No sexism implied here, I think. Throughout this whole series, I've been looking at these books from the point of view of my mother. Could someone who's used a computer for simple tasks for a few years learn how to use Linux from a book?
She's in the market for a new computer, and she wants Linux on it. (Now that Gnucash does everything she'd been using Quicken for, the only thing standing in the way is the OfficeJet.)
I've given her the RedHat book to read, and I think she can handle it. Then it's off to another book in this series of articles, because there's so much more to learn.
Thus, the reason for the feminine pronoun in this case is because I had a specific example in mind.
I think some people are critical of NVIDIA because the company can't quite seem to figure out what to do. It said it supports Linux, but couldn't seem to release decent drivers until pressured by a handful of articles on Slashdot, Linuxgames, and other sites. It was caught using GPL'd code in the driver, but removed it faster than it said it could. It uses a unified driver architecture so that improvements to the Windows driver show up in the Linux driver, but a marketing manager goes on the record as saying "The only reason anyone would open source something is because they can't do a good job of it on their own."
It get "caught" strongarming volunteer sites by sending out review hardware and then calling in the chips to get rid of information about competitors -- then explains it away by claiming it hires temporary workers with the authority to make those deals but not the oversight not to.
Personally, I think it's a normal company with good hardware that needs to get rid of the marketing department and give the technical people more control.
Read more at my NVIDIA history or analysis pages. I don't really understand what the company's doing, but at least my video card works.
I've had a handful of dealings with NVIDIA in the past. After a few people started reading my saga (including some folks at VA Linux -- which has a partnership with the company), a product manager contacted me.
He offered to let me in on beta testing their new Linux drivers, and offered to send me a shiny new GeForce for testing. A couple of days later, the beta drivers went fairly public. There are still internal betas to which I, in theory, have access... but I haven't heard from the fellow since.
I haven't seen the card, either, and it's been a few months.
Was it a bribe for me to take down the pages on my site critical of NVIDIA? I don't know. It certainly wasn't specified in the e-mails -- the card was coming so I could "publish updated benchmarks".
Don't get me wrong, giving the company the benefit of the doubt is kinda painful. Just remember that there are people working there who are tremendously helpful (Terrence Ripperda and the other official folks in #nvidia on irc.openprojects.net) and sympathetic to our concerns. Heck, they even got rid of the GPL violations in three days (instead of the weeks they said when the story originally broke).
Still... I can't immediately dismiss these reports.
Sounds like a quantum twist on the philosophy of George Berkeley. (His idea was a sort of idealism that stated, "Something must be perceived in order to exist." If God ever closed His eyes, the whole universe would disappear.)
But, part of that is standard rhetoric. If you pepper your argument with phrases like "I believe" and "It may be true that" and "This is probably how it works", you come off as wishy-washy. ESR's probably used to speaking directly, focusing more on his ideas than the supporting evidence.
(Whew, talk about a backhanded compliment.)
Anyway, about half the time he says something in public, I want to stand up and say "This man does not speak for me on these specific issues." It happened when he said that Linux users don't want China to use Linux. It happened when he talked about being unbelievably wealthy after the LNUX IPO.
It happened today, when, to the best of my knowledge, he argued that government is irrelevant in the face of his anarcho-capitalism. (One might wonder, if the U.S. federal government were judged on purely economic principles -- revenue, spending, employment statistics -- how would it compare to the corporations ESR mentions as examples?) I disagree.
I prefer Lessig's argument that giving up on The System (or denying that it exists, or, possibly, waiting to revolt against it and set up some kind of technocracy) would be a mistake. Coding skills, project administration ability, and even soi-disant expert sociological analysis are separate skills from governance. Some people might be gifted in both areas, but if you only have abilities in one, it's a grave fallacy to assume that the other is irrelevant. Both sides need to learn this lesson.
...but I agree with you and Steve on this one. Those questions are supposed to be rhetorical.
What I meant by that section is that the more traditional texts I've read have tried to explain C++ by teaching its C subset first. That includes those pitfall-ridden constructs that require strict error checking. Students who've learned from those texts might ask, "Why change? Why not teach the old basics we had to learn first?"
I think it's to the author's credit that he gets students used to classes that don't need bounds-checking, instead of introducing the old way to do things, then saying in an Appendix "But don't do it that way!".
All I said was, I hope it is marked as "Katz" so I don't have to read it.
You don't, no matter how it's marked. Unless someone clicks on "Read More" for you, the article won't come up. (You do realize that it is a suggestion, not a command?)
If it upsets you that some people don't want to read Katz, then you honestly have an problem you need to deal with.
On the other side, some people seem to hate his articles so much that simply seeing a link to one on the front page causes them distress. That doesn't sound normal to me.
But the powers that be have given you additional freedom. Hooray!
The whole point of Perl is to make it possible to write both ways, as well as up-and-down and sideways. I can write very quick and very dirty code that does a job fairly well. I can also write elegant code that's clean and understandable by someone of decent Perl ability.
Granted, there is a lot of the former roaming around (stuff that doesn't even compile under the strict pragma and that generates pages of warnings with -w enabled), but there are people like me, who write Perl every day, turning out good and clean code.
Would you say that Stephenson can't write excellent prose in English after reading a few pages of script-kiddie leet-speak in IRC?
Here's a fun new legal technique, similar to the unequivocably moral Unisys patent plan:
Take an open standard.
Add one small incompatibility.
Hide that incompatibility for a few months.
Write a paper describing the incompatibility in sufficient detail, so that implementing the necessary changes is trivial.
Post a warning on the paper that implementing the specification without advance written permission is illegal.
Wrap the paper in some sort of mechanism which presents the warning and a license agreement nominally waiving fair use, reverse engineering, and free speech rights.
Make sure the mechanism only works on platforms where your implementation is already present, and, thus, no clean-room version is necessary.
Distribute liberally, knowing that standard means of unwrapping the document -- on any platform but your own -- will not present the invalid license agreement anyway.
Wait for the information to spread to all interested parties.
Use the threat of legal force to intimidate anyone who might be considering writing a competing implementation -- not necessarily even based on information from the paper -- into cancelling the effort.
Bask in the glow that you have not used dumping, product tying, or buyouts to maintain your monopoly position. Just good old American justice.
As I am not aware of any patent or copyright our friends at Microsoft have on this business process, please be aware that it is copyright chromatic, 2000.
I found Andy Tanenbaum's Operating System Design and Implementation useful for learning the basics. It describes things in a Minix context (as he wrote Minix for a practical example) and has a strong microkernel bias, but it is an effective presentation of the important concepts.
Someone else mentioned the dinosaur book -- I think it's Operating System Concepts -- with case studies of NT and Linux 2.0. That's also good.
When I mentioned that LCKC could be a college text, I had the aforementioned books in mind as possible companions.
All the examples mentioned by Bob (sendmail, Apache) are more or less 'background' applications, outside the realm of lusers.
True... but saying something like, "Do you enjoy being able to view web sites? Thank the open source BIND and Apache. Do you like receiving e-mail? Thank Sendmail. Do you like interactive web sites and polls? Thank Perl and PHP and Python."
Really, 'normal end users' just don't know how much they depend on free software already. If it's good enough to run the Internet, maybe it's good enough to run your business.
Before Starship Titanic came out, there was talk of doing a CD-ROM based game consisting only of sound clips (no graphics, just radio). I see no mention of it on your site.
I hope this project is still in the pipeline -- it sounded very interesting. Is your first love still radio, or were you just looking for something unique (like Bureaucracy, perhaps)?
Don't forget that these drivers were released merely a week ago, catching everyone by surprise. I would expect that a couple of people at NVIDIA spent a couple of very hectic days preparing for this.
It wouldn't surprise me if this code was only there for the internal beta, to be distributed only to people who have NDAs with NVIDIA. The developer might have grabbed the code just to get the beta working, planning to rewrite it before the drivers went out to the general public.
Mistake? Yeah. Worth hate-mail? No. Give the company the chance to do the right thing. They've turned things around in just two weeks (from where I sit, and I have a bit of a perspective on this), so I'm willing to be patient. (Yes, I'll update the page in my.sig in the next day or two, too.)
Playing Angel's Advocate here, bear with me. What if I purchased a used CD from another consumer? Legal. What if she'd made a backup copy for personal use? Legal. What if she gave me the backup copy along with the CD? Super legal.
What if I listened to the backup copy she made? Not legal? Must I make my own backup copy myself for it to be legal, or can I copy or otherwise obtain a backup copy from another consumer who made it, legally, himself?
It seems to me that the fair use principle isn't in place so that we can spend our time ripping CDDA and encoding MP3s, but that, having purchased a license to listen to the music, we can listen to it in other formats.
--
Re:Um... this is a computer virus?
on
Sim Plague
·
· Score: 1
To put it in terms journalists might understand:
I have a picture of my dog stored on my computer. I hope he doesn't get loose and eat my homework.
Hey, let's give the company a chance. This time last week, I wasn't sure we'd see drivers at all until July -- beta or not. Kudos to Jim (et al.) at NVIDIA for working hard to get yesterday's drivers out.
I expect the company will be doing more for and with the free software community... just give them a little time to adjust.
I'm currently attempting to persuade a hardware manufacturer to provide unobfuscated source code and hardware documentation to free driver writers.
In your opinion, what is the best and/or most effective way to go about this? The court of public opinion? Economic arguments? Pointing out the higher quality of free drivers? Or should I just advise people to move to more enlightened hardware manufacturers.
If you're wondering why nVIDIA cards weren't reviewed, I've put up a couple of pages with information about the company and the saga of their "Linux support". The current rumor is that they have binary only drivers for XFree86 4.0 using their own straight-to-hardware pipeline (instead of DRI), and they're not really concerned about Linux users in general.
Otherwise, Microsoft could have argued, "We are required by law to do things that cause the government to prosecute us for breaking anti-trust law or risk being sued by our shareholders."
Yeah, a corporation is a legal fiction masquerading as a person. That doesn't mean the corp. has the right to do illegal or unethical things. Stockholders can also sue if the officers and/or board of directors do stupid, risky things (such as skirting the law).
Now maybe I've dealt with some very bad programmers here at work (and looking at some of the drivers and software they've released, that's a good possibility), but I've run into that very thing right now!
Some applications work only with certain versions of Internet Explorer (various patch levels and such). There are half a dozen numbers you have to check in the Help, About menu before you can be sure the program won't die a flaming death.
I won't even mention the number of hotfixes and upgrades and patches and enhancements and "Oh, you installed another Office Suite application that replaced this or that DLL" that introduced incompatibilities and requires me to keep a list of the software I've installed in the order I installed it.
And I'm only talking about Windows NT, Service Pack three. Don't get me started on the thre or four versions of 95 and at least two versions of 98.
Application developers are already pulling out their hair. They won't notice anything different.
I'll second that. Though it's based on kernel 2.2.5, it has a good blend of design and implementation details -- so you'll be able to follow 2.4 (when it appeaers) without too much trouble.
There's more to say, but I'll save that for the review.:)
I only speak for myself, not the other reviewers, but I *hate* reading bad books and am not likely to finish them, even if I do start them. (chromatic's rule of book covers: if it's colored like a fire truck, or has more exclamation points than recommendations from people I've heard of, it's probably not worth reading).
Slashdot doesn't tell me what to write, and hasn't changed a word of my reviews yet. Even if I don't care for a book, I try to find *some* appropriate audience for it. Besides that, technical book reviews are more interesting for *what* they say than how they say it (unless they're wrong or the writing gets in the way of the information -- which I've seen in a few books). That said, I'm slogging through a couple of books now that
You're less likely to see a review of a zero-star, rushed to market celebrity tell all here than an interesting discussion of the Linux Kernel or the latest Stephenson tour-de-force, simply because those of us who do reviews are not as much interested in the former. (At least, I'm not.)
I pick good books to read? I do positive reviews. Hemos likes positive reviews.
Hemos hasn't published anything from me in a while... uh oh.
I find it ironic that there are so many young people who are confused and hurting, who really need just one or two good adult role models to sit down with them and listen to them for a while...
... and those adults will hear some painful and confused feelings of isolation and loneliness and despair...
... and brand these kids as possible menaces to society.
When I was a teenager, I didn't want people to listen to me because they might be afraid of what I might do. I wanted people to listen to me because they cared about me and could identify with the way I was feeling and the thoughts I was thinking.
Don't alienate young people even further in the guise of helping them. Please.
Slashdot Mirror
No sexism implied here, I think. Throughout this whole series, I've been looking at these books from the point of view of my mother. Could someone who's used a computer for simple tasks for a few years learn how to use Linux from a book?
She's in the market for a new computer, and she wants Linux on it. (Now that Gnucash does everything she'd been using Quicken for, the only thing standing in the way is the OfficeJet.)
I've given her the RedHat book to read, and I think she can handle it. Then it's off to another book in this series of articles, because there's so much more to learn.
Thus, the reason for the feminine pronoun in this case is because I had a specific example in mind.
--
I think some people are critical of NVIDIA because the company can't quite seem to figure out what to do. It said it supports Linux, but couldn't seem to release decent drivers until pressured by a handful of articles on Slashdot, Linuxgames, and other sites. It was caught using GPL'd code in the driver, but removed it faster than it said it could. It uses a unified driver architecture so that improvements to the Windows driver show up in the Linux driver, but a marketing manager goes on the record as saying "The only reason anyone would open source something is because they can't do a good job of it on their own."
It get "caught" strongarming volunteer sites by sending out review hardware and then calling in the chips to get rid of information about competitors -- then explains it away by claiming it hires temporary workers with the authority to make those deals but not the oversight not to.
Personally, I think it's a normal company with good hardware that needs to get rid of the marketing department and give the technical people more control.
Read more at my NVIDIA history or analysis pages. I don't really understand what the company's doing, but at least my video card works.
--
I've had a handful of dealings with NVIDIA in the past. After a few people started reading my saga (including some folks at VA Linux -- which has a partnership with the company), a product manager contacted me.
He offered to let me in on beta testing their new Linux drivers, and offered to send me a shiny new GeForce for testing. A couple of days later, the beta drivers went fairly public. There are still internal betas to which I, in theory, have access... but I haven't heard from the fellow since.
I haven't seen the card, either, and it's been a few months.
Was it a bribe for me to take down the pages on my site critical of NVIDIA? I don't know. It certainly wasn't specified in the e-mails -- the card was coming so I could "publish updated benchmarks".
Don't get me wrong, giving the company the benefit of the doubt is kinda painful. Just remember that there are people working there who are tremendously helpful (Terrence Ripperda and the other official folks in #nvidia on irc.openprojects.net) and sympathetic to our concerns. Heck, they even got rid of the GPL violations in three days (instead of the weeks they said when the story originally broke).
Still... I can't immediately dismiss these reports.
--
Sounds like a quantum twist on the philosophy of George Berkeley. (His idea was a sort of idealism that stated, "Something must be perceived in order to exist." If God ever closed His eyes, the whole universe would disappear.)
--
That was actually Anselm's ontological argument. Read about Anselm here, if you're so inclined.
(chromatic studied a bit of philosophy as an undergraduate)
--
Omnipotence implies the ability to make logical fallacies -literally- true?
You might as well say, "Since there is no God, God must prove the impossibility of His existence to prove His existence."
Your definition of 'omnipotence' needs some work.
--
Yes.
But, part of that is standard rhetoric. If you pepper your argument with phrases like "I believe" and "It may be true that" and "This is probably how it works", you come off as wishy-washy. ESR's probably used to speaking directly, focusing more on his ideas than the supporting evidence.
(Whew, talk about a backhanded compliment.)
Anyway, about half the time he says something in public, I want to stand up and say "This man does not speak for me on these specific issues." It happened when he said that Linux users don't want China to use Linux. It happened when he talked about being unbelievably wealthy after the LNUX IPO.
It happened today, when, to the best of my knowledge, he argued that government is irrelevant in the face of his anarcho-capitalism. (One might wonder, if the U.S. federal government were judged on purely economic principles -- revenue, spending, employment statistics -- how would it compare to the corporations ESR mentions as examples?) I disagree.
I prefer Lessig's argument that giving up on The System (or denying that it exists, or, possibly, waiting to revolt against it and set up some kind of technocracy) would be a mistake. Coding skills, project administration ability, and even soi-disant expert sociological analysis are separate skills from governance. Some people might be gifted in both areas, but if you only have abilities in one, it's a grave fallacy to assume that the other is irrelevant. Both sides need to learn this lesson.
--
What I meant by that section is that the more traditional texts I've read have tried to explain C++ by teaching its C subset first. That includes those pitfall-ridden constructs that require strict error checking. Students who've learned from those texts might ask, "Why change? Why not teach the old basics we had to learn first?"
I think it's to the author's credit that he gets students used to classes that don't need bounds-checking, instead of introducing the old way to do things, then saying in an Appendix "But don't do it that way!".
--
All I said was, I hope it is marked as "Katz" so I don't have to read it.
You don't, no matter how it's marked. Unless someone clicks on "Read More" for you, the article won't come up. (You do realize that it is a suggestion, not a command?)
If it upsets you that some people don't want to read Katz, then you honestly have an problem you need to deal with.
On the other side, some people seem to hate his articles so much that simply seeing a link to one on the front page causes them distress. That doesn't sound normal to me.
But the powers that be have given you additional freedom. Hooray!
--
The whole point of Perl is to make it possible to write both ways, as well as up-and-down and sideways. I can write very quick and very dirty code that does a job fairly well. I can also write elegant code that's clean and understandable by someone of decent Perl ability.
Granted, there is a lot of the former roaming around (stuff that doesn't even compile under the strict pragma and that generates pages of warnings with -w enabled), but there are people like me, who write Perl every day, turning out good and clean code.
Would you say that Stephenson can't write excellent prose in English after reading a few pages of script-kiddie leet-speak in IRC?
--
Here's a fun new legal technique, similar to the unequivocably moral Unisys patent plan:
- Take an open standard.
- Add one small incompatibility.
- Hide that incompatibility for a few months.
- Write a paper describing the incompatibility in sufficient detail, so that implementing the necessary changes is trivial.
- Post a warning on the paper that implementing the specification without advance written permission is illegal.
- Wrap the paper in some sort of mechanism which presents the warning and a license agreement nominally waiving fair use, reverse engineering, and free speech rights.
- Make sure the mechanism only works on platforms where your implementation is already present, and, thus, no clean-room version is necessary.
- Distribute liberally, knowing that standard means of unwrapping the document -- on any platform but your own -- will not present the invalid license agreement anyway.
- Wait for the information to spread to all interested parties.
- Use the threat of legal force to intimidate anyone who might be considering writing a competing implementation -- not necessarily even based on information from the paper -- into cancelling the effort.
- Bask in the glow that you have not used dumping, product tying, or buyouts to maintain your monopoly position. Just good old American justice.
As I am not aware of any patent or copyright our friends at Microsoft have on this business process, please be aware that it is copyright chromatic, 2000.It's also unethical. I'm not surprised.
--
I found Andy Tanenbaum's Operating System Design and Implementation useful for learning the basics. It describes things in a Minix context (as he wrote Minix for a practical example) and has a strong microkernel bias, but it is an effective presentation of the important concepts.
Someone else mentioned the dinosaur book -- I think it's Operating System Concepts -- with case studies of NT and Linux 2.0. That's also good.
When I mentioned that LCKC could be a college text, I had the aforementioned books in mind as possible companions.
--
All the examples mentioned by Bob (sendmail, Apache) are more or less 'background' applications, outside the realm of lusers.
True... but saying something like, "Do you enjoy being able to view web sites? Thank the open source BIND and Apache. Do you like receiving e-mail? Thank Sendmail. Do you like interactive web sites and polls? Thank Perl and PHP and Python."
Really, 'normal end users' just don't know how much they depend on free software already. If it's good enough to run the Internet, maybe it's good enough to run your business.
--
Before Starship Titanic came out, there was talk of doing a CD-ROM based game consisting only of sound clips (no graphics, just radio). I see no mention of it on your site.
I hope this project is still in the pipeline -- it sounded very interesting. Is your first love still radio, or were you just looking for something unique (like Bureaucracy, perhaps)?
--
Don't forget that these drivers were released merely a week ago, catching everyone by surprise. I would expect that a couple of people at NVIDIA spent a couple of very hectic days preparing for this.
It wouldn't surprise me if this code was only there for the internal beta, to be distributed only to people who have NDAs with NVIDIA. The developer might have grabbed the code just to get the beta working, planning to rewrite it before the drivers went out to the general public.
Mistake? Yeah. Worth hate-mail? No. Give the company the chance to do the right thing. They've turned things around in just two weeks (from where I sit, and I have a bit of a perspective on this), so I'm willing to be patient. (Yes, I'll update the page in my .sig in the next day or two, too.)
--
Only a consumer can make a copy of their music
Playing Angel's Advocate here, bear with me. What if I purchased a used CD from another consumer? Legal. What if she'd made a backup copy for personal use? Legal. What if she gave me the backup copy along with the CD? Super legal.
What if I listened to the backup copy she made? Not legal? Must I make my own backup copy myself for it to be legal, or can I copy or otherwise obtain a backup copy from another consumer who made it, legally, himself?
It seems to me that the fair use principle isn't in place so that we can spend our time ripping CDDA and encoding MP3s, but that, having purchased a license to listen to the music, we can listen to it in other formats.
--
To put it in terms journalists might understand:
I have a picture of my dog stored on my computer. I hope he doesn't get loose and eat my homework.
--
Hey, let's give the company a chance. This time last week, I wasn't sure we'd see drivers at all until July -- beta or not. Kudos to Jim (et al.) at NVIDIA for working hard to get yesterday's drivers out.
I expect the company will be doing more for and with the free software community... just give them a little time to adjust.
--
I'm currently attempting to persuade a hardware manufacturer to provide unobfuscated source code and hardware documentation to free driver writers.
In your opinion, what is the best and/or most effective way to go about this? The court of public opinion? Economic arguments? Pointing out the higher quality of free drivers? Or should I just advise people to move to more enlightened hardware manufacturers.
(Thanks for the GNU/Abacus, by the way!)
--
If you're wondering why nVIDIA cards weren't reviewed, I've put up a couple of pages with information about the company and the saga of their "Linux support". The current rumor is that they have binary only drivers for XFree86 4.0 using their own straight-to-hardware pipeline (instead of DRI), and they're not really concerned about Linux users in general.
Link is here.
--
Um, no.
Otherwise, Microsoft could have argued, "We are required by law to do things that cause the government to prosecute us for breaking anti-trust law or risk being sued by our shareholders."
Yeah, a corporation is a legal fiction masquerading as a person. That doesn't mean the corp. has the right to do illegal or unethical things. Stockholders can also sue if the officers and/or board of directors do stupid, risky things (such as skirting the law).
--
Now maybe I've dealt with some very bad programmers here at work (and looking at some of the drivers and software they've released, that's a good possibility), but I've run into that very thing right now!
Some applications work only with certain versions of Internet Explorer (various patch levels and such). There are half a dozen numbers you have to check in the Help, About menu before you can be sure the program won't die a flaming death.
I won't even mention the number of hotfixes and upgrades and patches and enhancements and "Oh, you installed another Office Suite application that replaced this or that DLL" that introduced incompatibilities and requires me to keep a list of the software I've installed in the order I installed it.
And I'm only talking about Windows NT, Service Pack three. Don't get me started on the thre or four versions of 95 and at least two versions of 98.
Application developers are already pulling out their hair. They won't notice anything different.
--
I'll second that. Though it's based on kernel 2.2.5, it has a good blend of design and implementation details -- so you'll be able to follow 2.4 (when it appeaers) without too much trouble.
There's more to say, but I'll save that for the review. :)
--
I only speak for myself, not the other reviewers, but I *hate* reading bad books and am not likely to finish them, even if I do start them. (chromatic's rule of book covers: if it's colored like a fire truck, or has more exclamation points than recommendations from people I've heard of, it's probably not worth reading).
Slashdot doesn't tell me what to write, and hasn't changed a word of my reviews yet. Even if I don't care for a book, I try to find *some* appropriate audience for it. Besides that, technical book reviews are more interesting for *what* they say than how they say it (unless they're wrong or the writing gets in the way of the information -- which I've seen in a few books). That said, I'm slogging through a couple of books now that
You're less likely to see a review of a zero-star, rushed to market celebrity tell all here than an interesting discussion of the Linux Kernel or the latest Stephenson tour-de-force, simply because those of us who do reviews are not as much interested in the former. (At least, I'm not.)
I pick good books to read? I do positive reviews. Hemos likes positive reviews.
Hemos hasn't published anything from me in a while... uh oh.
--
I find it ironic that there are so many young people who are confused and hurting, who really need just one or two good adult role models to sit down with them and listen to them for a while...
When I was a teenager, I didn't want people to listen to me because they might be afraid of what I might do. I wanted people to listen to me because they cared about me and could identify with the way I was feeling and the thoughts I was thinking.
Don't alienate young people even further in the guise of helping them. Please.
--