You have to consider the possibility that there are some people whose way of life is fundamentally incompatable with yours. I believe that anyone capable of organizing the hijacking of several commerical airplanes and using them as weapons against both civillian and military targets in an unprovoked sneak attack is not someone that I can make amends with without making an unacceptable compromise of my integrity.
If possible, pepetrators of this kind of terrorism should be made to see the error of their ways, in the hopes of preventing future incidents. However, if that proves to be impossible, the only recourse is total annihalation.
I am a peaceful person. I would do just about anything to insure that we could all live harmoniously. But if other people force the issue into a "me or them" situation, I pick me, every time, and will defend that decision with ultimate force.
The fabricators are pretty expensive, and the amortized cost of the machine + materials is probably much greater than the replacement parts.
However, consider that an assembly line has thousands of different parts like this. The cost of keeping spares of every part around is much higher than the cost of just the failed parts. If you can just fabricate the part you need on the spot, you reduce downtime vs. waiting for a replacement part. On a high volume assembly line, my guess is the cost of a few of hours less downtime can pay for the prototyping machine.
School is what you make it. Yes, there are good profs and bad profs, and good course and bad courses, but at almost any university, you can make the experience worthwhile and educational, or you can make it a worthless waste of money and time.
I've taught myself way more than most professors.
Yes, that is a good sign. I would go so far as to say that it is impossible for anyone but yourself to teach you anything. The secret to getting a lot out of an education is to think of it in terms of learning, rather than teaching. Every class you take gives you an opportunity to learn something. You probably could learn it yourself, with appropriate books, and enough time, but nowhere near as fast as if you have a professor and a class full of other students to help you.
So, you have to approach your schoolwork with the attitude that you are the boss of you. Nobody is going to make you do things the way the professor tells you too, and if you are willing to accept the consequences, don't do things you don't think are beneficial.
A lot of people, some professors included, probably don't agree with this aproach to learning. They may, in fact, believe that professors are sacred fonts of wisdom that bless students with their presence. But it doesn't matter -- you can learn almost as much from those professors as anyone else, more if you count learning to suffer fools gracefully.
Finally, I have gone to a lot of schools and had a lot of professors/teachers. Some were brilliant, some were idiots, some were wonderful people, some were complete assholes, some I loved, some I hated. But I have never -- repeat NEVER -- met a teacher who wasn't willing to help a student learn, if that student sincerly asked for it.
If you get any less than a stellar education, you have primarily yourself to blame.
Actually, both this solution and LinuxBIOS can get to a command prompt (or, more likely, load an embedded application) in 3 seconds.
The 0.8s to LILO prompt is nice because it demonstrates how fast the BIOS is, but is irrelevant from a system designer standpoint: The important thing is how fast they can get their program running.
In 2.4.X, linux defaults to using the same outgoing port that the originating machine used, then tries sequentially from there, rather than using ports in the 60,000 range (as prior kernel series did).
You could still probably detect it, but A) it would be inconclusive, and B) it wouldn't mean anything if it were. I have two computers behind a NAT based firewall, and lots of people use those linksys (or other) "Cable/DSL routers" that do NAT automatically.
Well, first of all, when your lease times out, you will get switched automatically. Of course, any active connections you have will be dropped, but your connectivity will continue. This will hork up SSH connections, but HTTP should be fine.
Assuming you can get information from the lan card when it switches base stations, you should be able to automate the release/renew cycle (on *ix or Win*). Even if it isn't a clean message, you should be able to monitor the MAC of the active base station on a short enough interval you can switch almost seamlessly.
The guys at O'Reilly working on this have an idea that uses a hash of your MAC as your IP address, so you can keep the same address when you move from cell to cell.
Finally, I believe that if you put point-to-point links between adjacent cells, and substantially increse the sophistication of the gateway software, you should be able to handle forwarding established connections when someone moves from cell to cell. Plus, then you could do bandwidth agregation, if you were really clever, and had a relatively dense network.
Of course, that involves a substantial amount of extra hardware, and correspondingly higher costs.
Hm. Are there any reasonably low cost capture systems that have high quality S-Video replay? I was thinking about doing this sort of thing for archival purposes. I have a TiVo, which I love, but it is a 2 drive 30 hour model, and I don't want to screw it up attempting and upgrade.
I already have scads of unused drive space, I just need the capture card, but I want one that has S-Video in and out at high quality. I couldn't care less about VGA.
IIRC, the TiVo uses an MPEG chipset designed to go with the PowerPC CPU they use, and it isn't available as a consumer board.
Also, how easy/hard is it to cut a mpeg video, say, to remove commercials? It seems to me that roughly every 5th frame is uncompressed, so you ought to be able to slice it on those boundaries.
First of all, there is a limited market for these things. Very few people are willing to pay the extra it would cost to get a tricked out case.
Second, they would be really expensive. If you make these as a hobby, part of the goal is to have fun building it, and in the end you have a unique case that you made yourself. If you make these to sell, you have to charge people for your time, and that bumps the price way up.
In the PC market, you just can't make money off of this sort of thing. The only place it works is when you have a single supplier of something. Hence, the distinctive looks of Apples, SGIs, NeXT, and Cray computers (to name a few). You buy the hardware you want and automatically get a nice looking case to go with it (depending on your tastes).
The other factor is availability of peripherals. It is difficult for an end user to even find black CD-ROMs, much less ones to go with a specific case design, or other peripherals that match their color scheme. Again, the only way this works is with a single supplier, or if you are going to do the mods yourself. Witness how you can get a keyboard, mouse, USB hub, scanner, and floppy drive to all match a G4 or an Imac, but in the PC world you don't even get perfectly matching shades of beige.
If properly designed, network servers don't have to be particularly slow. As long as you don't have too many people sharing a 10 MBit segment, bandwidth wont be a problem. Latency will be, but that is pretty much fixed. I would have gone with cheap 10/100 NICs (you can get them for $10), to at least allow the option of moving to 100BaseTx.
If the application set is relatively small, a network server with >=256 MB of RAM is going to have them all loaded all the time. So users aren't going to have to wait that long to start up *office or netscape.
Sharing files becomes really easy when everyone is on the same machine, or a small cluster of servers. Persumably the server cluster would be connected with 100BaseT as well.
A number of offices (Windows based, too) use WordPerfect not because they migrated, but because they never migrated to Office. This also dramatically reduces training, since to be honest, the majority of training issues are using applications, not the OS. If they start KDE and are presented with a button that says "Corel WordPerfect", even the most addle brained users are going to figure out what to do.
I wonder if you could set up an automated office translation server? If the filter APIs of office are exposed via COM or somesuch, someone should be able to whip up some perl and/or VB that would do the filtering on a Windows machine transparently Of course, you would have to pick an office suite whose native format MS Office had good import/export filters for.
First, he isn't trying to stick it to MS. He is trying to save his company big bucks by using nearly free equipment and software.
That said, this really does hurt MS. They make lots of money off of cilent side software. While the server side software (W2K Server, BackOffice, etc) are very expensive, client outnumber servers by a huge factor. And W2K client + Office Pro is not exactly cheap per seat.
Also most trends move from the client to the enterprise server. As long as MS has a large foothold in the client market, there will be a push to migrate servers to it, out of a desire to unify platforms, simplify management, and use untrained lackeys as system administrators. These are mostly myths. Running an NT/W2K server is nearly as different from running office on W9X as Linux is, and knowing how to play minesweeper does not qualify someone to manage an enterprise server, but many times people making these decisions don't understand this.
Linux gaining credibility as an extremely low cost desktop solution that can be used by accountants, lawyers, secretaries, and other mostly non-technical users is a big win for Linux on the client and the server.
Identical hardware is not so important when you are using the clients for X servers only, and running applications on a server. Also, if your computers cost $100 we really are talking about disposable computing. If it fails, throw it out. Maybe keep the hard drive and save yourself 20 minutes reinstalling on a new PC. You can do that 6 times before you touch the baseline of the cheapest machines you are likely to find are.
I don't know why you think you would need a different image for each machine. Unlike windows which has to reinstall drivers every time you move a card from one slot to another, a single kernel image can support just about any hardware you are likely to throw at it, at least for the purposes of an X terminal, where the only relevant devices are keyboard, mouse, NIC, HDD, and VGA. Just about every NIC every made is supported under linux, and most graphics cards you are likely to run into on obsolete machines. If it became a problem, you could probably find a large lot of obsolete VGA cards or NICs being sold for a few bucks per unit somewhere, and just drop those into every machine.
If you want sound, things are more complicated, but sound is really just a nicety for most business settings.
I personally think the latency when running apps over remote X is too high for comfort, but the management issues probably make that a worthwhile tradeoff for non-technical users.
The only thing I strongly disagree with him is the use of thinnet. Thinnet is fine for connecting a handful of machines in a single office, or even connecting computers in a large lab, but to deal with it on a larger / more spread out scale than that is idiotic. Unless you have a large base of installed cable, use 10BaseTx. You can probably pick up 24 port 10BaseT hubs pretty cheap these days, what with everyone migrating to 100MBit switches. Plus, you have a lot more flexibility to upgrade to 100BaseTx, or repartition your network to keep from getting bandwidth starved as you add more clients.
Well, when you buy or download a CD with a linux distribution, there is a lot of GNU software on it. Caldera's OpenUNIX product linked to yesterday demonstrates that for most users, the software environment is a much bigger piece of the "user experience" than the kernel.
However, there is a lot of other non-GNU software, too. It isn't really practical to say you use Mandrake X/KDE/GNU/Apache/Mozilla/Linux.
RMS, as a developer, thinks the development toolchain is the most important part of a system. Since almost all the development tools, plus the basic UNIX toolset (fileutils, shellutis, etc.) and emacs are all GNU software, GNU provides most of the software he uses directly.
To many other people, X, or their desktop environment, web browser, or irc client are the most important part of their system.
That is why I say Linux, but RMSs crusade for GNU/Linux doesn't bother me, nor do others who choose to say GNU/Linux.
However, what it sounds like he tried with glibc2 is inexcusable. I am going to withold final judgment until I see a statement from him, but I just lost a lot of respect for him.
RMS can be extremely obnoxious and hostile, but 99% of the time he is right. Here he is wrong.
Nope. Just as I can hire people to come over to my house and modify web pages before I view them
Actually, I am not sure you can do this, either. The fair use doctrine only applies to the "owner" of copyrighted material. In this case, that would be the person viewing the webpage. Nobody else can modify that page without the consent of the copyright owner.
This has been repeatedly validated in courts. The ruling against myMP3.com's internet jukebox said that the service was illegal because only the owner of a CD can make a copy. MP3.com could not make a copy, even if specifically requested by the user.
Similarly, it would be illegal for a digial cable company to implement TiVo-like PVR functionality into the cable subscription, unless they got permission from the copyright holders.
There are two issues remaining here. First, I won't say it is entirely clear that this interpretation of fair use is what it should be, but it is the way the legal system is now.
Second, it isn't clear who is making changes to a webpage when you view it with these plugins installed. If the software were entirely standalone, I would say that the end user was doing it: he had installed a tool that made modifications to web pages. This is analogous to me downloading cdparanoia and ripping a CD. It was me, not the authors of cdparanoia who made the copy. On the other hand, if I have software that uses information from an outside server to decide what changes to make, it is a little more ambigious. In the case of smartlinks (not to pick on MS, but it is an example most people are familiar with), it certainly sounds like MS is the one doing the altering.
Finally, I still say that if the changes made by said software misrepresent the original author, while it may not be copyright violation, it could certainly be fraud or libel (depending on what was changed, and to what end). If smartlinks, for instance, puts a link on the FSF homepage that says "Use Windows, we love it", and the web user viewing that believed it represented the FSF, that would definately be libelous--the FSF's credibility would have suffered measureably from a seeming endorement of Windows.
This is an extreme example, but much more subtle things are possible. Howabout something that detects "Now" buttons (Linux NOW!/Apache NOW!/Netscape NOW!...) and adds one that says W2K NOW!
If you would get upset at a newspaper misquoting you (or making something up entirely), you should be upset if people alter your webpage in a way that looks like you said something you didn't.
Keep in mind that the two ready sources of hydrogen are 1) electrolysis of water - no net energy gain, and 2) craking of hydrocarbons, which pollutes less than burning them, but has the same CO2 production.
I am a big fan of solar energy, but I don't think it will be useful for vehicles for some time yet, though you could use solar power to electolyze water for H2 vehicles.
In any case, I think solar energy is better suited to stationary or low power mobile devices, not transportation. I am a big fan of biomass energy for cars. Biomass methanol has a very high net energy value, a closed carbon cycle, and is safer than compressed hydrogen.
Absolutely false. I can deface your site all day long, as long as it's for my own personal use
Actually, I think it is indeed copyright violation (or some other illegal act), unless the software makes it completely clear what content was put there by the author, and what content came from a different source.
This is especially true when the "alteration" is an advertisment, or could be construed as endorsment of a product or service by the original author, or is in anyway for commercial gain on the part of the software author.
If MS wants to do something that provides links to additional content, they should do something like the netscape "What's related" or the mozilla sidebar (neither of which I use), or even put an item on the context menu for a link, rather than editing the page.
There is no real evidence to support the claim that users are made aware of what they are installing. Certainly if SmartTags are ever enabled by default, and installed on new computers, the user cannot be reasonable expected to know they have software editing the webpages they view. I can't imagine anyone who understood what Gator does actually wanting it, so I conclude that most users did not knowingly install it.
But the relevent point here is actually your tagline:
It's unethical to block ads. Don't like them? Don't visit sites that use them. Else, you are stealing.
Bottom line is, what Gator does is stealing. They are stealing the ad revenue from web authors.
I am not completely convinced that personal ad filters are stealing, though I don't use them because I think it is unethical. HTML makes no explicit guarantees on if or how something will be displayed. It is certainly not stealing to browse with Lynx, or disable automatic image loading, so I don't know that using junkbuster or another ad proxy is really theft. I could probably be convinced either way.
In any case, I stand by my claim that adding or altering content and representing it as the work of the site's author is a much more serious offence than removing content the user wishes not to see.
If you believe in fair use, then you'll delete that tag. If you don't, and you believe in fair use, then you are a hypocrite.
Not really. I don't think it is at all a violation of fair use to request that a web site not be automatically altered. It isn't like he encrypted his webpage and requires a signed executable to decrypt it.
I knew about the SmartTags meta tag, but I think there should be something more general. Just like robots.txt would be useless if every spider looked for a different file in a different format, this is only really useful if there is a more-or-less standard way to do it.
Best of all would be an opt-in system instead of an opt-out system, but I think it will be easier to get all parties to agree on an opt-out system.
I don't personally have any websites. I just am upset by this because I regularly visit a large number of ad-supported sites, many of which are in financial trouble right now. It really pisses me off when companies try to steal what little revenue these sites generate.
No, but if my Tivo replaced ads put in by the broadcasters, they would be sued, and it wouldn't really matter whether I had agreed to it or not.
I would say there is a difference between removing something (ie, adding a 30 second skip or allowing me to fast forward through ads) and replacing it with different content that is represented as the original.
Like I said, I wouldn't have a problem if the popped up ads were clearly seperate from the original content, such as in the task bar, or the toolbar of your browser, nor would I object to software that allowed the user to block out some or all ads. That is merely allowing the consumer to choose what parts of a webpage they view. Replacing content is fraudulent. End of story.
Since I can't think of any reason a customer who actually understood what Gator was doing would consent to having it installed on his/her computer, I have to assume that what they are doing is at least slightly deceptive.
If they had a disclamer which read in bold letters "Gator will attempt to drive out of business the free websites you most frequently visit by damaging their revenue stream from adversing, yet making you still look at other ads, do you want to do this?" just about everyone would say no.
There is no question here. This is Just Plain Wrong, and must be stopped. Same thing with smart links. It shouldn't even be an option.
Unfortunately, it seems like this is going to come up again and again. The best solution I can think of is a HTML meta tag or HTTP header like "HTTP-Dont-Fuck-With: yes". Adding or replacing content on such a page would be prohibited, and doing so would be considered fraudulent.
Now, I have no problem with something that doesn't affect the display of the page being viewed. If MS wants to add a button to the toolbar that serves the same function as smart links, or if Gator wants to add something to the system tray, or whatever, that is fine. But altering the content of a web page for comercial gain should be considered, as mentioned in the article, the same as cliping and replacing ads in a print magazine before you reveive it in the mail.
You don't need a GUI to admin the machine, but it is rather convenient on occasion, especially when doing the initial setup/configuration.
Many 3rd party softare packages (think, Oracle) don't have a console based installation. Sure, you can run X remotely, and I have done that plenty often, but it wouldn't be a bad idea to have Gnome there as a convenience. Presumably once the system is running "for real", you would turn off the X server to conserve resources.
Supposedly there is some ambiguity in exactly what VIA aquired with S3. I expect that VIA scrunitized the agreement extremely carefully, and has to be pretty confident that Intel would lose in court, but if there is any ambiguity, Intel could conceivable get a restraining order and slow them down a while.
At least one claim was that the S3 licensing agreement only applies do graphics chips and/or chipsets with integrated graphics.
Anyway, it will be interesting to see how this works out. My guess is that Intel is trying to stall this until their SDRAM chipset comes out, and doesn't expect/plan to kill it off entirely. Perhaps there will be some midnight licensing deal where Intel accepts much lower royalties than they had originally asked for (something like $20+/chipset) -- that would open up the market and speed P4 acceptence, while still saving face for Intel, and not set any dangerous (to Intel) precedents about enforcing patents.
We also have planetary formation theories, which while we don't have a lot of data to test them, can be useful.
The problem is that not only do you need a planet with a circular orbit in the habitation zone, you need a solar system free of other massive objects with eliptical orbits. If 50 % of the large objects in a solar system have highly elliptical orbits, it is going to be hard for life to form on one of the other 50%. If you have a large gas giant like Jupiter in a circular orbit, it is going to capture or eject most of those objects, leaving the habitible zone relatively clean.
the concept of molecular bifurcation (bifrucation?) communication is steadily gaining acceptance as something that might be viable in the not-so-distant future.
Only in the world of science fiction and dilber cartoons. The idea of "molecular bifurcation communication" in all of its forms is based on a misinterpretation of J.S. Bells' theorom (referenced in the Dilbert cartoon where he tries to make one, sorry I don't have the date).
Bells theorom losely states that it is possible to construct two particle systems -- two photon, two halves of a molecular decay, whatever -- in which the state of the system cannot be explained in terms of the state of part 1, and the state of part 2.. ie, there is some non-local correlation between them.
However, you can't actually *do* anything to one particle and have it affect the other. You can only see this correlation after you measure both particles, and compare their states -- which means you have to communicate over a "classical" channel first.
As for FTL travel/communication in general, I am not an expert in general relativity, and it is still actually an open question, but I don't believe that FTL communication will ever be possible.
SGI doesn't make any 32 CPU systems with uniform memory access, which was my point. I know how well IRIX scales on ccNUMA systems (Origin2K, Origin3K). Speculating on how they scale on a 16 or 32 CPU "normal" shared memory computer can be nothing but speculation.
Slashdot Mirror
You have to consider the possibility that there are some people whose way of life is fundamentally incompatable with yours. I believe that anyone capable of organizing the hijacking of several commerical airplanes and using them as weapons against both civillian and military targets in an unprovoked sneak attack is not someone that I can make amends with without making an unacceptable compromise of my integrity.
If possible, pepetrators of this kind of terrorism should be made to see the error of their ways, in the hopes of preventing future incidents. However, if that proves to be impossible, the only recourse is total annihalation.
I am a peaceful person. I would do just about anything to insure that we could all live harmoniously. But if other people force the issue into a "me or them" situation, I pick me, every time, and will defend that decision with ultimate force.
The fabricators are pretty expensive, and the amortized cost of the machine + materials is probably much greater than the replacement parts.
However, consider that an assembly line has thousands of different parts like this. The cost of keeping spares of every part around is much higher than the cost of just the failed parts. If you can just fabricate the part you need on the spot, you reduce downtime vs. waiting for a replacement part. On a high volume assembly line, my guess is the cost of a few of hours less downtime can pay for the prototyping machine.
School is what you make it. Yes, there are good profs and bad profs, and good course and bad courses, but at almost any university, you can make the experience worthwhile and educational, or you can make it a worthless waste of money and time.
I've taught myself way more than most professors.
Yes, that is a good sign. I would go so far as to say that it is impossible for anyone but yourself to teach you anything. The secret to getting a lot out of an education is to think of it in terms of learning, rather than teaching. Every class you take gives you an opportunity to learn something. You probably could learn it yourself, with appropriate books, and enough time, but nowhere near as fast as if you have a professor and a class full of other students to help you.
So, you have to approach your schoolwork with the attitude that you are the boss of you. Nobody is going to make you do things the way the professor tells you too, and if you are willing to accept the consequences, don't do things you don't think are beneficial.
A lot of people, some professors included, probably don't agree with this aproach to learning. They may, in fact, believe that professors are sacred fonts of wisdom that bless students with their presence. But it doesn't matter -- you can learn almost as much from those professors as anyone else, more if you count learning to suffer fools gracefully.
Finally, I have gone to a lot of schools and had a lot of professors/teachers. Some were brilliant, some were idiots, some were wonderful people, some were complete assholes, some I loved, some I hated. But I have never -- repeat NEVER -- met a teacher who wasn't willing to help a student learn, if that student sincerly asked for it.
If you get any less than a stellar education, you have primarily yourself to blame.
Actually, both this solution and LinuxBIOS can get to a command prompt (or, more likely, load an embedded application) in 3 seconds.
The 0.8s to LILO prompt is nice because it demonstrates how fast the BIOS is, but is irrelevant from a system designer standpoint: The important thing is how fast they can get their program running.
In 2.4.X, linux defaults to using the same outgoing port that the originating machine used, then tries sequentially from there, rather than using ports in the 60,000 range (as prior kernel series did).
You could still probably detect it, but A) it would be inconclusive, and B) it wouldn't mean anything if it were. I have two computers behind a NAT based firewall, and lots of people use those linksys (or other) "Cable/DSL routers" that do NAT automatically.
Well, first of all, when your lease times out, you will get switched automatically. Of course, any active connections you have will be dropped, but your connectivity will continue. This will hork up SSH connections, but HTTP should be fine.
Assuming you can get information from the lan card when it switches base stations, you should be able to automate the release/renew cycle (on *ix or Win*). Even if it isn't a clean message, you should be able to monitor the MAC of the active base station on a short enough interval you can switch almost seamlessly.
The guys at O'Reilly working on this have an idea that uses a hash of your MAC as your IP address, so you can keep the same address when you move from cell to cell.
Finally, I believe that if you put point-to-point links between adjacent cells, and substantially increse the sophistication of the gateway software, you should be able to handle forwarding established connections when someone moves from cell to cell. Plus, then you could do bandwidth agregation, if you were really clever, and had a relatively dense network.
Of course, that involves a substantial amount of extra hardware, and correspondingly higher costs.
Hm. Are there any reasonably low cost capture systems that have high quality S-Video replay? I was thinking about doing this sort of thing for archival purposes. I have a TiVo, which I love, but it is a 2 drive 30 hour model, and I don't want to screw it up attempting and upgrade.
I already have scads of unused drive space, I just need the capture card, but I want one that has S-Video in and out at high quality. I couldn't care less about VGA.
IIRC, the TiVo uses an MPEG chipset designed to go with the PowerPC CPU they use, and it isn't available as a consumer board.
Also, how easy/hard is it to cut a mpeg video, say, to remove commercials? It seems to me that roughly every 5th frame is uncompressed, so you ought to be able to slice it on those boundaries.
First of all, there is a limited market for these things. Very few people are willing to pay the extra it would cost to get a tricked out case.
Second, they would be really expensive. If you make these as a hobby, part of the goal is to have fun building it, and in the end you have a unique case that you made yourself. If you make these to sell, you have to charge people for your time, and that bumps the price way up.
In the PC market, you just can't make money off of this sort of thing. The only place it works is when you have a single supplier of something. Hence, the distinctive looks of Apples, SGIs, NeXT, and Cray computers (to name a few). You buy the hardware you want and automatically get a nice looking case to go with it (depending on your tastes).
The other factor is availability of peripherals. It is difficult for an end user to even find black CD-ROMs, much less ones to go with a specific case design, or other peripherals that match their color scheme. Again, the only way this works is with a single supplier, or if you are going to do the mods yourself. Witness how you can get a keyboard, mouse, USB hub, scanner, and floppy drive to all match a G4 or an Imac, but in the PC world you don't even get perfectly matching shades of beige.
If properly designed, network servers don't have to be particularly slow. As long as you don't have too many people sharing a 10 MBit segment, bandwidth wont be a problem. Latency will be, but that is pretty much fixed. I would have gone with cheap 10/100 NICs (you can get them for $10), to at least allow the option of moving to 100BaseTx.
If the application set is relatively small, a network server with >=256 MB of RAM is going to have them all loaded all the time. So users aren't going to have to wait that long to start up *office or netscape.
Sharing files becomes really easy when everyone is on the same machine, or a small cluster of servers. Persumably the server cluster would be connected with 100BaseT as well.
A number of offices (Windows based, too) use WordPerfect not because they migrated, but because they never migrated to Office. This also dramatically reduces training, since to be honest, the majority of training issues are using applications, not the OS. If they start KDE and are presented with a button that says "Corel WordPerfect", even the most addle brained users are going to figure out what to do.
I wonder if you could set up an automated office translation server? If the filter APIs of office are exposed via COM or somesuch, someone should be able to whip up some perl and/or VB that would do the filtering on a Windows machine transparently Of course, you would have to pick an office suite whose native format MS Office had good import/export filters for.
First, he isn't trying to stick it to MS. He is trying to save his company big bucks by using nearly free equipment and software.
That said, this really does hurt MS. They make lots of money off of cilent side software. While the server side software (W2K Server, BackOffice, etc) are very expensive, client outnumber servers by a huge factor. And W2K client + Office Pro is not exactly cheap per seat.
Also most trends move from the client to the enterprise server. As long as MS has a large foothold in the client market, there will be a push to migrate servers to it, out of a desire to unify platforms, simplify management, and use untrained lackeys as system administrators. These are mostly myths. Running an NT/W2K server is nearly as different from running office on W9X as Linux is, and knowing how to play minesweeper does not qualify someone to manage an enterprise server, but many times people making these decisions don't understand this.
Linux gaining credibility as an extremely low cost desktop solution that can be used by accountants, lawyers, secretaries, and other mostly non-technical users is a big win for Linux on the client and the server.
Identical hardware is not so important when you are using the clients for X servers only, and running applications on a server. Also, if your computers cost $100 we really are talking about disposable computing. If it fails, throw it out. Maybe keep the hard drive and save yourself 20 minutes reinstalling on a new PC. You can do that 6 times before you touch the baseline of the cheapest machines you are likely to find are.
I don't know why you think you would need a different image for each machine. Unlike windows which has to reinstall drivers every time you move a card from one slot to another, a single kernel image can support just about any hardware you are likely to throw at it, at least for the purposes of an X terminal, where the only relevant devices are keyboard, mouse, NIC, HDD, and VGA. Just about every NIC every made is supported under linux, and most graphics cards you are likely to run into on obsolete machines. If it became a problem, you could probably find a large lot of obsolete VGA cards or NICs being sold for a few bucks per unit somewhere, and just drop those into every machine.
If you want sound, things are more complicated, but sound is really just a nicety for most business settings.
I personally think the latency when running apps over remote X is too high for comfort, but the management issues probably make that a worthwhile tradeoff for non-technical users.
The only thing I strongly disagree with him is the use of thinnet. Thinnet is fine for connecting a handful of machines in a single office, or even connecting computers in a large lab, but to deal with it on a larger / more spread out scale than that is idiotic. Unless you have a large base of installed cable, use 10BaseTx. You can probably pick up 24 port 10BaseT hubs pretty cheap these days, what with everyone migrating to 100MBit switches. Plus, you have a lot more flexibility to upgrade to 100BaseTx, or repartition your network to keep from getting bandwidth starved as you add more clients.
Well, when you buy or download a CD with a linux distribution, there is a lot of GNU software on it. Caldera's OpenUNIX product linked to yesterday demonstrates that for most users, the software environment is a much bigger piece of the "user experience" than the kernel.
However, there is a lot of other non-GNU software, too. It isn't really practical to say you use Mandrake X/KDE/GNU/Apache/Mozilla/Linux.
RMS, as a developer, thinks the development toolchain is the most important part of a system. Since almost all the development tools, plus the basic UNIX toolset (fileutils, shellutis, etc.) and emacs are all GNU software, GNU provides most of the software he uses directly.
To many other people, X, or their desktop environment, web browser, or irc client are the most important part of their system.
That is why I say Linux, but RMSs crusade for GNU/Linux doesn't bother me, nor do others who choose to say GNU/Linux.
However, what it sounds like he tried with glibc2 is inexcusable. I am going to withold final judgment until I see a statement from him, but I just lost a lot of respect for him.
RMS can be extremely obnoxious and hostile, but 99% of the time he is right. Here he is wrong.
Yep. You can tell O1 owners because their right arm is slightly longer than their left.
I still have mine around, somewhere.
Nope. Just as I can hire people to come over to my house and modify web pages before I view them
Actually, I am not sure you can do this, either. The fair use doctrine only applies to the "owner" of copyrighted material. In this case, that would be the person viewing the webpage. Nobody else can modify that page without the consent of the copyright owner.
This has been repeatedly validated in courts. The ruling against myMP3.com's internet jukebox said that the service was illegal because only the owner of a CD can make a copy. MP3.com could not make a copy, even if specifically requested by the user.
Similarly, it would be illegal for a digial cable company to implement TiVo-like PVR functionality into the cable subscription, unless they got permission from the copyright holders.
There are two issues remaining here. First, I won't say it is entirely clear that this interpretation of fair use is what it should be, but it is the way the legal system is now.
Second, it isn't clear who is making changes to a webpage when you view it with these plugins installed. If the software were entirely standalone, I would say that the end user was doing it: he had installed a tool that made modifications to web pages. This is analogous to me downloading cdparanoia and ripping a CD. It was me, not the authors of cdparanoia who made the copy. On the other hand, if I have software that uses information from an outside server to decide what changes to make, it is a little more ambigious. In the case of smartlinks (not to pick on MS, but it is an example most people are familiar with), it certainly sounds like MS is the one doing the altering.
Finally, I still say that if the changes made by said software misrepresent the original author, while it may not be copyright violation, it could certainly be fraud or libel (depending on what was changed, and to what end). If smartlinks, for instance, puts a link on the FSF homepage that says "Use Windows, we love it", and the web user viewing that believed it represented the FSF, that would definately be libelous--the FSF's credibility would have suffered measureably from a seeming endorement of Windows.
This is an extreme example, but much more subtle things are possible. Howabout something that detects "Now" buttons (Linux NOW!/Apache NOW!/Netscape NOW!...) and adds one that says W2K NOW!
If you would get upset at a newspaper misquoting you (or making something up entirely), you should be upset if people alter your webpage in a way that looks like you said something you didn't.
Keep in mind that the two ready sources of hydrogen are 1) electrolysis of water - no net energy gain, and 2) craking of hydrocarbons, which pollutes less than burning them, but has the same CO2 production.
I am a big fan of solar energy, but I don't think it will be useful for vehicles for some time yet, though you could use solar power to electolyze water for H2 vehicles.
In any case, I think solar energy is better suited to stationary or low power mobile devices, not transportation. I am a big fan of biomass energy for cars. Biomass methanol has a very high net energy value, a closed carbon cycle, and is safer than compressed hydrogen.
Absolutely false. I can deface your site all day long, as long as it's for my own personal use
Actually, I think it is indeed copyright violation (or some other illegal act), unless the software makes it completely clear what content was put there by the author, and what content came from a different source.
This is especially true when the "alteration" is an advertisment, or could be construed as endorsment of a product or service by the original author, or is in anyway for commercial gain on the part of the software author.
If MS wants to do something that provides links to additional content, they should do something like the netscape "What's related" or the mozilla sidebar (neither of which I use), or even put an item on the context menu for a link, rather than editing the page.
There is no real evidence to support the claim that users are made aware of what they are installing. Certainly if SmartTags are ever enabled by default, and installed on new computers, the user cannot be reasonable expected to know they have software editing the webpages they view. I can't imagine anyone who understood what Gator does actually wanting it, so I conclude that most users did not knowingly install it.
But the relevent point here is actually your tagline:
It's unethical to block ads. Don't like them? Don't visit sites that use them. Else, you are stealing.
Bottom line is, what Gator does is stealing. They are stealing the ad revenue from web authors.
I am not completely convinced that personal ad filters are stealing, though I don't use them because I think it is unethical. HTML makes no explicit guarantees on if or how something will be displayed. It is certainly not stealing to browse with Lynx, or disable automatic image loading, so I don't know that using junkbuster or another ad proxy is really theft. I could probably be convinced either way.
In any case, I stand by my claim that adding or altering content and representing it as the work of the site's author is a much more serious offence than removing content the user wishes not to see.
If you believe in fair use, then you'll delete that tag. If you don't, and you believe in fair use, then you are a hypocrite.
Not really. I don't think it is at all a violation of fair use to request that a web site not be automatically altered. It isn't like he encrypted his webpage and requires a signed executable to decrypt it.
I knew about the SmartTags meta tag, but I think there should be something more general. Just like robots.txt would be useless if every spider looked for a different file in a different format, this is only really useful if there is a more-or-less standard way to do it.
Best of all would be an opt-in system instead of an opt-out system, but I think it will be easier to get all parties to agree on an opt-out system.
I don't personally have any websites. I just am upset by this because I regularly visit a large number of ad-supported sites, many of which are in financial trouble right now. It really pisses me off when companies try to steal what little revenue these sites generate.
No, but if my Tivo replaced ads put in by the broadcasters, they would be sued, and it wouldn't really matter whether I had agreed to it or not.
I would say there is a difference between removing something (ie, adding a 30 second skip or allowing me to fast forward through ads) and replacing it with different content that is represented as the original.
Like I said, I wouldn't have a problem if the popped up ads were clearly seperate from the original content, such as in the task bar, or the toolbar of your browser, nor would I object to software that allowed the user to block out some or all ads. That is merely allowing the consumer to choose what parts of a webpage they view. Replacing content is fraudulent. End of story.
Since I can't think of any reason a customer who actually understood what Gator was doing would consent to having it installed on his/her computer, I have to assume that what they are doing is at least slightly deceptive.
If they had a disclamer which read in bold letters "Gator will attempt to drive out of business the free websites you most frequently visit by damaging their revenue stream from adversing, yet making you still look at other ads, do you want to do this?" just about everyone would say no.
There is no question here. This is Just Plain Wrong, and must be stopped. Same thing with smart links. It shouldn't even be an option.
Unfortunately, it seems like this is going to come up again and again. The best solution I can think of is a HTML meta tag or HTTP header like "HTTP-Dont-Fuck-With: yes". Adding or replacing content on such a page would be prohibited, and doing so would be considered fraudulent.
Now, I have no problem with something that doesn't affect the display of the page being viewed. If MS wants to add a button to the toolbar that serves the same function as smart links, or if Gator wants to add something to the system tray, or whatever, that is fine. But altering the content of a web page for comercial gain should be considered, as mentioned in the article, the same as cliping and replacing ads in a print magazine before you reveive it in the mail.
You don't need a GUI to admin the machine, but it is rather convenient on occasion, especially when doing the initial setup/configuration.
Many 3rd party softare packages (think, Oracle) don't have a console based installation. Sure, you can run X remotely, and I have done that plenty often, but it wouldn't be a bad idea to have Gnome there as a convenience. Presumably once the system is running "for real", you would turn off the X server to conserve resources.
Supposedly there is some ambiguity in exactly what VIA aquired with S3. I expect that VIA scrunitized the agreement extremely carefully, and has to be pretty confident that Intel would lose in court, but if there is any ambiguity, Intel could conceivable get a restraining order and slow them down a while.
At least one claim was that the S3 licensing agreement only applies do graphics chips and/or chipsets with integrated graphics.
Anyway, it will be interesting to see how this works out. My guess is that Intel is trying to stall this until their SDRAM chipset comes out, and doesn't expect/plan to kill it off entirely. Perhaps there will be some midnight licensing deal where Intel accepts much lower royalties than they had originally asked for (something like $20+/chipset) -- that would open up the market and speed P4 acceptence, while still saving face for Intel, and not set any dangerous (to Intel) precedents about enforcing patents.
I believe such a satelite would probably be tide locked, so that one side always faced the gas giant, and the other either faced the sun, or away.
It seems to me that a thicker-than-earth greenhouse gas layer would help out the problem a lot.
Of course, you have to be far enough away from the planet to not get your atmospher sucked away.
We also have planetary formation theories, which while we don't have a lot of data to test them, can be useful.
The problem is that not only do you need a planet with a circular orbit in the habitation zone, you need a solar system free of other massive objects with eliptical orbits. If 50 % of the large objects in a solar system have highly elliptical orbits, it is going to be hard for life to form on one of the other 50%. If you have a large gas giant like Jupiter in a circular orbit, it is going to capture or eject most of those objects, leaving the habitible zone relatively clean.
Only in the world of science fiction and dilber cartoons. The idea of "molecular bifurcation communication" in all of its forms is based on a misinterpretation of J.S. Bells' theorom (referenced in the Dilbert cartoon where he tries to make one, sorry I don't have the date).
Bells theorom losely states that it is possible to construct two particle systems -- two photon, two halves of a molecular decay, whatever -- in which the state of the system cannot be explained in terms of the state of part 1, and the state of part 2.. ie, there is some non-local correlation between them.
However, you can't actually *do* anything to one particle and have it affect the other. You can only see this correlation after you measure both particles, and compare their states -- which means you have to communicate over a "classical" channel first.
As for FTL travel/communication in general, I am not an expert in general relativity, and it is still actually an open question, but I don't believe that FTL communication will ever be possible.
SGI doesn't make any 32 CPU systems with uniform memory access, which was my point. I know how well IRIX scales on ccNUMA systems (Origin2K, Origin3K). Speculating on how they scale on a 16 or 32 CPU "normal" shared memory computer can be nothing but speculation.