Napster is an enabling technology that allows its users to break the law. It is generally legal in the U.S. to sell things to people even when the purpose of the things is to break the law. Selling radar detectors is legal in most states, even though the only possible purpose of a radar detector is to help its owner break the law. Are radar detector manufacturers liable for contributing to millions of cases of speeding? If not, why should Napster be liable for millions of cases of copyright infringement?
Actually, as I recall (someone check me on this) there was a v6 ODBC driver on the last beta that was sent out before the open source decision. The problem is, Interbase Corp. doesn't own the rights to it and therefore can't open-source it, at least not without permission. It does bring up a good point: What will the licensing terms be for the 6.0 ODBC driver? I think I'll go ask on their newsgroup...
They could refund money to the distributors for any returned copies they haven't sold yet.
That could be a lot of copies. Top management probably has some reservations about this whole weird "open source" thing. Throw in a couple hundred thousand on the wrong side of the ledger and you might wind up with obsolete binaries rather than released source.
Or they could simply say that the "end of life" for the product is (some deadline) and promise to release the source once that deadline is past.
Certainly. Of course, depending on the situation, that might well result in the eventual source release being even further delayed.
If nothing else, couldn't the first open-source version be the minimum cleanup necessary from the last proprietary version, not a development version with lots of half-finished features?
It depends on the situation. If the new version is a total rewrite and doesn't work yet, well, sure, the old version might be interesting. But most of the time, I'd want the latest code. Specifically in the case of Interbase, 6.0 is in late beta, feature-complete, and close to releasable; in fact, they might wind up putting out the source and the 'official' 6.0 release on the same day.
The scary thing about this is that if the same situation were to arise today, the story would be very different. With DMCA, UCITA, etc., IBM would almost certainly have been able not only to win licensing control from CMU, but also to prevent MIT from suceeding with the X project. Certainly, good software needs to evolve. Clearly, though, the only reason it's been able to do so is that the greed-inspired have had no way to stop it; Bill Gates famous phobia regarding "some guy in a garage" was actually well warranted. Until now.
I don't think this stems from laziness on the part of the company doing the open-sourcing. The problem is, there are existing contractual agreements with distributors, VARs, etc., regarding the current product release. It is these people, not the software company itself, who are often the main force behind keeping a product proprietary. The software company might be able to find other revenue sources, but the distributors are just flatly out of luck. Keep in mind that up to the moment of making the open-source decision, these people have been the software company's major customers, and the full weight of customer service culture pushes towards doing what they want--it takes a clear and powerful vision to make a decision that you know is going to totally alienate all the customers that built your business to the point it is today. And quite frankly, it's absolutely the wrong thing to do, 99% of the time.
So what it boils down to is, if you open-source the existing version, you cause all the existing inventory at the distributors to have a value of zero; this constitutes damages, and is probably actionable. Expect your pants to be sued off shortly. If you open-source the next, as yet undistributed version, all the distributors angrily drop your product (doesn't matter how much mark-up you give them if the MSRP is $0.00), but they don't have a case against you for any real damages, since you haven't done anything to reduce the value of their existing inventory, assuming you continue to claim, without snickering too loud, that the old version will still be sold at full price.
And whatever happens, you'd better do a flawless execution of your open source business plan, because the bridge you just rode in on is now merrily in flames.
Not to start a flame war, but you're much more likely to succeed in hooking MS-Project to Interbase. All the stuff on this thread that's being claimed for PostgreSQL some time in the future, Interbase already has. My company has run on Interbase for years. The Philadelphia Stock Exchange uses an Interbase back-end. Now it's (shortly going to be) open source. What more could you ask for? Yes, PostgreSQL is a worthy database in many ways, certainly more so than MySQL. It's a serious contender for Web-served applications, for example--maybe even as good as Interbase, as long as you don't want to do more-than-trivial stored procedures, sub-select queries, domain-based schemas, triggers, event alerters, long-duration isolated transactioning (i.e. snapshots), or any of the other things Interase does well but PostgreSQL does not. But have you ever tried to connect PostgreSQL to a Windows client? The ODBC drivers, quite frankly, suck. Before Interbase was free, I actually deployed PosgreSQL/ODBC to an office full of Windows clients. I had unsurmountable problems with stability, compatibility issues, etc, etc. The database itself was fine, but the Windows client support was flatly unacceptable. This was about a year ago, maybe it's improved since then, but the simple problem is: Why would anyone on the PostgreSQL team be motivated to write for Windows?
So if you want a free Linux database serving Windows clients, there simply is nothing to compare to Interbase. You can download the binaries free today, and the source will be out as soon as they've built a makefile that works on anyone else's box, removed all the non-politically-correct comments, got final sign-off from the lawyers, etc, etc. The Web site is www.interbase.com.
Strangely enough, many other countries seem to have developed fairly reasonable compromises on this issue. I don't recall this being front-page news in my time in Canada or the UK. The Germans, Japanese and French have all pretty much figured out systems they're happy with. Why is it so hard for the U.S. to accept that in some cases, collective freedom (ie, things like physical safety, medical care and clean air) can be more important than individual freedom (ie, ownership of handguns, low taxes, unrestricted consumption)? That this balance must be struck in all societies? That restriction of some individual freedoms is necessary, not because Big Brother wants to make you unhappy, but because we the people desire not to have our children shot by that handgun you so desparately insist on keeping and bearing?
The interesting questin here is what happens when Congress passes two obviously contradictory pieces of legislation. The DMCA has specific exemptions that permit copying to and from computer hard drives. What happens when two laws of equal priority (ie, neither is a constitutional amendment or whatever) are in obvious conflict? If you sue me under AHRA, can I defend myself using the provisions from DMCA?
Not that I'm saying DMCA and AHRA are necessarily contradictory; I don't know a thing about what AHRA says.
I did stop getting the paper, but the reasons people have brought up so far really aren't why. I think most of the Slashdot readership is coming at this from the perspective of: I read news on the Web now, so why don't I switch to a paper? In my case, I used to read the paper, but I switched to the Web. Perhaps, as an old man of 31, I'm not quite jogging to the same Walkman as most of my Slashdot compatriots.
Several comments have referred to the timeliness of the Web as compared to newspapers. It's certainly easy to get addicted to news-as-it-happens, but unless you're a political campaign manager, FEMA director, Fortune 500 CEO or (real) day trader, newspapers are already more than fast enough to get you the information you actually need. So this wasn't a major concern for me. I would be just as happy waiting until after Super Tuesday before awareness even dawns that the primaries have started. Certainly, hearing about the results the next day is not a problem for me.
The other major issue seems to be quality of content. Again, I don't find this to be particularly relevent. After all, most of the newspaper articles are the exact same AP or Reuters wire feeds that you can read on Yahoo. It's true that the Web has some really good specialized news sites, but on the other hand, good newspapers have excellent investigative reporters. In the end, it comes out about even in my book.
Katz and a few other posters seem to be unduly concerned about the lack of graphic content in newspapers. I would like to point out that the Web isn't any more graphical than a newspaper. How many of Katz's articles on Slashdot have ever had so much as a single accompanying photo? Does anyone else notice that the photos accompanying Yahoo news items seem jarringly out of place somehow? The Web is clearly a textual medium, just like a newspaper--the opinions of Web pundits with fond memories of the days when Wired was actually cool notwithstanding.After all, do you watch Slashdot or do you read Slashdot? The Web is therefore a much better place than, say, television, for the presentation of complex and difficult topics. But so are newspapers. Print journalism is in a very sad state at the moment, but this is not inherent to the medium; there used to be excellent reporting in the papers, and if the Web figures out how to do it well, the papers can for the most part pick it up without any trouble. Again, it's a wash.
If these weren't the reasons, why did I stop my subscription to the Raleigh News & Observer several years ago? The main reasons are:
The physical act itself
I've never found reading a newspaper to be a very comfortable experience. It's too big. If you're in a crowded area, like a typical RTP lunch spot, you practically have to take up two seats. Newspapers were obviously designed to be read in overstuffed armchairs. I don't own an overstuffed armchair. I'm perfectly comfortable reading a magazine or novel over lunch, but a newspaper is too cumbersome. Actually, the only way I really enjoy reading a newspaper is if I take up the whole dining room table with it. Big, slow and spread out all over the place just doesn't fit my personality or lifestyle.
The lack of searchability
I don't read the paper for its entertainment value. My recreational reading runs to novels, the Atlantic Monthly and Scientific American. I (used to) read the paper to keep up with current events in areas which interest me. However, newspapers only provide one index method (so to speak) - the date. I frequently find myself wanting to know what's happening in some specialized area, like: what's happening today in Kosovo. I'm not a passive vessel to fill with whatever the newspaper chooses to publish that day. I'm an active, self-directed seeker of information. I want to be able to ask a question and get an answer. I can do this with the Web, but it's hard to see how a newspaper could work this way.
Local news
I'm going to go against the grain here and say that I much prefer getting my local news from the WRAL-TV Web site, rather than the News & Observer, so I credit this as a strenght of the Web. I understand this might not be true in other areas, because they might not have a resource like WRAL. However, I think WRAL has made it clear that the Web has inherent possibilities for local news reporting that newspapers can't duplicate. With this year's hurricanes and snowstorms, I want to know, in the evening, whether the kids' schools will be open the following morning. As I get ready to leave for work, I want to know, at that particular moment, what the traffic situation is like on I-40. I want to know what that big construction project I saw driving to work was all about. I want to see the milestone dates for the Outer Loop--and whether they're being met or not. And I want it all to appear at the moment I want to know, not when the paper happens to run a story.
Heavy-handed marketing
The News & Observer went through a couple years where they made extensive use of outbound telemarketing as a promotional vehicle. I'm not one of these knee-jerk boycotters, but when a firm interrupts your dinner half three times in the same month to pummel you with a hard-sell for a product you already buy from them, you tend to notice. It's not enough of a big deal to make me stop the paper for just this reason, but the ongoing sense of vague irritation was certainly a contributing factor.
Editorial bias
Everyone in this area says the N&O is too liberal. Personally, I find it far too conservative. But that's not the problem. The problem is that by having a single reporting structure under an editor-in-chief, every newspaper can't help but show some kind of editorial bias. When I want to read opinion pieces, I'll pick up a magazine. I'm interested in a newspaper for factual reporting of what actually happened. Much too often, the interpretation and opinion gets in the way of telling the story as it actually occurred. Reading newspaper pieces, I often find myself disagreeing with the writer and wishing I had some source of simple factual data to check up on it for myself - so I go to the Web. Once this had happened often enough, I started going to the Web first, rather than reading the paper at all. When I found that a whole week of papers had come in and I hadn't read a single one, I decided to quit.
Incidentally, now I know why Katz and Jesse Burst and so forth close their articles with "opinions? talk back by clicking..." type statements. It's much easier than thinking of an effective closing paragraph.
It's not a GPL violation if they provide source when they send it out on campus. A lot of Slashdot readers would be well advised to actually read the GPL. Click Here Now. You will not find anywhere that it says you have a personal obligation to give copies to anyone. It's just that if you do distribute a copy, you must also give out source, or provide a way to get it.
Think about it. How much free software would be written if all free software authors were required to maintain multiple T1s to handle the load, if they happened to write something popular?
Now, if they try to say that students aren't allowed to redistribute CAEN Linux from a student, or whatever, they're violating the GPL. But they haven't done that. They've just put it on an on-campus-only download site. If you offered to mirror it off-campus, no doubt you could get a copy.
However - if it's anything like Eos Linux from NCSU, it'll be full of site-specific code that's meaningless outside the UMICH network. So why would you want it?
North Carolina State University has had a group of students supporting a Red Hat-based distribution called EOS Linux for some time. Just recently, the university announced that they would begin treating it as an officialy supported platform. Details are sketchy - the announcement just came out - but it's a pretty big deal for the NCSU folks. Perhaps some Eos Linux people can provide more depth.
Actually, recent versions of EOS Linux aren't distributions, but a set of rpms that you install against a specific version of Red Hat. The Eos Linux group felt it would be easier to it this way, rather than taking on all the day-to-day maintenance responsibilities of supporting a full distribution. I think this is a better model for site-specific Linuxes. It's a good discussion topic, anyway.
Last but not least - I don't know if it would be possible to build a global "University Linux" as the article suggests. The main benefit to Eos Linux, and presumably to the University of Michigan Linux as well, is that it integrates tightly with all the on-campus systems and networks. It's site-specific, and that's why you want it. It's hard to imagine how you could build a single distribution that would be site-specific at all sites.
Hey, hold on a minute, there. Remember how Netscape performed before they open sourced it? On my machine, it crashed multiple times a day. It felt like a race to see if you could get the answer to your question (or whatever) before the browser died. We all knew, well before the source was ever made public, that it was a pig.
Interbase is not like that. Interbase is one of the best-running, bug-free, and robust pieces of software I've ever worked with, at least from a "black box" perspective. I'd love to see the source, and I'd give you long odds indeed against it turning out to be a "pile of crud."
I'll work on it. I'm a database developer and administrator at a 400-employee firm, and our database is Interbase. Technically, it's unbeatable. We are not looking at another Netscape fiasco here (unless Inprise screws it up somehow). Navigator was practically unworkable. Before they even released the source, everyone knew it would stink, just because of how poorly Navigator performed. Interbase is not like this. The _only_ thing wrong with Interbase is a lack of awareness/marketing/mindshare. And I can tell you that it scales flawlessly to at least tens of gigabytes and four-way SMP (though not on Linux). I can't comment on nutty mainframe-like setups, but if I had the money and time, I'd love to see a two- or four-CPU server, with a gig or so dataset, running performance and stability comparisons between Interbase and anything else in the world. Open sourcing Interbase would give me a bigger woody than anything else that's happened this year or is likely to next year.
Requiring an unique IP would exclude all but one user behind each NAT-based corporate firewall. I have 300 analytical chemists here, many of whom play chess, and quite possibly more than one of whom might participate in something like this. Do you think they would be happy to learn that only one person in our company would be allowed to submit votes?
I know that people will disagree with this, but the X server is like the kernel in that it is supposed to be able to deal with *anything* that a misbehaving user application could throw at it. If you're non-root, there should be no possible code you could ever execute that crashes the kernel *or* the X server.
So the best you can say is that if there's an app (cough...Netscape...cough) that crashes X all the time, then *both* the app and the X server have a bug.
The difference, of course, is that McDonald's Corporation is not trying to follow a free software ideology. Red Hat has stated publicly, on numerous occasions, that anyone can copy the distro and sell it for money. So the question now is, if people choose to do so, what rules do we want them to follow? It would be much better if these were stated explicitly up front in a defensible way. Given the choice, who do you want to set the standard for interpreting copyright and trademark law in the context of free software: a member of the community like Red Hat, or the court system? As you consider your answer, consider any three recent computer-related precedents set by the courts...
How much effort do you have to go to in order to be "calling it" something different? Is it sufficient to just put a different name on the outside of the CD, or do you also have to update/etc/issue, redhat-release, etc? What are the rules about what the name can be? Could you, for example, make an exact copy of the Official Red Hat Linux 6.0 CD, label it "Unofficial Red Hat Linux 6.0" and then sell it? Or would you have to do more than that?
Also, does it have to actually be Red Hat Linux in order to call it Red Hat Linux? For example, could you make a copy of Debian 2.1 and sell it under the product name "This Is Not Red Hat Linux?" Or could you make copies of FreeBSD and market it as as "Red Hat FreeBSD" - not because it has anything to do with Red Hat, but just to gain some of the benefit of Red Hat's brand identity? If these extreme cases are unacceptable, where exactly is the line drawn and how do you know if you've crossed it?
I've never used OpenLinux, but I work in a Novell shop, and I was quite excited by the NetWare port that Caldera released for Linux. I was even prepared to spend good money on it. Unfortunately, they released it as binary only, it required their customized 2.0.35 kernel with STREAMS support, and they never did one lick of work to support it. Fortunately, I noticed this before I put down any actual cash.
But at Linux Expo in Raleigh, when I made my concerns known to their sales reps, they told me to wait a couple months for some really exciting announcements that would change my mind about Caldera.
Can you name another client that lets you store something other than mail on an IMAP store?
We run Novell Groupwise, and it stores everything, including calendar information, in the same message store, all of which is accessible via IMAP. Schedule items are nothing but e-mail messages with a second set of RFC822-like header lines at the beginning of the body text, so if you read your mail via a non-Groupwise IMAP client, you just see "Time:" "Location:" etc. headers at the top of your message.
There's no reason in the world why this would break anything, and it's open and accessible to anyone who wants to mess with it (eg: I've written Perl scripts to check my own calendar and page me via qpage if I have a meeting coming up that meets certain criteria for being important).
If I were a wide-eyed innocent, I would say something like "Gee, I don't understand why everyone doesn't do this."
Me too, actually. And while I was never an OS/2 fan, a lot of OS/2 people are quite unhappy about the way Borland dropped that platform. It's true that Borland could burn a lot of people by shipping a Linux product and then dropping it later. But the interesting thing is, they will then have done a lot of work teaching us (the community) why a RAD tool is useful, how it should work, etc, etc. If they release and then drop a Delphi-for-Linux, they'll have done a great deal towards the free Delphi-clone-for-Linux project that will result. Nothing like a huge number of frustrated developers with an itch to scratch...
Borland's compilers have been called bcc for longer than the GPL'ed bcc has existed. If someone has to change their name, shouldn't priority go to the older usage?
No. The demo was on the big screen at the convention center ballroom. There was no opportunity to inspect the setup or see much of what was going on.
Remember, it's very early in the development cycle. Who knows what kind of jury-rigging was used to get it to work at all. And I don't think they're really trying to put down gcc. They just want to show that they have a compiler and it has strengths. We'll have to wait for an actual product before we can really see what the deal is.
Your company would FIRE you if you ran a browser on a server machine? Would you mind disclosing where you work, so I can take steps to ensure that neither me nor any of my friends ever, ever go to work there?
Yeah, CP/M had drive letters like DOS, and (if you weren't running ZCPR3 or something) you might very well see an A> or B> prompt under CP/M.
But it's very unlikely you would see a C> prompt. Upder CP/M the only way that would happen is if you had three floppy drives. Or a sense of humor.
But remember, this movie wasn't made for us. It was made for Them. They don't know the difference between a C> prompt and a paper tape, it's all just neat-looking props.
Slashdot Mirror
Napster is an enabling technology that allows its users to break the law. It is generally legal in the U.S. to sell things to people even when the purpose of the things is to break the law. Selling radar detectors is legal in most states, even though the only possible purpose of a radar detector is to help its owner break the law. Are radar detector manufacturers liable for contributing to millions of cases of speeding? If not, why should Napster be liable for millions of cases of copyright infringement?
-Graham
Actually, as I recall (someone check me on this) there was a v6 ODBC driver on the last beta that was sent out before the open source decision. The problem is, Interbase Corp. doesn't own the rights to it and therefore can't open-source it, at least not without permission. It does bring up a good point: What will the licensing terms be for the 6.0 ODBC driver? I think I'll go ask on their newsgroup...
-Graham
That could be a lot of copies. Top management probably has some reservations about this whole weird "open source" thing. Throw in a couple hundred thousand on the wrong side of the ledger and you might wind up with obsolete binaries rather than released source.
Or they could simply say that the "end of life" for the product is (some deadline) and promise to release the source once that deadline is past.
Certainly. Of course, depending on the situation, that might well result in the eventual source release being even further delayed.
If nothing else, couldn't the first open-source version be the minimum cleanup necessary from the last proprietary version, not a development version with lots of half-finished features?
It depends on the situation. If the new version is a total rewrite and doesn't work yet, well, sure, the old version might be interesting. But most of the time, I'd want the latest code. Specifically in the case of Interbase, 6.0 is in late beta, feature-complete, and close to releasable; in fact, they might wind up putting out the source and the 'official' 6.0 release on the same day.
-Graham
The scary thing about this is that if the same situation were to arise today, the story would be very different. With DMCA, UCITA, etc., IBM would almost certainly have been able not only to win licensing control from CMU, but also to prevent MIT from suceeding with the X project. Certainly, good software needs to evolve. Clearly, though, the only reason it's been able to do so is that the greed-inspired have had no way to stop it; Bill Gates famous phobia regarding "some guy in a garage" was actually well warranted. Until now.
-Graham
I don't think this stems from laziness on the part of the company doing the open-sourcing. The problem is, there are existing contractual agreements with distributors, VARs, etc., regarding the current product release. It is these people, not the software company itself, who are often the main force behind keeping a product proprietary. The software company might be able to find other revenue sources, but the distributors are just flatly out of luck. Keep in mind that up to the moment of making the open-source decision, these people have been the software company's major customers, and the full weight of customer service culture pushes towards doing what they want--it takes a clear and powerful vision to make a decision that you know is going to totally alienate all the customers that built your business to the point it is today. And quite frankly, it's absolutely the wrong thing to do, 99% of the time.
So what it boils down to is, if you open-source the existing version, you cause all the existing inventory at the distributors to have a value of zero; this constitutes damages, and is probably actionable. Expect your pants to be sued off shortly. If you open-source the next, as yet undistributed version, all the distributors angrily drop your product (doesn't matter how much mark-up you give them if the MSRP is $0.00), but they don't have a case against you for any real damages, since you haven't done anything to reduce the value of their existing inventory, assuming you continue to claim, without snickering too loud, that the old version will still be sold at full price.
And whatever happens, you'd better do a flawless execution of your open source business plan, because the bridge you just rode in on is now merrily in flames.
-Graham
Not to start a flame war, but you're much more likely to succeed in hooking MS-Project to Interbase. All the stuff on this thread that's being claimed for PostgreSQL some time in the future, Interbase already has. My company has run on Interbase for years. The Philadelphia Stock Exchange uses an Interbase back-end. Now it's (shortly going to be) open source. What more could you ask for? Yes, PostgreSQL is a worthy database in many ways, certainly more so than MySQL. It's a serious contender for Web-served applications, for example--maybe even as good as Interbase, as long as you don't want to do more-than-trivial stored procedures, sub-select queries, domain-based schemas, triggers, event alerters, long-duration isolated transactioning (i.e. snapshots), or any of the other things Interase does well but PostgreSQL does not. But have you ever tried to connect PostgreSQL to a Windows client? The ODBC drivers, quite frankly, suck. Before Interbase was free, I actually deployed PosgreSQL/ODBC to an office full of Windows clients. I had unsurmountable problems with stability, compatibility issues, etc, etc. The database itself was fine, but the Windows client support was flatly unacceptable. This was about a year ago, maybe it's improved since then, but the simple problem is: Why would anyone on the PostgreSQL team be motivated to write for Windows?
So if you want a free Linux database serving Windows clients, there simply is nothing to compare to Interbase. You can download the binaries free today, and the source will be out as soon as they've built a makefile that works on anyone else's box, removed all the non-politically-correct comments, got final sign-off from the lawyers, etc, etc. The Web site is www.interbase.com.
-Graham
Strangely enough, many other countries seem to have developed fairly reasonable compromises on this issue. I don't recall this being front-page news in my time in Canada or the UK. The Germans, Japanese and French have all pretty much figured out systems they're happy with. Why is it so hard for the U.S. to accept that in some cases, collective freedom (ie, things like physical safety, medical care and clean air) can be more important than individual freedom (ie, ownership of handguns, low taxes, unrestricted consumption)? That this balance must be struck in all societies? That restriction of some individual freedoms is necessary, not because Big Brother wants to make you unhappy, but because we the people desire not to have our children shot by that handgun you so desparately insist on keeping and bearing?
-GrahamThe interesting questin here is what happens when Congress passes two obviously contradictory pieces of legislation. The DMCA has specific exemptions that permit copying to and from computer hard drives. What happens when two laws of equal priority (ie, neither is a constitutional amendment or whatever) are in obvious conflict? If you sue me under AHRA, can I defend myself using the provisions from DMCA?
Not that I'm saying DMCA and AHRA are necessarily contradictory; I don't know a thing about what AHRA says.
-Graham
Several comments have referred to the timeliness of the Web as compared to newspapers. It's certainly easy to get addicted to news-as-it-happens, but unless you're a political campaign manager, FEMA director, Fortune 500 CEO or (real) day trader, newspapers are already more than fast enough to get you the information you actually need. So this wasn't a major concern for me. I would be just as happy waiting until after Super Tuesday before awareness even dawns that the primaries have started. Certainly, hearing about the results the next day is not a problem for me.
The other major issue seems to be quality of content. Again, I don't find this to be particularly relevent. After all, most of the newspaper articles are the exact same AP or Reuters wire feeds that you can read on Yahoo. It's true that the Web has some really good specialized news sites, but on the other hand, good newspapers have excellent investigative reporters. In the end, it comes out about even in my book.
Katz and a few other posters seem to be unduly concerned about the lack of graphic content in newspapers. I would like to point out that the Web isn't any more graphical than a newspaper. How many of Katz's articles on Slashdot have ever had so much as a single accompanying photo? Does anyone else notice that the photos accompanying Yahoo news items seem jarringly out of place somehow? The Web is clearly a textual medium, just like a newspaper--the opinions of Web pundits with fond memories of the days when Wired was actually cool notwithstanding.After all, do you watch Slashdot or do you read Slashdot? The Web is therefore a much better place than, say, television, for the presentation of complex and difficult topics. But so are newspapers. Print journalism is in a very sad state at the moment, but this is not inherent to the medium; there used to be excellent reporting in the papers, and if the Web figures out how to do it well, the papers can for the most part pick it up without any trouble. Again, it's a wash.
If these weren't the reasons, why did I stop my subscription to the Raleigh News & Observer several years ago? The main reasons are:
The physical act itself
I've never found reading a newspaper to be a very comfortable experience. It's too big. If you're in a crowded area, like a typical RTP lunch spot, you practically have to take up two seats. Newspapers were obviously designed to be read in overstuffed armchairs. I don't own an overstuffed armchair. I'm perfectly comfortable reading a magazine or novel over lunch, but a newspaper is too cumbersome. Actually, the only way I really enjoy reading a newspaper is if I take up the whole dining room table with it. Big, slow and spread out all over the place just doesn't fit my personality or lifestyle.
The lack of searchability
I don't read the paper for its entertainment value. My recreational reading runs to novels, the Atlantic Monthly and Scientific American. I (used to) read the paper to keep up with current events in areas which interest me. However, newspapers only provide one index method (so to speak) - the date. I frequently find myself wanting to know what's happening in some specialized area, like: what's happening today in Kosovo. I'm not a passive vessel to fill with whatever the newspaper chooses to publish that day. I'm an active, self-directed seeker of information. I want to be able to ask a question and get an answer. I can do this with the Web, but it's hard to see how a newspaper could work this way.
Local news
I'm going to go against the grain here and say that I much prefer getting my local news from the WRAL-TV Web site, rather than the News & Observer, so I credit this as a strenght of the Web. I understand this might not be true in other areas, because they might not have a resource like WRAL. However, I think WRAL has made it clear that the Web has inherent possibilities for local news reporting that newspapers can't duplicate. With this year's hurricanes and snowstorms, I want to know, in the evening, whether the kids' schools will be open the following morning. As I get ready to leave for work, I want to know, at that particular moment, what the traffic situation is like on I-40. I want to know what that big construction project I saw driving to work was all about. I want to see the milestone dates for the Outer Loop--and whether they're being met or not. And I want it all to appear at the moment I want to know, not when the paper happens to run a story.
Heavy-handed marketing
The News & Observer went through a couple years where they made extensive use of outbound telemarketing as a promotional vehicle. I'm not one of these knee-jerk boycotters, but when a firm interrupts your dinner half three times in the same month to pummel you with a hard-sell for a product you already buy from them, you tend to notice. It's not enough of a big deal to make me stop the paper for just this reason, but the ongoing sense of vague irritation was certainly a contributing factor.
Editorial bias
Everyone in this area says the N&O is too liberal. Personally, I find it far too conservative. But that's not the problem. The problem is that by having a single reporting structure under an editor-in-chief, every newspaper can't help but show some kind of editorial bias. When I want to read opinion pieces, I'll pick up a magazine. I'm interested in a newspaper for factual reporting of what actually happened. Much too often, the interpretation and opinion gets in the way of telling the story as it actually occurred. Reading newspaper pieces, I often find myself disagreeing with the writer and wishing I had some source of simple factual data to check up on it for myself - so I go to the Web. Once this had happened often enough, I started going to the Web first, rather than reading the paper at all. When I found that a whole week of papers had come in and I hadn't read a single one, I decided to quit.
Incidentally, now I know why Katz and Jesse Burst and so forth close their articles with "opinions? talk back by clicking..." type statements. It's much easier than thinking of an effective closing paragraph.
-Graham
Think about it. How much free software would be written if all free software authors were required to maintain multiple T1s to handle the load, if they happened to write something popular?
Now, if they try to say that students aren't allowed to redistribute CAEN Linux from a student, or whatever, they're violating the GPL. But they haven't done that. They've just put it on an on-campus-only download site. If you offered to mirror it off-campus, no doubt you could get a copy.
However - if it's anything like Eos Linux from NCSU, it'll be full of site-specific code that's meaningless outside the UMICH network. So why would you want it?
-Graham
North Carolina State University has had a group of students supporting a Red Hat-based distribution called EOS Linux for some time. Just recently, the university announced that they would begin treating it as an officialy supported platform. Details are sketchy - the announcement just came out - but it's a pretty big deal for the NCSU folks. Perhaps some Eos Linux people can provide more depth.
Actually, recent versions of EOS Linux aren't distributions, but a set of rpms that you install against a specific version of Red Hat. The Eos Linux group felt it would be easier to it this way, rather than taking on all the day-to-day maintenance responsibilities of supporting a full distribution. I think this is a better model for site-specific Linuxes. It's a good discussion topic, anyway.
Last but not least - I don't know if it would be possible to build a global "University Linux" as the article suggests. The main benefit to Eos Linux, and presumably to the University of Michigan Linux as well, is that it integrates tightly with all the on-campus systems and networks. It's site-specific, and that's why you want it. It's hard to imagine how you could build a single distribution that would be site-specific at all sites.
-Graham
Hey, hold on a minute, there. Remember how Netscape performed before they open sourced it? On my machine, it crashed multiple times a day. It felt like a race to see if you could get the answer to your question (or whatever) before the browser died. We all knew, well before the source was ever made public, that it was a pig.
Interbase is not like that. Interbase is one of the best-running, bug-free, and robust pieces of software I've ever worked with, at least from a "black box" perspective. I'd love to see the source, and I'd give you long odds indeed against it turning out to be a "pile of crud."
-Graham
I'll work on it. I'm a database developer and administrator at a 400-employee firm, and our database is Interbase. Technically, it's unbeatable. We are not looking at another Netscape fiasco here (unless Inprise screws it up somehow). Navigator was practically unworkable. Before they even released the source, everyone knew it would stink, just because of how poorly Navigator performed. Interbase is not like this. The _only_ thing wrong with Interbase is a lack of awareness/marketing/mindshare. And I can tell you that it scales flawlessly to at least tens of gigabytes and four-way SMP (though not on Linux). I can't comment on nutty mainframe-like setups, but if I had the money and time, I'd love to see a two- or four-CPU server, with a gig or so dataset, running performance and stability comparisons between Interbase and anything else in the world. Open sourcing Interbase would give me a bigger woody than anything else that's happened this year or is likely to next year.
-Graham
Well then change it back! Fast!
Requiring an unique IP would exclude all but one user behind each NAT-based corporate firewall. I have 300 analytical chemists here, many of whom play chess, and quite possibly more than one of whom might participate in something like this. Do you think they would be happy to learn that only one person in our company would be allowed to submit votes?
I know that people will disagree with this, but the X server is like the kernel in that it is supposed to be able to deal with *anything* that a misbehaving user application could throw at it. If you're non-root, there should be no possible code you could ever execute that crashes the kernel *or* the X server.
So the best you can say is that if there's an app (cough...Netscape...cough) that crashes X all the time, then *both* the app and the X server have a bug.
-Graham
The difference, of course, is that McDonald's Corporation is not trying to follow a free software ideology. Red Hat has stated publicly, on numerous occasions, that anyone can copy the distro and sell it for money. So the question now is, if people choose to do so, what rules do we want them to follow? It would be much better if these were stated explicitly up front in a defensible way. Given the choice, who do you want to set the standard for interpreting copyright and trademark law in the context of free software: a member of the community like Red Hat, or the court system? As you consider your answer, consider any three recent computer-related precedents set by the courts...
How much effort do you have to go to in order to be "calling it" something different? Is it sufficient to just put a different name on the outside of the CD, or do you also have to update /etc/issue, redhat-release, etc? What are the rules about what the name can be? Could you, for example, make an exact copy of the Official Red Hat Linux 6.0 CD, label it "Unofficial Red Hat Linux 6.0" and then sell it? Or would you have to do more than that?
Also, does it have to actually be Red Hat Linux in order to call it Red Hat Linux? For example, could you make a copy of Debian 2.1 and sell it under the product name "This Is Not Red Hat Linux?" Or could you make copies of FreeBSD and market it as as "Red Hat FreeBSD" - not because it has anything to do with Red Hat, but just to gain some of the benefit of Red Hat's brand identity? If these extreme cases are unacceptable, where exactly is the line drawn and how do you know if you've crossed it?
-Graham
I've never used OpenLinux, but I work in a Novell shop, and I was quite excited by the NetWare port that Caldera released for Linux. I was even prepared to spend good money on it. Unfortunately, they released it as binary only, it required their customized 2.0.35 kernel with STREAMS support, and they never did one lick of work to support it. Fortunately, I noticed this before I put down any actual cash.
But at Linux Expo in Raleigh, when I made my concerns known to their sales reps, they told me to wait a couple months for some really exciting announcements that would change my mind about Caldera.
I assume this is it. Yawn.
-Graham
We run Novell Groupwise, and it stores everything, including calendar information, in the same message store, all of which is accessible via IMAP. Schedule items are nothing but e-mail messages with a second set of RFC822-like header lines at the beginning of the body text, so if you read your mail via a non-Groupwise IMAP client, you just see "Time:" "Location:" etc. headers at the top of your message.
There's no reason in the world why this would break anything, and it's open and accessible to anyone who wants to mess with it (eg: I've written Perl scripts to check my own calendar and page me via qpage if I have a meeting coming up that meets certain criteria for being important).
If I were a wide-eyed innocent, I would say something like "Gee, I don't understand why everyone doesn't do this."
-Graham
Me too, actually. And while I was never an OS/2 fan, a lot of OS/2 people are quite unhappy about the way Borland dropped that platform. It's true that Borland could burn a lot of people by shipping a Linux product and then dropping it later. But the interesting thing is, they will then have done a lot of work teaching us (the community) why a RAD tool is useful, how it should work, etc, etc. If they release and then drop a Delphi-for-Linux, they'll have done a great deal towards the free Delphi-clone-for-Linux project that will result. Nothing like a huge number of frustrated developers with an itch to scratch...
Borland's compilers have been called bcc for longer than the GPL'ed bcc has existed. If someone has to change their name, shouldn't priority go to the older usage?
No. The demo was on the big screen at the convention center ballroom. There was no opportunity to inspect the setup or see much of what was going on.
Remember, it's very early in the development cycle. Who knows what kind of jury-rigging was used to get it to work at all. And I don't think they're really trying to put down gcc. They just want to show that they have a compiler and it has strengths. We'll have to wait for an actual product before we can really see what the deal is.
-Graham
Your company would FIRE you if you ran a browser on a server machine? Would you mind disclosing where you work, so I can take steps to ensure that neither me nor any of my friends ever, ever go to work there?
-Graham
Yeah, CP/M had drive letters like DOS, and (if you weren't running ZCPR3 or something) you might very well see an A> or B> prompt under CP/M.
But it's very unlikely you would see a C> prompt. Upder CP/M the only way that would happen is if you had three floppy drives. Or a sense of humor.
But remember, this movie wasn't made for us. It was made for Them. They don't know the difference between a C> prompt and a paper tape, it's all just neat-looking props.