I can just imagine the day when OCR becomes reliable enough to read books of code... then books will be software!
It already is reliable enough, especially when the book is printed with that specific objective in mind, as was the case with the EFF's book Cracking DES.
The print is in an OCR-friendly font, with special provisions for whitespace, and with checksums. But it's still entirely human-readable and expressive.
The tools for both printing code in that format and decoding it once it's been OCR'd are available for FTP from the directory ftp://ftp.pgpi.com/pub/pgp/5.5/books/
What's the bet that the US government knew that the export laws could be appealed on constitutional grounds, but went ahead and got all these countries (including my own, Australia) to sign the Wassenar agreement.
If the Wassenar Agreement is a Treaty (is it?) that the US signs and ratifies, it will have the force of constitutional law here. This makes it completely unclear as to how it would fare in a showdown with the First Amendment.
So... now the US can quite legimately claim that it can't honour the agreement because of the constitutional appeal, giving US software companies an advantage over all the poor countries that were duped into signing over their rights
I think a much more credible theory would be that they wanted to get Wassenar signed before the export regs were once again found unconstitutional, so that they could cite a new source of authority in an appeal.
Your suggestion that the US Administration cares about the well-being of US software companies is dubious. They care about it only in so far as it doesn't interfere with their intelligence-gathering capabilities.
The real issue is protecting communication among people. All we want is to protect our communications, domestic or international.
I dispute that there is a single real issue. But what you bring up is certainly a valid issue. And the Ninth Circuit Court of Appeals actually observed that; they stated:
the government's efforts to retard progress in cryptography may implicate the Fourth Amendment, as well as the right to speak anonymously
The court does not take these issues lightly, and has demonstrated that they (unlike Congress) are willing to study the issues in depth rather then making knee-jerk responses.
Upgradeable? Officially, probably not. But in practice, what I think you'll find inside this box is a standard laptop IDE drive, and a small adapter card with a "tailgate" chip (Symbios SYM13FW500 or equivalent). So you probably could replace the drive. The box is only 17mm tall, so at most there's only room for a 12.5mm tall drive, or maybe only a 9.5mm or 8.5mm. They can't suck enough power from 1394 for a huge drive; even their external power supply is probably not sufficient.
It would be nice to find a cheap source for tailgate adapters. Last time I checked, Symbios (now absorbed by LSI Logic) wanted several thousand dollars for the development kit, so I didn't buy one. But the actual chips are under $10, and only need a handful of inexpensive passive components to make them work.
This makes me think they just glued some cheap IDE drives to the FireWire port.
Yes, that's exactly what it is. A laptop IDE drive with a tailgate chip (Symbios SYM13FW500 or equivalent) tacked on.
AFAIK, none of the drive vendors are mass producing drives with native 1394 interfaces yet. It's a catch-22; not enough computers have a 1394 port to justify making devices, and there are not enough devices to justify making 1394 a standard interface on computers. Apple and Sony are pushing it, but will they be enough?
We have Intel to blame for this. They claimed that they would support both USB and 1394 in their chip sets, so that both would become ubiquitous. They put USB in, and that has been fairly successful. Now it's hard to buy a computer without a USB interface. But they forgot 1394, probably because they want to push USB 2.
SCSI cabling is such a crock that I'd really like to move all my SCSI devices (such as scanner, DAT drive, and 8x CDR) to 1394. It's too bad there's not a single-chip adapter for SCSI devices like there is for IDE.
The following comments reflect my experiences with Red Hat vs. several other Linux distributions, FreeBSD, NetBSD, and commercial operating systems. I have not tried Debian.
1. Timely security updates don't always happen.
They almost always happen in a more timely manner for Red Hat than any other distribution (perhaps excluding Debian, I wouldn't know).
2. A history of upgrade problems. Every distribution has had upgrade difficulties at some point, of course, but it seems like RH has had more than its share.
I disagree. Red Hat upgrades have required less effort on my part than other distributions I have tried (although I haven't tried Debian). Some releases have had quite a bit of errata, but applying the errata was trivial.
3. Immature bug tracking. Red Hat did not have a public bug tracking system before bugzilla was released.
So? Many distributions still don't have public bug tracking. In what possible way can the fact that Red Hat does have it be construed as a disadvantage?
4. Lack of update tools. e.g.: no single, approved command to check for and install any "errata" patches over the internet.
Your use of the words "single" and "approved" seem intended to deliberately cloud the issue. Does Red Hat include any "single, approved" commands for anything? Do any distributions? Who approves these things?
It's not clear to me that automatically installing errata is a good thing. I make it a point to read the errata carefully to decide whether to install things.
Nevertheless, there are readily available tools that do exactly what you want. And if you don't like the various contributed programs, it can be done in three lines from the shell:
wget -m ftp://updates.redhat.com/5.2/alpha/ for p in *.rpm; do rpm -q `rpm -qp --queryformat '%{name}' $p` && pl="$pl $p"; done rpm -U $pl
Mind the ticks and backticks. If you're not sufficiently familiar with RPM to do things like this, put it in a script and preface it with a '#!/bin/sh' line.
5. Frequently inaccessible ftp server, hard to locate list of mirrors.
Firstly, you should be using a mirror. And secondly, the mirrors are trivial to find. It took me less than thirty seconds to find the list of mirror sites
6. $99 for a proprietary secure web server.
When there is a competing non-proprietary secure web server that can be freely distributed in the US for commercial use, then you might have some grounds for complaining about this. I don't want to be US-centric, but the reality is that currently Red Hat products have to be packaged for sale in the US (and, in the case of secure web servers, only in the US, due to our stupid laws).
There seem to be two different issues involved here, privacy and information ownership. They are not the same issue, although they are related.
With regard to privacy, it has never been the case that domain registration information was private. It makes some sense for this data to be a matter of public record; the purpose of collecting the data is to provide a means of contacting the people responsible for a domain. Since DNS is a public resource, and no one is forced to register a domain, there is some justification for the data being publicly available in order to facilitate the proper management of the public resource. I realize that there are also strong arguments to be made in favor of privacy. However, this is not my major concern. It is easy enough to register a domain with a mail drop and a voice mailbox as the contact information. Or some people provide entirely false information. I don't think that is a good idea, but it can be done. Personally I am not in favor of supporting anonymous domain registration, any more than I would be in favor of anonymous automobile registration.
What I am concerned about is the concept that NSI claims to own the data. It might perhaps seem counterintuitive, but I am claiming that I don't mind NSI giving away my data, yet I object to their selling it. And it's not really even the selling per se that is the problem. It's that since they claim to own it, they aren't really selling it. They are licensing it, and claim that no one else has the right to distribute it.
Effectively they have stolen my data, by licensing it to third parties under different terms than I've previously granted them permission. If they want to actually own my data, to the extent that they can license it as they please, then they had damn well better buy it from me first.
In addition, they appear to be combining the data with other data they obtain from third parties. Yet I am given no opportunity to review this additional data and correct any errors. Others have pointed out that much of the information NSI is selling can be obtained from D&B or other sources. However, when you obtain that information from D&B, it doesn't come pre-packaged with domain name registration information, and that is the crucial difference. NSI may have the right to resell information they have legitimately purchased from another source such as D&B, but they do not have the right to comingle it with my domain registration information, since they don't own the latter.
I feel doubly screwed by this, since I've paid them for this great privilege. What a great business to be in: one group of people pays them $70 each to collect the data, and another group buys the resulting database!
When I signed up for my domain in December of 1995, I don't recall that the domain agreement gave the Internic (or NSI) ownership of my data. The information I provided was for the sole purpose of allowing the Internic to effectively maintain the domain name system under contract to the NSF.
Now they claim that they own the data, and that the database is proprietary such that they can do with it as they will. I think this is a load of crap, and am tempted to have a lawyer draft a cease and desist order demanding that NSI refrain from claiming that the data I've provided to them regarding my domain is proprietary to them, selling that data to a third party, or using the data in any other way not explicity authorized by their contract with the NSF.
A kernel module has to be linked to the kernel in order to work. The process of linking is the creation of a derived work.
Linus has explicitly granted permission (beyond the scope of the GPL) for distribution of binary modules that install into the kernel, provided that they don't require any other kernel modifications.
If you want to distribute a binary-only kernel module for Linux, and you need to modify the standard kernel code to make it work, you're SOL.
We don't want proprietary crap in the kernel, and allowing binary-only modules at all is a big concession; allowing that to be combined with arbitrary hacks would allow anyone to make arbitrary amounts of the kernel proprietary.
any source code control/configuration management system that requires KERNEL MODS to run is right out in my book
And, as has been discussed to death in another article here, if they require kernel mods (rather than just an installable module), they will have to supply the entire source code of the kernel components of their system in order to meet the requirements of the GPL.
I don't believe a judge is ever going to buy the "linux is owned by 1000 copyright holders" argument. I think he/she is more inclined to rule that linux is owned by no copyright holders.
Why do you thing a judge would decide that? Do you think that a judge would deliberately ignore the clear evidence of ownership present in the source files and documentation, and would ignore Linus's statements about ownership? Who do you think that the judge would claim to be the author, and why?
In my limited experience with the judicial system, the courts appear to accept claims of ownership at face value unless there is hard evidence to the contrary.
I can't speak for the FSF, but I don't believe that they oppose information ownership. Rather, they strongly support the rights of the owner, which include the right to make the information publicly available under terms that prohibit others from restricting it.
Richard Stallman certainly tries to encourage people to release GPL'd software, because that benefits society. But supporting Free Software and opposing ownership are not the same concept.
I would really like it if things were this way for more open source projects- you put them under GPL, but also allow licensing of the product in a non-GPL product.
In other words, you are asking for exactly how the GPL already works.
There is absolutely NOTHING in the GPL that precludes the software from being provided by its owner under a different license. Many people have already done this. Examples include Ghostscript and Kaffe. In fact, paragraph 12 of the GPL actually explicitly states this. Of course, the owner is not required to do so.
What the GPL prohibits is someone other than the owner from taking the GPL'd code and releasing it under a different license.
Presumably, as the author of code that you want to offer under multiple licenses, you must surely want this protection?
You might be shocked to hear this but there are a lot of people that get paid to develop "free" software.
Why should I be shocked? I've been there. But the rest of the people that use the resulting software get it for free, so they've gotten "free work out of the developers". My point is that I don't see how trying "to get free work out of the developers" is a sham. Aren't the people who are perpetually asking for new features in Linux, EMACS, the GIMP, etc., trying to do exactly that? Are those people guilty of a sham?
I agree, and I'd take that one step further: we should be glad that there are people willing to pay $50 for something they can download for free or buy elsewhere for $2, because that's what keeps Red Hat in business.
I'm actually one of the lazy idiots that pays $50 for each new release of Red Hat. In fact, I pay $100 since I buy both Intel and Alpha versions. But my real motivation isn't that I'm lazy; I often download the release when it first becomes available for FTP, and then buy the official box later. Why? Because I want to support Red Hat; I think that they are doing great stuff, and that they deserve the money. The fact that they give away their code and documentation under the GPL makes me more inclined to buy their product, not less.
I may be an idiot, but the people getting the Red Hat distribution for $2 or for free should thank me.
No, the 20th Century is from 1-Jan-1901 through 31-Dec-2000. The 21st Century begins on 1-Jan-2001.
There was no year 0 A.D.; the 1st Century A.D. was from the year 1 to 100, inclusive. The 1st Century A.D. had to start with the first year A.D., which we write as 1 A.D.
Someone wrote in to PC magazine a while back about how stupid this was, and that common sense shows that the century starts with 2000. He proceded to give an argument that disproved his own point; he said something to the effect of "Everyone would agree that your 100th penny is part of your first dollar." That is, of course, entirely true, which means that your 2nd dollar starts with your 101st penny. Centuries work in exactly the same way.
This in no way precludes having a great party on 31-Dec-1999. I fully plan to have a huge party on that date, to celebrate the "rolling over of the odometer", as it were, and another huge party a year later, to celebrate the end of the century.
Is IBM going to share its technology with other companies like Intel, Sun, and SGI.
IBM's normal policy is to make licenses to their patents available at reasonable and nondiscriminatory terms. This was a result of the consent decree from one of the antitrust suits.
IBM publishes an amazing amount of research results, not all of which even get patented, so they effectively make a lot of things available free.
I see that Slashdot still lets just any idiot post here.
No, I most certainly did not pull them out of any part of my anatomy.
If you'd bother to actually read what I posted before you decide to spout off, and maybe even think about it for a moment (yes, I know, that's asking an awful lot), you would have noticed that:
Those numbers were very explicitly given in megabits per second, not megabytes per second.
The numbers were for the raw transfer rate from the media to the heads (inside the drive), not across the bus.
The numbers came from the manufacturers specifications for the drives
So please crawl back under whatever rock you came from.
Ever used some of those new DVD-ROm drives. they read just as fast as a hard drive.
No, they don't even come close. The drive on my desktop machine isn't the latest thing Seagate makes, but it has a raw transfer rate (from the heads) of 137 to 240 Mbit/s. The newest DVD-ROM drives have a raw transfer rate (peak) of under 55 Mbit/s.
The latest Seagate disk have transfer rates ranging up to 308 Mbit/s.
Slashdot Mirror
Jeez, sounds just like the Cuborg 2000 from the Dilbert strip. IIRC, they planned to add ventilation holes to next year's model.
The print is in an OCR-friendly font, with special provisions for whitespace, and with checksums. But it's still entirely human-readable and expressive.
The tools for both printing code in that format and decoding it once it's been OCR'd are available for FTP from the directory ftp://ftp.pgpi.com/pub/pgp/5.5/books/
Your suggestion that the US Administration cares about the well-being of US software companies is dubious. They care about it only in so far as it doesn't interfere with their intelligence-gathering capabilities.
It would be nice to find a cheap source for tailgate adapters. Last time I checked, Symbios (now absorbed by LSI Logic) wanted several thousand dollars for the development kit, so I didn't buy one. But the actual chips are under $10, and only need a handful of inexpensive passive components to make them work.
AFAIK, none of the drive vendors are mass producing drives with native 1394 interfaces yet. It's a catch-22; not enough computers have a 1394 port to justify making devices, and there are not enough devices to justify making 1394 a standard interface on computers. Apple and Sony are pushing it, but will they be enough?
We have Intel to blame for this. They claimed that they would support both USB and 1394 in their chip sets, so that both would become ubiquitous. They put USB in, and that has been fairly successful. Now it's hard to buy a computer without a USB interface. But they forgot 1394, probably because they want to push USB 2.
SCSI cabling is such a crock that I'd really like to move all my SCSI devices (such as scanner, DAT drive, and 8x CDR) to 1394. It's too bad there's not a single-chip adapter for SCSI devices like there is for IDE.
It's not clear to me that automatically installing errata is a good thing. I make it a point to read the errata carefully to decide whether to install things.
Nevertheless, there are readily available tools that do exactly what you want. And if you don't like the various contributed programs, it can be done in three lines from the shell:
wget -m ftp://updates.redhat.com/5.2/alpha/
for p in *.rpm; do rpm -q `rpm -qp --queryformat '%{name}' $p` && pl="$pl $p"; done
rpm -U $pl
Mind the ticks and backticks. If you're not sufficiently familiar with RPM to do things like this, put it in a script and preface it with a '#!/bin/sh' line.
Firstly, you should be using a mirror. And secondly, the mirrors are trivial to find. It took me less than thirty seconds to find the list of mirror sites When there is a competing non-proprietary secure web server that can be freely distributed in the US for commercial use, then you might have some grounds for complaining about this. I don't want to be US-centric, but the reality is that currently Red Hat products have to be packaged for sale in the US (and, in the case of secure web servers, only in the US, due to our stupid laws).With regard to privacy, it has never been the case that domain registration information was private. It makes some sense for this data to be a matter of public record; the purpose of collecting the data is to provide a means of contacting the people responsible for a domain. Since DNS is a public resource, and no one is forced to register a domain, there is some justification for the data being publicly available in order to facilitate the proper management of the public resource. I realize that there are also strong arguments to be made in favor of privacy. However, this is not my major concern. It is easy enough to register a domain with a mail drop and a voice mailbox as the contact information. Or some people provide entirely false information. I don't think that is a good idea, but it can be done. Personally I am not in favor of supporting anonymous domain registration, any more than I would be in favor of anonymous automobile registration.
What I am concerned about is the concept that NSI claims to own the data. It might perhaps seem counterintuitive, but I am claiming that I don't mind NSI giving away my data, yet I object to their selling it. And it's not really even the selling per se that is the problem. It's that since they claim to own it, they aren't really selling it. They are licensing it, and claim that no one else has the right to distribute it.
Effectively they have stolen my data, by licensing it to third parties under different terms than I've previously granted them permission. If they want to actually own my data, to the extent that they can license it as they please, then they had damn well better buy it from me first.
In addition, they appear to be combining the data with other data they obtain from third parties. Yet I am given no opportunity to review this additional data and correct any errors. Others have pointed out that much of the information NSI is selling can be obtained from D&B or other sources. However, when you obtain that information from D&B, it doesn't come pre-packaged with domain name registration information, and that is the crucial difference. NSI may have the right to resell information they have legitimately purchased from another source such as D&B, but they do not have the right to comingle it with my domain registration information, since they don't own the latter.
I feel doubly screwed by this, since I've paid them for this great privilege. What a great business to be in: one group of people pays them $70 each to collect the data, and another group buys the resulting database!
Now they claim that they own the data, and that the database is proprietary such that they can do with it as they will. I think this is a load of crap, and am tempted to have a lawyer draft a cease and desist order demanding that NSI refrain from claiming that the data I've provided to them regarding my domain is proprietary to them, selling that data to a third party, or using the data in any other way not explicity authorized by their contract with the NSF.
Maybe they didn't crack the satellite; maybe they cracked the Reuters wire service?
A kernel module has to be linked to the kernel in order to work. The process of linking is the creation of a derived work.
Linus has explicitly granted permission (beyond the scope of the GPL) for distribution of binary modules that install into the kernel, provided that they don't require any other kernel modifications.
If you want to distribute a binary-only kernel module for Linux, and you need to modify the standard kernel code to make it work, you're SOL.
We don't want proprietary crap in the kernel, and allowing binary-only modules at all is a big concession; allowing that to be combined with arbitrary hacks would allow anyone to make arbitrary amounts of the kernel proprietary.
In my limited experience with the judicial system, the courts appear to accept claims of ownership at face value unless there is hard evidence to the contrary.
Richard Stallman certainly tries to encourage people to release GPL'd software, because that benefits society. But supporting Free Software and opposing ownership are not the same concept.
There is absolutely NOTHING in the GPL that precludes the software from being provided by its owner under a different license. Many people have already done this. Examples include Ghostscript and Kaffe. In fact, paragraph 12 of the GPL actually explicitly states this. Of course, the owner is not required to do so.
What the GPL prohibits is someone other than the owner from taking the GPL'd code and releasing it under a different license.
Presumably, as the author of code that you want to offer under multiple licenses, you must surely want this protection?
I'm actually one of the lazy idiots that pays $50 for each new release of Red Hat. In fact, I pay $100 since I buy both Intel and Alpha versions. But my real motivation isn't that I'm lazy; I often download the release when it first becomes available for FTP, and then buy the official box later. Why? Because I want to support Red Hat; I think that they are doing great stuff, and that they deserve the money. The fact that they give away their code and documentation under the GPL makes me more inclined to buy their product, not less.
I may be an idiot, but the people getting the Red Hat distribution for $2 or for free should thank me.
There was no year 0 A.D.; the 1st Century A.D. was from the year 1 to 100, inclusive. The 1st Century A.D. had to start with the first year A.D., which we write as 1 A.D.
Someone wrote in to PC magazine a while back about how stupid this was, and that common sense shows that the century starts with 2000. He proceded to give an argument that disproved his own point; he said something to the effect of "Everyone would agree that your 100th penny is part of your first dollar." That is, of course, entirely true, which means that your 2nd dollar starts with your 101st penny. Centuries work in exactly the same way.
This in no way precludes having a great party on 31-Dec-1999. I fully plan to have a huge party on that date, to celebrate the "rolling over of the odometer", as it were, and another huge party a year later, to celebrate the end of the century.
Reference: the specifications on the offical Palm website
IBM publishes an amazing amount of research results, not all of which even get patented, so they effectively make a lot of things available free.
No, I most certainly did not pull them out of any part of my anatomy.
If you'd bother to actually read what I posted before you decide to spout off, and maybe even think about it for a moment (yes, I know, that's asking an awful lot), you would have noticed that:
So please crawl back under whatever rock you came from.
The latest Seagate disk have transfer rates ranging up to 308 Mbit/s.