietary Unix or mainframe systems usually have multiple very high speed buses; a single 32-bit PCI bus is rather low-end in comparison. Now of course this is not Linux's fault; but then again Linux is not just a PC operating system! So I guess my question is if this is about benchmarking Linux for enterprise use, how about some information about Linux running on enterprise-class hardware rather than suped up PC's. I'm sure IBM must have a few resources there.
In particular I'm interested in how the Linux kernel is designed to handle multiple independent I/O buses. Are the I/O schedulers weighted down with locking issuesor interrupt contention. Or what about the allocation of memory buffers between faster and slower I/O devices. Or even it's support for advisory I/O operations (hinting) that some proprietary OS's provide? What about asynchronous I/O?
And of course Linux suffers from the general Unix philosophy when it come's to giving I/O the same level of attention as CPU. For instance there are lots of processor use controls, such as process nice levels, processor affinities, real-time schedulers, threading options galore, etc. But how do you say that a given process may only use 30% of the I/O bandwidth on a particular bus? And those are things that mainframes were good at, so how does Linux on mainframes compare?
I'm all for choice and openness, and if these front-ends bring more people to the wonderful world of nethack then all the better.
But if you have not weened yourself from the graphics to give the ASCII text mode a serious try then you are really depriving yourself of a lot of fun. And yes, after just a little period of adjustment the text only mode is actually a lot easier. After all there are over 320 different creature types in just the standard Nethack; using just letters and a few symbols you can readily recognize 58 categories, and then throw in color text and other symbols for objects and it's much better. Pretty graphics are going to almost always be ambiguous and similar looking and disappointing (the sense of scale is always an issue, you have everything from a spider to a dragon). Text is wonderfully expressive...that's why modern languages use small letter-like alphabets rather than artistic pictographs. And it's also why nethack is best played like you're reading a book rather than watching a movie. 3D graphics can never live up to what your imagination dreams up. I don't want to see a picture of a disenchanter thank you, that R will do just fine and lets my imagination do the rest. And yes my heart rate usually jumps when I see an L approaching me! Seriously, if you've been playing the graphical front ends and are starting to get bored, try the text version and use your imagination!
I've been playing nethack for about 15 years and still love it. And yes I am listed in the guidebook, but to be fair I only helped a little bit with the Amiga port a long time ago. I owe the real devteam members a lot of gratitude for all these wonderful years of play.
How does Intel's Hyperthreading Technology differ from the dual core? I realize the obvious, such as one in in the Pentium line and the other in the Itanium, and the physical differences of packaging.
But how how different will the architecture of a dual-die chip differ from hyperthreading, such as which CPU components will be shared (like cache, instruction decode/scheduler, etc.)?
Also would the Linux kernel's logical processor abstraction used to enable hyperthreading support (see IBM developerWorks Article) also continue to work effectively with a dual-die chip?
It's also cached "permamently" on the wonderful
WayBack Machine with the latest cached copy on October 2001 here
It will at least be there unless Jim (or lawyers) contact them to have it removed.
Everybody should grab your copies now so this information remains public and useful. And by all means give lots of credit to the hardworking Jim Boemler and just remove the offensive PCI(tm) logo.
And how does it distinguish a DDoS of a nameserver from /. linking to a site, causing it's DNS entry to be looked up by bazillions of people in under 7.3 seconds? There's just not much state in typical DNS traffic for an IDS to analyze. And if an IDS says something look fishy what are you supposed to do, take the nameserver offline?
Not necessarily, it depends upon what you are protecting against. The advantage of ICMP or ICMPv6 (the equivalent layer in IPv6) is that they are very lightweight. There is no expensive crypto operations or other computation, so it is ideal to help protect against DoS floods.
IPv6 can though provide a very secure layer (IPsec) but it comes at an expense. It is not something that you would want to use for DNS queries, where the name of the game is speed and the number of hosts involved can be thousands or even millions.
But for the less voluminous DNS messages, such as zone transfers which occur between mirrors, authenticity is much more of a concern. IPsec could be very useful there, but it is probably unnecessary as DNS already has it's own security protocol built into it (DNSSEC).
In general though IPv6 does provide many benefits over IPv4 and in some ways does provide many new tools to address the DDoS and script kiddies; but like any single technology it is not a super pill that makes all the ills go away.
Any caching system must have a way to update itself or its data will decay and not keep up with changes. Companies change ISP or hosting services all the time, so there DNS entries must be able to change in a timely manner to reflect the IP address changes. Also when a domain name is not renewed it's DNS entries should likewise expire. There are many reasons why an out-of-date cache is bad.
Generally there are two ways to keep caches relatively fresh: expire records based on some precondition (such as time) or have the master source send out notifications when data was changed. And DNS can do BOTH.
First, there are three kinds of expirations in DNS, all time based where the periods are selected by the owner of the domain. The first is when you attempt to look up a name which doesn't exist; that's called negative caching and is typically set to just and hour or two. The next is the refresh time which indicates when an entry in a cache should be checked to see if it is still current and is typically about a half a day. And finally the time-to-live is the time after which the cache entry is forcibly thrown away, and is usually set to a couple weeks or more.
Finally DNS servers can coordinate notification messages, whereby the primary name server for a domain will send a message to any secondaries whenever the data has changed. This allows dirty cache entries to be flushed out almost immediately . But DNS notifications are usually used only between coordinated DNS servers, and not all the way to your home PC.
It should be noted though that most end users' operating systems do not really perform DNS caching very well if at all...usually it is your ISP that is doing the caching. Windows users are mostly out of luck unless you are running in a server or enterprise configuration. Linux can very easily run a caching nameserver if you install the package. I don't know what the Macs do by default.
How much do you want to bet that the MS vouchers will be full of those colorful MS holographs and other ultra-paranoid anti-counterfeit measures, and will probably require an activation code to use. I noticed they're even putting holographs on the bottom of mice, even though I've never heard of a pirated mouse. Watch out US Treasury Dept., MS is getting in the business of printing money now!
More seriously though concerning the distribution of these vouchers. How much of the potential claims were for all the silicon valley dot-coms which no longer exist? (hey they may have a new revenue model now!) Or what about all those PC's with Windows preloaded that were ordered from the many CA mail order vendors which were bought by out of state purchasers? Are they entitled to some refund too? Or businesses that bought Windows multiple times, perhaps without even knowing. I truly suspect that very little of the earmarked amount will actually be claimed. And the left-over pretty much goes back into MS's pockets.
I'd rather they keep their money and send me file format specs instead.
How does one actually qualify to make a claim for a voucher, especially if you double or triple purchased a Microsoft license. Most companies have a hard time just knowing what their current licensed products are, not to mention knowing how much they are eligible to claim going back to 1995. I can just imagine the helpline: "okay let the BSA do an audit and they'll tell you what we owe you (er, you owe us!)".
And did anybody else catch that of the unclaimed refunds schools get 1/3, and of that amount half will be for vouchers of MS products only.