This is not the first time the ketchup comparison has been made by Young--the first time is in O'Reilly's book Open Sources, I believe.
The full analogy is something like this: Anybody could make ketchup on their own if they wanted to, but most people don't want to take the time to do it, so they'll buy it bottled. For practical purposes, ketchup is ketchup, and it doesn't matter what the label says on it. But Heinz still has 80% of the ketchup market, solely from brand recognition: they've gotten the buying public trained to think of Heinz when they think of ketchup.
So what's this mean for Linux? Essentially, Red Hat's business strategy is to become the Heinz of the Linux world. Linux's licensing status makes it a commodity OS--anyone can package it or even "roll their own" without a distribution--just like (his argument goes) ketchup is a commodity. But if people think of Red Hat when they think of Linux, they'll be the market leader. Quality isn't irrelevant, of course, but it's not as important as the network effect: "everybody else uses Heinz, it must be the best."
(One presumes this makes Caldera the Hunt's of the Linux world. Debian is an organic farmer's co-op that sells their ketchup in Mason jars.)
I think some of what you're seeing here is simply that Opera supports HTML 3.2, not HTML 4.0. I find this a frustrating choice on their part at times, but most of the time it isn't an issue. It does let out Dynamic HTML, though, and there are occasionally pages that I hit that simply don't render in Opera.
Opera 4.0 will theoretically support HTML 4.0 and CSS Level 2. We'll see.
As an unrelated note, Opera ports are in progress to a fair number of platforms. It's quite possible that the reason it took so long for them to start porting has less to do with code portability than it does with the nature of commercial software (particularly something as Quixotic as a commercial web browser)--you do ports when you (a) believe there's a market there and (b) have the resources to do so. They've addressed point (b) by contracting with other development companies.
Because of that (the subcontracting), people should probably nag the Opera Unix developers if they don't like MDI. The BeOS version of Opera uses standalone windows like other browsers do (and it still implements Opera's cascade and tile buttons, just in case someone wants to claim you need MDI to have that functionality).
While I'm not a huge fan of national (or multinational) corporations in general, Disney has had the rather unusual experience of being bashed for decades because they are Disney. There's very few corporations that have reached the status of cultural icon, and Disney is perhaps unique among that handful--their primary "product" is stories and experiences for children. Generations of Americans have grown up on Disney.
Because of that, it seems people have unrealistic expectations for everything they do--too high and too low, depending on how the viewer regards them. This is how a movie like Pocahontas can be fiercely attacked for being "insufferably politically correct" and "stereotyping Native Americans" at the same time. Never mind that Dances with Wolves was starkly black-and-white in its presentation of "Indians good, White people bad" in a way which Pocahontas avoided--Kevin Costner can do what Disney can't. (At least until "Waterworld," but we digress.) As for being an unrealistic portrayal of Indians, as Mel Gibson put it, "it has a talking tree in it, for God's sake."
With Celebration, I think people have had similar sets of unrealistic expectations. If the town had been set up by another major company with comparable resources--AT&T, say--it probably wouldn't have been subject to the same firestorm of criticism, and on the flip side, it probably wouldn't have attracted people expecting it to be nearly utopian. Pocahontas is just a movie, and Celebration is just a planned development.
Personally, I wouldn't mind living in Celebration's apartments. Sure, you'd probably have to buy some things from stores along U.S. 192, a couple miles away. So what? Unless you're working in its modest office park (or in retail), you're commuting to work anyway. And I like the idea of living near a pretty central park area within walking distance of a grocery shop, bakery, bookstore, coffee shop, movie theatre and four restaurants.
It's not Utopia, and it's not something that couldn't be duplicated elsewhere, and maybe duplicated better. But despite that, and despite the inevitable problems with a community like it (it's not the first "New Urban" town but it may be the first on its scale, and certainly the first with the level of media attention it gets), Celebration does work, and if it inspires other places to try similar development, I think it'll only be for the better.
Press that hot-topic button!
on
Quack!
·
· Score: 3
Sigh.
I don't think it's bad to have a "features columnist" for Slashdot, and one who writes about the intersection between society and technology is a good choice. People of the True Engineering Mindset often ignore that intersection, and that can lead to the "all problems will be solved by technology" mindset which ignores the fact that the problems technology creates often aren't technological problems.
But--okay. Maybe it's just me. What I've seen since Jon was writing for Slashdot, though, has roughly followed this path:
Various articles with varying degrees of insight.
Complaints from readers.
More strident articles, whose insights became pithier and, well, less insightful.
More complaints from readers.
"Voices from the Hellmouth."
Hundreds of "Yeah! Right ons!" from readers and national media attention.
Dropping of nearly all other technology issues and all pretense of analysis to constantly sound the alarm of "young people's rights."
I don't like to accuse Jon of being an opportunist, but it's difficult not to start drawing that conclusion. Sounding shrill alarms about movie theatres actually stopping people who are under 17 from attending R-rated movies is honestly a little dubious; you may not like the rating system, but it's not exactly new. And it's not even a government agency--the power of the MPAA is a creation of market forces. (Nobody forces studios to have movies rated, and nobody forces theatres that play unrated movies to police the age of attendees. You just won't make any money if you're limited to showing at the few theatres that play unrated movies.)
But now we're going beyond that, and sounding the alarm about the new great force for fascism in the country: pediatricians. Yes. You thought Dr. Spock was well-intentioned, but no, he was Mussolini with a lollipop!
Come on. Recommending that parents not put TVs in the bedrooms of their preteens and not show any children under the age of two television shows is an assault on free speech?
Really?
Jon, I understand the desire to be popular, but couldn't we get back to the pithy insights about technology sometime?
Actually, the specs for the G3 processor are not the problem. BeOS will run on Macintoshes with G3s in them--the compatibility problems lie in the surrounding hardware.
To make an operating system run on Mac hardware, you need specific design documents from Apple. (There is a name for them, but I don't remember it right now.) These documents are not publicly available, and concern the deisgn of the logic board, custom logic chips, address spaces, and so on.
In the pre-Jobs days of clones, these reference documents were available to a lot of Apple partners. After Jobs' return, it was decided these documents should only be made available to MacOS licensees, so that they can build MacOS-compatible hardware. This cut Be off from the information (and, indeed, there are no more MacOS licensees anyway).
Be has requested this information repeatedly, going up through various levels up until Gassee emailed Jobs personally and got no answer. This was around the middle of 1998, as I recall, and they haven't made any effort since then.
So does Be really want to be helped in this? I don't know. There's no love lost between Gassee and Jobs. Be has stated publicly that they could make BeOS work on Apple's G3 hardware without these documents--but they're not going to, unless Apple is willing to support them directly.
The flaw with this argument is not technical. It's that the typical business user does need Windows applications, if for no other reason than that they've been using them since the early '90s and you won't find any other platform that runs them. I'm not just talking about the ability to open Word documents. I'm talking about the service order entry system your company has been using for the last five years built around Microsoft Access. The spreadsheets lurking in every third or forth director's hard drive that contain Visual Basic for Applications programs. The weirdass vertical market application that looks like a star of "It Escaped From The Mainframe" that all your field service personnel depend on.
If you were building a business from the ground up you would be better off with non-Windows solutions, although you should be prepared to deal with the Word, Excel and occasionally PowerPoint documents that businesses you work with will send you.
Personally, if I were doing that business from the ground up, I wouldn't give everyone Linux, because the desktop application support still isn't quite there. (Yes, I spent a lot of time playing with StarOffice, and it ultimately sold me on Gobe Productive for BeOS. StarOffice is pretty and powerful, but my God, it's like a hog in molasses!) Servers, sure. Give the people who know what they're doing with them Linux boxes. Give the people who want to run office applications iMacs.
Getting xfstt up and running cut down on much ranting on my part about how pug-ugly X's font displays are. It's still not as nice as BeOS's display--no anti-aliased fonts, and some applications are a touch on the clueless side when it comes to using TrueType (Netscape being a notable example). Even so, it made a world of difference. If you're using Red Hat 6.0, I understand it comes with xfstt installed, and even a logical arrangement to share fonts between applications that logically should be sharing them, like X and Ghostscript. Somebody else could correct me if I'm wrong on that. RH51 didn't, and getting it running was a bit wonky, although not too bad if you pay close attention to the instructions. (I thought I had, but we won't go there.) When I switched to Debian 2.1, I discovered that when I installed the xfstt package it automatically configured itself and came up the next time I started things. The most work I had to do was to look in/usr/doc/xfstt to find what directory I needed to stuff the font files in. Somebody tell me why the Debian distribution is supposed to be the difficult-to-use one again?
Uh, perhaps the affluence there has affected your vision. The only theatrical animated releases Disney has made under their name that haven't been from one of the two Disney Feature Animation studios have been "Duck Tales: Legend of the Lost Lamp" and "A Goofy Movie," which were done by their television division. There may indeed be a Disney TV Animation Canada, but they would be one of several contractors working on various TV series and direct-to-video releases.
Yes, some movie theatres are actually making an effort to enforce the ratings now. Yes, that'll probably piss people off. Yes, ramping up the enforcement because of skitterishness over school violence is silly. But none of those things present a convincing argument for abolishing the MPAA ratings system, and they're not even particularly convincing of the apparent argument that children should be let into R-rated movies. Sorry, I just don't think the right to free speech is endangered by telling a 13-year-old he can't go see someone masturbate into a pastry.
As for the argument that Mr. Winter's DVD home theatre with the GNOME interface is the harbinger of doom for the movie industry, sorry again, Jon. Last weekend was the biggest box office weekend in history in America. People are not failing to go to the movies, and it seems that those teenagers being turned away from The Blair Witch Project aren't significant enough to affect the bottom line.
I can't address the argument that liking "South Park" is a form of asserting one's individuality without collapsing into giggle fits, so I won't try.
...people are absolutely correct when they say installs have gotten very easy. In fact, they've been pretty easy; most of the changes in installation programs I've seen over the last couple years have been more cosmetic than functional.
The usability issue is configuration after the install. It's things like being able to use a control panel to change your resolution, color depth and refresh rate to anything your card is capable of on the fly and have it stick (sorry, XFree86's hot keys just don't cut it in that regard). Like being able to install a new font by dropping it in a directory and having all the programs on your system be able to use it for both screen and print.
This extends to applications, too. KDE and Gnome are making strides in the right direction, but people shouldn't underestimate how nice it is for all but the most diehard CLI-or-death types to be able to do something (seemingly) simple like select font and font size for applications while they're running with a few mouse clicks, or creating symbolic links with drag-and-drop. Linux developers should not disdain the UIs of MacOS and BeOS, they should start learning from them.
Of course he's trolling, because there's nothing more fun for a columnist than getting a bunch of people frothing at the mouth. It's probably one of punditry's best perks to watch people go out and spew righteous flame with relatively little clue.
Bob Metcalfe has been the editor of Infoworld for quite some time now, including the time when Infoworld was the only corporate computer magazine taking Linux seriously. They reported on Novell's Corsair project--you know, the one that became Caldera--on their front page when other Microsoft-happy journals were ignoring it or, in PC Week's case, subtly slamming Infoworld for reporting on such nonsense.
Look, follks, every time someone casts doubt on Linux winning the world doesn't mean they're in the pay of Microsoft. It doesn't necessarily even mean they don't like Linux. Metcalfe likes to play the devil's advocate--he's been doing it for years. (Unlike folks like John Dvorak, Metcalfe also does admit when he's wrong, even going so far as to eat one of his own columns when he made some dramatic prediction that didn't come true.) But the record at Infoworld under his editorship doesn't exactly suggest an anti-Linux atmosphere, and it doesn't really suggest a particularly pro-Microsoft one. Remember, this is the publication that named Linux their top network operating system last year--and one that for several years kept naming OS/2 the best business operating system over Windows. (It also won for six years straight, I believe, in the readers' choice awards, suggesting that most of Infoworld's audience isn't comprised of Microsoft lackey's any more than the staff is.)
And, sure, Metcalfe cares about attracting attention. That doesn't mean he doesn't care about being right. If you're a pundit, you call 'em as you see 'em. You may see 'em wrong. Does Metcalfe see it wrong this time? The danger of the attention storm that Microsoft will undoubtedly try to create around Windows 2000 should not be underestimated in its capacity to derail the attention Linux is getting now from corporate users, and that's the mindset ol' Bob is writing about when he talks about the potential of Linux "failing"--not that Linux will blow up and dry away, but that Microsoft will be successful in putting up a brick wall for it to run headlong into.
Look: you can either puff up at the "trolling" and squawk a lot, or you can look for the points Metcalfe has that maybe should be kept in mind as Linux moves forward.
While I found Katz's "nightmare for Libertarians" comment to be typically uninformed, this "the world doesn't owe anything" huff-and-puffery is also a typical net-libertarian response. Why is it anytime someone makes any suggestion like "monopolies are bad," much less something truly radical like "maybe the way to reform government would be to make it more efficient at helping people rather than reducing its role to that of enforcer," this old chestnut comes out? Who said the world owed anyone anything? News flash: us bleeding hearts know the world ain't fair. We just have the quaint notion that working to make it a little fairer isn't a bad thing.
The American economy over the 20th century has been an absolute marvel, and Libertarians need to open their eyes and notice that that economy has been mixed capitalism, not pure capitalism. The IMF has been forcing pure capitalism down the throats of countries they've been trying to help over the last decade or two, abandoning the principles of the Breton Woods compact, and by any honest accounting it's been an absolute disaster. We gave them our theory, not our practice, and it's dismaying to see presumably well-intentioned people push those theories so hard here.
The market rewards mind share more than it does good products, and investors only reward high profits and continued growth--how that profit and growth is obtained is irrelevant to the reward. But if it's obtained through means other than providing the best possible product at the best possible price, very strange effects start happening, Microsoft being one of the comparatively innocuous ones. (HMOs are one of the less innocuous ones.)
Incidentally, I don't despise Bill Gates for his wealth. (This is another bugaboo with libertarians, I've noticed--mention "growing income gap" and they start frothing at the mouth, as if merely noting its existence is a call to sell Texaco's assets to the homeless.) That doesn't mean I can't despise him for flogging useless garbage, and you know what? You can too!
So Leonard praises Slashdot in a manner which is not only effusive but accurate, and you call him a BIG FAT PRICK. Clever. Today's thought for the day, children: does this kind of logic discourage "anti-Linux FUD" by the press, or does it encourage it, particularly of the sort criticizing the Linux userbase for being hypersensitive, highly unsocialized and apparently illiterate?
If the underlying operating system is as irrelevant as they seem to imply that it is, I'm not sure we should be looking for significant kernel modifications any time soon. I think they went with Linux as opposed to QNX or BeOS for three reasons, all of which boil down to "free" as in "free beer":
No licensing fees.
It's getting a hell of a lot of "buzz" about it right now, and they get some free marketing because of that.
Other buzz-aware companies are trying to jump on the Linux bandwagon, too, so again Amiga gets a free ride. Assuming the MCC can run x86 binaries (not guaranteed, but hardly impossible), it'll have more software out of the starting gate than BeOS does--and certainly much more than if they'd gone with QNX or Chorus.
I'm interested in seeing what happens, although obviously I'm a bit jaundiced (well-funded or not, this is an awful lot for a company that was less than a dozen people at the start of this year to deliver on this fast--a lot of this strategy apparently doesn't go back much further than Jim Collas' involvement).
But you're right: if nothing else, we'll end up with another desktop manager with unique features that aren't compatible with Gnome and KDE. To hell with those "closed" desktop environments--why should we make it so application programmers have any reasonable expectation of what services the GUI they're running on can deliver?:-)
Granted, unlike iToaster the new Amiga (if it happens) will presumably actually use Linux instead of just sticking its name in the press release, but one can't help but suspect it's being done for similar reasons. Linux is the buzzword to be using now if you're challenging Microsoft.
And, if you need to get a new operating system out real fast, what better approach can you take than to use someone else's operating system... that you can use for free?
Personally, I'm skeptical as to whether or not the "great new hardware" for the next generation Amiga is still in the works, too. What about all the developers that are supposedly waiting in the wings to support the system? How many of them were developing for QNX? How many of them were developing at all? Will all of them be willing to develop for Linux? Will they be able to develop for the Amiga "OE" without developing for Linux itself?
That sounds like an odd question, but it isn't--there's no reason to believe this "operating environment" will also be GPLed. Of course, if it isn't, it'll have to run in a "closed box" on top of Linux. Is this going to be able to compete in real-time performance with BeOS? Or for that matter, QNX?
While you've raised some pretty good points, I think that BeOS R4.5 does have an adequate level of hardware support for common configurations. Not a good level, but an adequate level.
The SCSI support isn't that dramatic, but it supports most of the Adaptec chipsets and some Buslogic and Symbios chips.
It supports a fair number of the major sound cards, too--the only popular ones missing at the moment are SoundBlaster Live and Aureal-based cards, both of which should be out soon. (The SB Live drivers already exist, but haven't been released).
While it doesn't support hardware acceleration on anything but Voodoo 2/3 cards yet, that's not the same as saying the cards are unusable. It has 2D acceleration on many popular cards--and can use nearly any card that's VESA 2.0-compatible.
Would it be better if it could do hardware-accelerated OpenGL on TNT-based cards? Sure, and I suspect there'll be drivers that support that for non-Voodoo cards before R5.
But even if you just subscribe to the "media" schtick, the lack of 3D hardware acceleration isn't that important if your definition of media is, say, audio or video editing. With those the framework is already there, and in practice, not just theory. (Visit Cirque de Soleil in Orlando, or the Broadway production of "Ragtime," or the video-editing demonstration at the ZEUM hands-on museum in LA.) What about web design work? Or 2D "cel-style" animation with Lost Marble's Moho, a program I haven't seen the likes of on any other platform yet?
Sure, there are holes there, but it's not like the initial Intel release anymore. We're talking potholes now, not gaping sinkholes. And, I agree with the commenter who disputed that remote administration and multiuser capability are make-or-break features for a large percentage of the audience. (They're certainly not important to me--I'm running Linux now, but I'm just one user, after all. And maybe your office is better than mine, but at work I really don't want my company's IS department trying to adminster my PC for me. NT is annoying enough without SMS futzing with it, thank you very much!)
While I've seen this argument before, I'm not sure I buy it. Office got mindshare at first because it was the only "pro-level" game in town for Windows--Word and Excel on Windows predate WordPerfect and 1-2-3 on Windows by at least two years, and the first Windows versions of both of those were terrible. By the time anyone mounted serious competition, both of them had over 50% of their respective markets on Windows, and Microsoft started using aggressive bundling tactics to make sure that many users had Office installed with their computers when they got them.
The "only idiots want to use too-friendly software" argument belies the fact that bloated or not, Word and Excel are very powerful programs. All of the obnoxious automatic features can be turned off quite easily, and there are fairly powerful features--outlining, mail merge, revision changes, groupware routing--that are as close to intuitive as you can get in Word. And I haven't seen a better spreadsheet than Excel yet. It does useful things that no other program in its class does; while there are better programs just for statistical analysis and just for flatfile databases, for example, there's no program I've found that does both of them together as well and as easily. I write statistics-gathering tools for our company's frame relay networks (information about which is scattered in SQL and Access databases), and Excel is simply better tool than Access for such things. (Its database handling is nearly as powerful, and its reporting tools are better.)
Scott Hacker, for example, provided no data on the percentage of flames versus constructive criticisms. It may be that the negative comments were not at all representative of the whole.
While you're right, this is also somewhat misleading; he said the responses he received ran the gamut from complete agreement to obscene flames, and never implied that most of what he received was flames rather than constructive criticism. Beyond those introductory paragraphs, in fact, he didn't address flames at all, but instead did respond to the constructive criticisms, including acknowledgement of points he hadn't considered or wasn't clear on. Obviously people can still disagree with his conclusions, but just as obviously he's not just ignoring well-considered responses.
It should also be noted that consistently misspelling Scot's first name as "Scott" probably doesn't help things.:-)
I'd been looking for a good message to point BeOS users toward on how not to advocate for our favorite operating system. Thanks for this one--it's just perfect!
In other words, if you look through, the first format to hit critical mass with the product people buy it to use seems to win, unless a very compelling reason comes along to change.
This is the "network effect," and is probably more responsible than anything else both for Windows' de facto OS monopoly on the PC architecture and for why PC architecture greatly outsells other microcomputer designs. (I say "de facto" monopology because there have always been alternatives to DOS and Windows, yet asserting that Compaq had the choice in 1995 to install OS-9000, QNX or Concurrent CP/M-86 is akin to asserting General Motors has the choice to stop making cars and trucks that can run on interstate highways.)
While I don't think we're going to see the end of the network effect for Windows any time soon, I think we are seeing it start for Linux, and possibly for BeOS. None of these are likely to get bigger than MacOS, but that's had enough of a network effect to carry it through years of corporate mismanagement.
Windows' multi-hundred-million user base tends to skew people's ideas of what a sustainable market size is, I think; a market of only one million could sustain small companies, and it only takes a market of a few hundred thousand to sustain "cottage software" houses--something I think we'll start seeing more of in the future again. (What's old is new....)
Actually, Scot really doesn't seem to dislike Linux; he's written in the past about how BeOS and Linux have their own strengths and weaknesses. I think what happens with articles like this could be called a "defensiveness feedback loop"; I won't say there aren't BeOS zealots out there (I've seen some fairly wacky claims from supporters), but there are more Linux zealots (if for no other reason than there are more Linux users). And, Linux is unique in being blessed or cursed with a small subset of advocates who give "OS religion" a more literal meaning than we've ever seen in the past--the idea that open source is the only way to go because closed source is unethical. You see hints of that in this thread in the argument that Be can't be said to support open source if they only provide an egcs-based IDE, a GNU toolchain and the source to most of their drivers and applications. If it's not "total" support (i.e., all source opened), so this logic goes, it's no support at all. While this isn't an indefensible viewpoint, it approaches a religious argument (you're completely with us, or you're against us, and to be shunned).
As a staunch BeOS supporter, I'd say that Linux is more mature. Hey, you should call 'em as you see 'em. I think people tend to forget that despite the fact that Be, Inc. has been around since 1990, BeOS hasn't. DR1 came out in late 1994, and the first release that didn't require a BeBox was DR8.2 in January 1997.
There are things I like more about BeOS than about other operating systems, but I still can't do everything that I want in it. Of course, I still can't do everything that I want under Linux, either: there's no equivalent to the word processor I use (Nota Bene), and while the DOS version does run under dosemu, it doesn't run perfectly in console mode and runs even less perfectly under xdos. And there are tradeoffs even between the two operating systems. (I can't get TinyFugue to run correctly under BeOS, and I can't get a Linux version of e-Picture or Pe. BeOS sucks moose droppings if I want to use it as a server, but its GUI--whether or not people like the color choices--just isn't matched by Gnome or KDE. [For the morbidly curious, I duck that debate in practice and just use Window Maker.])
...why would you want to get BeOS off it and put Linux on it?
Well, I wouldn't, but I was giving what seemed to be a good answer to someone asking about the possibility of that.
What would be the point?
What would be the point of putting Linux on Furby? It's a hacker thing. (I know some are dreaming of $200 network gateways, but with the iToaster design that's not real likely to happen.)
The box was made to be used with BeOS; Linux isn't the end all and be all.
I think people need to stop thinking of this as a "$200 BeOS box." It's kind of like referring to the Philips "MyWeb" Internet TV device as a "QNX box." It's not that it's inaccurate, but it's somewhat misleading. Neither iToaster or MyWeb are meant to be PCs, even cheap ones. It's a different market.
BeOS is $70 from Be directly, although you can get it online at some other places more cheaply (PC Mall, I think, sells it for $53). It should be in retail boxes next month, I think, too, so stores may have it lower. It's not dirt cheap, but it's not horrid, either. (Red Hat 6.0 is more expensive if you get the "official" release.)
From what I've read about the iToaster, getting BeOS off it and putting Linux out it would be a challenge--it has no CD-ROM and only dial-up networking, and is about as expandable as a styrofoam brick. It's more "PC-like" than a real PC. It might be interesting to try to get Linux on it for someone up to the challenge, but success probably wouldn't bring you much useful--it'd have to be done for the "because it is there" feeling.
Slashdot Mirror
This is not the first time the ketchup comparison has been made by Young--the first time is in O'Reilly's book Open Sources, I believe.
The full analogy is something like this: Anybody could make ketchup on their own if they wanted to, but most people don't want to take the time to do it, so they'll buy it bottled. For practical purposes, ketchup is ketchup, and it doesn't matter what the label says on it. But Heinz still has 80% of the ketchup market, solely from brand recognition: they've gotten the buying public trained to think of Heinz when they think of ketchup.
So what's this mean for Linux? Essentially, Red Hat's business strategy is to become the Heinz of the Linux world. Linux's licensing status makes it a commodity OS--anyone can package it or even "roll their own" without a distribution--just like (his argument goes) ketchup is a commodity. But if people think of Red Hat when they think of Linux, they'll be the market leader. Quality isn't irrelevant, of course, but it's not as important as the network effect: "everybody else uses Heinz, it must be the best."
(One presumes this makes Caldera the Hunt's of the Linux world. Debian is an organic farmer's co-op that sells their ketchup in Mason jars.)
I think some of what you're seeing here is simply that Opera supports HTML 3.2, not HTML 4.0. I find this a frustrating choice on their part at times, but most of the time it isn't an issue. It does let out Dynamic HTML, though, and there are occasionally pages that I hit that simply don't render in Opera.
Opera 4.0 will theoretically support HTML 4.0 and CSS Level 2. We'll see.
As an unrelated note, Opera ports are in progress to a fair number of platforms. It's quite possible that the reason it took so long for them to start porting has less to do with code portability than it does with the nature of commercial software (particularly something as Quixotic as a commercial web browser)--you do ports when you (a) believe there's a market there and (b) have the resources to do so. They've addressed point (b) by contracting with other development companies.
Because of that (the subcontracting), people should probably nag the Opera Unix developers if they don't like MDI. The BeOS version of Opera uses standalone windows like other browsers do (and it still implements Opera's cascade and tile buttons, just in case someone wants to claim you need MDI to have that functionality).
While I'm not a huge fan of national (or multinational) corporations in general, Disney has had the rather unusual experience of being bashed for decades because they are Disney. There's very few corporations that have reached the status of cultural icon, and Disney is perhaps unique among that handful--their primary "product" is stories and experiences for children. Generations of Americans have grown up on Disney.
Because of that, it seems people have unrealistic expectations for everything they do--too high and too low, depending on how the viewer regards them. This is how a movie like Pocahontas can be fiercely attacked for being "insufferably politically correct" and "stereotyping Native Americans" at the same time. Never mind that Dances with Wolves was starkly black-and-white in its presentation of "Indians good, White people bad" in a way which Pocahontas avoided--Kevin Costner can do what Disney can't. (At least until "Waterworld," but we digress.) As for being an unrealistic portrayal of Indians, as Mel Gibson put it, "it has a talking tree in it, for God's sake."
With Celebration, I think people have had similar sets of unrealistic expectations. If the town had been set up by another major company with comparable resources--AT&T, say--it probably wouldn't have been subject to the same firestorm of criticism, and on the flip side, it probably wouldn't have attracted people expecting it to be nearly utopian. Pocahontas is just a movie, and Celebration is just a planned development.
Personally, I wouldn't mind living in Celebration's apartments. Sure, you'd probably have to buy some things from stores along U.S. 192, a couple miles away. So what? Unless you're working in its modest office park (or in retail), you're commuting to work anyway. And I like the idea of living near a pretty central park area within walking distance of a grocery shop, bakery, bookstore, coffee shop, movie theatre and four restaurants.
It's not Utopia, and it's not something that couldn't be duplicated elsewhere, and maybe duplicated better. But despite that, and despite the inevitable problems with a community like it (it's not the first "New Urban" town but it may be the first on its scale, and certainly the first with the level of media attention it gets), Celebration does work, and if it inspires other places to try similar development, I think it'll only be for the better.
Sigh.
I don't think it's bad to have a "features columnist" for Slashdot, and one who writes about the intersection between society and technology is a good choice. People of the True Engineering Mindset often ignore that intersection, and that can lead to the "all problems will be solved by technology" mindset which ignores the fact that the problems technology creates often aren't technological problems.
But--okay. Maybe it's just me. What I've seen since Jon was writing for Slashdot, though, has roughly followed this path:
I don't like to accuse Jon of being an opportunist, but it's difficult not to start drawing that conclusion. Sounding shrill alarms about movie theatres actually stopping people who are under 17 from attending R-rated movies is honestly a little dubious; you may not like the rating system, but it's not exactly new. And it's not even a government agency--the power of the MPAA is a creation of market forces. (Nobody forces studios to have movies rated, and nobody forces theatres that play unrated movies to police the age of attendees. You just won't make any money if you're limited to showing at the few theatres that play unrated movies.)
But now we're going beyond that, and sounding the alarm about the new great force for fascism in the country: pediatricians. Yes. You thought Dr. Spock was well-intentioned, but no, he was Mussolini with a lollipop!
Come on. Recommending that parents not put TVs in the bedrooms of their preteens and not show any children under the age of two television shows is an assault on free speech?
Really?
Jon, I understand the desire to be popular, but couldn't we get back to the pithy insights about technology sometime?
Actually, the specs for the G3 processor are not the problem. BeOS will run on Macintoshes with G3s in them--the compatibility problems lie in the surrounding hardware.
To make an operating system run on Mac hardware, you need specific design documents from Apple. (There is a name for them, but I don't remember it right now.) These documents are not publicly available, and concern the deisgn of the logic board, custom logic chips, address spaces, and so on.
In the pre-Jobs days of clones, these reference documents were available to a lot of Apple partners. After Jobs' return, it was decided these documents should only be made available to MacOS licensees, so that they can build MacOS-compatible hardware. This cut Be off from the information (and, indeed, there are no more MacOS licensees anyway).
Be has requested this information repeatedly, going up through various levels up until Gassee emailed Jobs personally and got no answer. This was around the middle of 1998, as I recall, and they haven't made any effort since then.
So does Be really want to be helped in this? I don't know. There's no love lost between Gassee and Jobs. Be has stated publicly that they could make BeOS work on Apple's G3 hardware without these documents--but they're not going to, unless Apple is willing to support them directly.
The flaw with this argument is not technical. It's that the typical business user does need Windows applications, if for no other reason than that they've been using them since the early '90s and you won't find any other platform that runs them. I'm not just talking about the ability to open Word documents. I'm talking about the service order entry system your company has been using for the last five years built around Microsoft Access. The spreadsheets lurking in every third or forth director's hard drive that contain Visual Basic for Applications programs. The weirdass vertical market application that looks like a star of "It Escaped From The Mainframe" that all your field service personnel depend on.
If you were building a business from the ground up you would be better off with non-Windows solutions, although you should be prepared to deal with the Word, Excel and occasionally PowerPoint documents that businesses you work with will send you.
Personally, if I were doing that business from the ground up, I wouldn't give everyone Linux, because the desktop application support still isn't quite there. (Yes, I spent a lot of time playing with StarOffice, and it ultimately sold me on Gobe Productive for BeOS. StarOffice is pretty and powerful, but my God, it's like a hog in molasses!) Servers, sure. Give the people who know what they're doing with them Linux boxes. Give the people who want to run office applications iMacs.
Getting xfstt up and running cut down on much ranting on my part about how pug-ugly X's font displays are. It's still not as nice as BeOS's display--no anti-aliased fonts, and some applications are a touch on the clueless side when it comes to using TrueType (Netscape being a notable example). Even so, it made a world of difference. If you're using Red Hat 6.0, I understand it comes with xfstt installed, and even a logical arrangement to share fonts between applications that logically should be sharing them, like X and Ghostscript. Somebody else could correct me if I'm wrong on that. RH51 didn't, and getting it running was a bit wonky, although not too bad if you pay close attention to the instructions. (I thought I had, but we won't go there.) When I switched to Debian 2.1, I discovered that when I installed the xfstt package it automatically configured itself and came up the next time I started things. The most work I had to do was to look in /usr/doc/xfstt to find what directory I needed to stuff the font files in. Somebody tell me why the Debian distribution is supposed to be the difficult-to-use one again?
Uh, perhaps the affluence there has affected your vision. The only theatrical animated releases Disney has made under their name that haven't been from one of the two Disney Feature Animation studios have been "Duck Tales: Legend of the Lost Lamp" and "A Goofy Movie," which were done by their television division. There may indeed be a Disney TV Animation Canada, but they would be one of several contractors working on various TV series and direct-to-video releases.
Yes, some movie theatres are actually making an effort to enforce the ratings now. Yes, that'll probably piss people off. Yes, ramping up the enforcement because of skitterishness over school violence is silly. But none of those things present a convincing argument for abolishing the MPAA ratings system, and they're not even particularly convincing of the apparent argument that children should be let into R-rated movies. Sorry, I just don't think the right to free speech is endangered by telling a 13-year-old he can't go see someone masturbate into a pastry.
As for the argument that Mr. Winter's DVD home theatre with the GNOME interface is the harbinger of doom for the movie industry, sorry again, Jon. Last weekend was the biggest box office weekend in history in America. People are not failing to go to the movies, and it seems that those teenagers being turned away from The Blair Witch Project aren't significant enough to affect the bottom line.
I can't address the argument that liking "South Park" is a form of asserting one's individuality without collapsing into giggle fits, so I won't try.
...people are absolutely correct when they say installs have gotten very easy. In fact, they've been pretty easy; most of the changes in installation programs I've seen over the last couple years have been more cosmetic than functional.
The usability issue is configuration after the install. It's things like being able to use a control panel to change your resolution, color depth and refresh rate to anything your card is capable of on the fly and have it stick (sorry, XFree86's hot keys just don't cut it in that regard). Like being able to install a new font by dropping it in a directory and having all the programs on your system be able to use it for both screen and print.
This extends to applications, too. KDE and Gnome are making strides in the right direction, but people shouldn't underestimate how nice it is for all but the most diehard CLI-or-death types to be able to do something (seemingly) simple like select font and font size for applications while they're running with a few mouse clicks, or creating symbolic links with drag-and-drop. Linux developers should not disdain the UIs of MacOS and BeOS, they should start learning from them.
Of course he's trolling, because there's nothing more fun for a columnist than getting a bunch of people frothing at the mouth. It's probably one of punditry's best perks to watch people go out and spew righteous flame with relatively little clue.
Bob Metcalfe has been the editor of Infoworld for quite some time now, including the time when Infoworld was the only corporate computer magazine taking Linux seriously. They reported on Novell's Corsair project--you know, the one that became Caldera--on their front page when other Microsoft-happy journals were ignoring it or, in PC Week's case, subtly slamming Infoworld for reporting on such nonsense.
Look, follks, every time someone casts doubt on Linux winning the world doesn't mean they're in the pay of Microsoft. It doesn't necessarily even mean they don't like Linux. Metcalfe likes to play the devil's advocate--he's been doing it for years. (Unlike folks like John Dvorak, Metcalfe also does admit when he's wrong, even going so far as to eat one of his own columns when he made some dramatic prediction that didn't come true.) But the record at Infoworld under his editorship doesn't exactly suggest an anti-Linux atmosphere, and it doesn't really suggest a particularly pro-Microsoft one. Remember, this is the publication that named Linux their top network operating system last year--and one that for several years kept naming OS/2 the best business operating system over Windows. (It also won for six years straight, I believe, in the readers' choice awards, suggesting that most of Infoworld's audience isn't comprised of Microsoft lackey's any more than the staff is.)
And, sure, Metcalfe cares about attracting attention. That doesn't mean he doesn't care about being right. If you're a pundit, you call 'em as you see 'em. You may see 'em wrong. Does Metcalfe see it wrong this time? The danger of the attention storm that Microsoft will undoubtedly try to create around Windows 2000 should not be underestimated in its capacity to derail the attention Linux is getting now from corporate users, and that's the mindset ol' Bob is writing about when he talks about the potential of Linux "failing"--not that Linux will blow up and dry away, but that Microsoft will be successful in putting up a brick wall for it to run headlong into.
Look: you can either puff up at the "trolling" and squawk a lot, or you can look for the points Metcalfe has that maybe should be kept in mind as Linux moves forward.
While I found Katz's "nightmare for Libertarians" comment to be typically uninformed, this "the world doesn't owe anything" huff-and-puffery is also a typical net-libertarian response. Why is it anytime someone makes any suggestion like "monopolies are bad," much less something truly radical like "maybe the way to reform government would be to make it more efficient at helping people rather than reducing its role to that of enforcer," this old chestnut comes out? Who said the world owed anyone anything? News flash: us bleeding hearts know the world ain't fair. We just have the quaint notion that working to make it a little fairer isn't a bad thing.
The American economy over the 20th century has been an absolute marvel, and Libertarians need to open their eyes and notice that that economy has been mixed capitalism, not pure capitalism. The IMF has been forcing pure capitalism down the throats of countries they've been trying to help over the last decade or two, abandoning the principles of the Breton Woods compact, and by any honest accounting it's been an absolute disaster. We gave them our theory, not our practice, and it's dismaying to see presumably well-intentioned people push those theories so hard here.
The market rewards mind share more than it does good products, and investors only reward high profits and continued growth--how that profit and growth is obtained is irrelevant to the reward. But if it's obtained through means other than providing the best possible product at the best possible price, very strange effects start happening, Microsoft being one of the comparatively innocuous ones. (HMOs are one of the less innocuous ones.)
Incidentally, I don't despise Bill Gates for his wealth. (This is another bugaboo with libertarians, I've noticed--mention "growing income gap" and they start frothing at the mouth, as if merely noting its existence is a call to sell Texaco's assets to the homeless.) That doesn't mean I can't despise him for flogging useless garbage, and you know what? You can too!
So Leonard praises Slashdot in a manner which is not only effusive but accurate, and you call him a BIG FAT PRICK. Clever. Today's thought for the day, children: does this kind of logic discourage "anti-Linux FUD" by the press, or does it encourage it, particularly of the sort criticizing the Linux userbase for being hypersensitive, highly unsocialized and apparently illiterate?
Of course, when Amelio looks back I bet he really wishes he'd paid the $400M for Be to not get Steve Jobs. :-)
If the underlying operating system is as irrelevant as they seem to imply that it is, I'm not sure we should be looking for significant kernel modifications any time soon. I think they went with Linux as opposed to QNX or BeOS for three reasons, all of which boil down to "free" as in "free beer":
I'm interested in seeing what happens, although obviously I'm a bit jaundiced (well-funded or not, this is an awful lot for a company that was less than a dozen people at the start of this year to deliver on this fast--a lot of this strategy apparently doesn't go back much further than Jim Collas' involvement).
But you're right: if nothing else, we'll end up with another desktop manager with unique features that aren't compatible with Gnome and KDE. To hell with those "closed" desktop environments--why should we make it so application programmers have any reasonable expectation of what services the GUI they're running on can deliver? :-)
Granted, unlike iToaster the new Amiga (if it happens) will presumably actually use Linux instead of just sticking its name in the press release, but one can't help but suspect it's being done for similar reasons. Linux is the buzzword to be using now if you're challenging Microsoft.
And, if you need to get a new operating system out real fast, what better approach can you take than to use someone else's operating system... that you can use for free?
Personally, I'm skeptical as to whether or not the "great new hardware" for the next generation Amiga is still in the works, too. What about all the developers that are supposedly waiting in the wings to support the system? How many of them were developing for QNX? How many of them were developing at all? Will all of them be willing to develop for Linux? Will they be able to develop for the Amiga "OE" without developing for Linux itself?
That sounds like an odd question, but it isn't--there's no reason to believe this "operating environment" will also be GPLed. Of course, if it isn't, it'll have to run in a "closed box" on top of Linux. Is this going to be able to compete in real-time performance with BeOS? Or for that matter, QNX?
Or for that matter, the original AmigaDOS?
While you've raised some pretty good points, I think that BeOS R4.5 does have an adequate level of hardware support for common configurations. Not a good level, but an adequate level.
Would it be better if it could do hardware-accelerated OpenGL on TNT-based cards? Sure, and I suspect there'll be drivers that support that for non-Voodoo cards before R5.
But even if you just subscribe to the "media" schtick, the lack of 3D hardware acceleration isn't that important if your definition of media is, say, audio or video editing. With those the framework is already there, and in practice, not just theory. (Visit Cirque de Soleil in Orlando, or the Broadway production of "Ragtime," or the video-editing demonstration at the ZEUM hands-on museum in LA.) What about web design work? Or 2D "cel-style" animation with Lost Marble's Moho, a program I haven't seen the likes of on any other platform yet?
Sure, there are holes there, but it's not like the initial Intel release anymore. We're talking potholes now, not gaping sinkholes. And, I agree with the commenter who disputed that remote administration and multiuser capability are make-or-break features for a large percentage of the audience. (They're certainly not important to me--I'm running Linux now, but I'm just one user, after all. And maybe your office is better than mine, but at work I really don't want my company's IS department trying to adminster my PC for me. NT is annoying enough without SMS futzing with it, thank you very much!)
While I've seen this argument before, I'm not sure I buy it. Office got mindshare at first because it was the only "pro-level" game in town for Windows--Word and Excel on Windows predate WordPerfect and 1-2-3 on Windows by at least two years, and the first Windows versions of both of those were terrible. By the time anyone mounted serious competition, both of them had over 50% of their respective markets on Windows, and Microsoft started using aggressive bundling tactics to make sure that many users had Office installed with their computers when they got them.
The "only idiots want to use too-friendly software" argument belies the fact that bloated or not, Word and Excel are very powerful programs. All of the obnoxious automatic features can be turned off quite easily, and there are fairly powerful features--outlining, mail merge, revision changes, groupware routing--that are as close to intuitive as you can get in Word. And I haven't seen a better spreadsheet than Excel yet. It does useful things that no other program in its class does; while there are better programs just for statistical analysis and just for flatfile databases, for example, there's no program I've found that does both of them together as well and as easily. I write statistics-gathering tools for our company's frame relay networks (information about which is scattered in SQL and Access databases), and Excel is simply better tool than Access for such things. (Its database handling is nearly as powerful, and its reporting tools are better.)
Scott Hacker, for example, provided no data on the percentage of flames versus constructive criticisms. It may be that the negative comments were not at all representative of the whole.
While you're right, this is also somewhat misleading; he said the responses he received ran the gamut from complete agreement to obscene flames, and never implied that most of what he received was flames rather than constructive criticism. Beyond those introductory paragraphs, in fact, he didn't address flames at all, but instead did respond to the constructive criticisms, including acknowledgement of points he hadn't considered or wasn't clear on. Obviously people can still disagree with his conclusions, but just as obviously he's not just ignoring well-considered responses.
It should also be noted that consistently misspelling Scot's first name as "Scott" probably doesn't help things. :-)
I'd been looking for a good message to point BeOS users toward on how not to advocate for our favorite operating system. Thanks for this one--it's just perfect!
In other words, if you look through, the first format to hit critical mass with the product people buy it to use seems to win, unless a very compelling reason comes along to change.
This is the "network effect," and is probably more responsible than anything else both for Windows' de facto OS monopoly on the PC architecture and for why PC architecture greatly outsells other microcomputer designs. (I say "de facto" monopology because there have always been alternatives to DOS and Windows, yet asserting that Compaq had the choice in 1995 to install OS-9000, QNX or Concurrent CP/M-86 is akin to asserting General Motors has the choice to stop making cars and trucks that can run on interstate highways.)
While I don't think we're going to see the end of the network effect for Windows any time soon, I think we are seeing it start for Linux, and possibly for BeOS. None of these are likely to get bigger than MacOS, but that's had enough of a network effect to carry it through years of corporate mismanagement.
Windows' multi-hundred-million user base tends to skew people's ideas of what a sustainable market size is, I think; a market of only one million could sustain small companies, and it only takes a market of a few hundred thousand to sustain "cottage software" houses--something I think we'll start seeing more of in the future again. (What's old is new....)
...I'd like to see more stories about Hurd and Berlin, too. I know nothing about Berlin, in fact.
Whoops, that wasn't a flame. Sorry. I'll do better next time. :)
Actually, Scot really doesn't seem to dislike Linux; he's written in the past about how BeOS and Linux have their own strengths and weaknesses. I think what happens with articles like this could be called a "defensiveness feedback loop"; I won't say there aren't BeOS zealots out there (I've seen some fairly wacky claims from supporters), but there are more Linux zealots (if for no other reason than there are more Linux users). And, Linux is unique in being blessed or cursed with a small subset of advocates who give "OS religion" a more literal meaning than we've ever seen in the past--the idea that open source is the only way to go because closed source is unethical. You see hints of that in this thread in the argument that Be can't be said to support open source if they only provide an egcs-based IDE, a GNU toolchain and the source to most of their drivers and applications. If it's not "total" support (i.e., all source opened), so this logic goes, it's no support at all. While this isn't an indefensible viewpoint, it approaches a religious argument (you're completely with us, or you're against us, and to be shunned).
As a staunch BeOS supporter, I'd say that Linux is more mature. Hey, you should call 'em as you see 'em. I think people tend to forget that despite the fact that Be, Inc. has been around since 1990, BeOS hasn't. DR1 came out in late 1994, and the first release that didn't require a BeBox was DR8.2 in January 1997.
There are things I like more about BeOS than about other operating systems, but I still can't do everything that I want in it. Of course, I still can't do everything that I want under Linux, either: there's no equivalent to the word processor I use (Nota Bene), and while the DOS version does run under dosemu, it doesn't run perfectly in console mode and runs even less perfectly under xdos. And there are tradeoffs even between the two operating systems. (I can't get TinyFugue to run correctly under BeOS, and I can't get a Linux version of e-Picture or Pe. BeOS sucks moose droppings if I want to use it as a server, but its GUI--whether or not people like the color choices--just isn't matched by Gnome or KDE. [For the morbidly curious, I duck that debate in practice and just use Window Maker.])
Well, I wouldn't, but I was giving what seemed to be a good answer to someone asking about the possibility of that.
What would be the point?
What would be the point of putting Linux on Furby? It's a hacker thing. (I know some are dreaming of $200 network gateways, but with the iToaster design that's not real likely to happen.)
The box was made to be used with BeOS; Linux isn't the end all and be all.
I think people need to stop thinking of this as a "$200 BeOS box." It's kind of like referring to the Philips "MyWeb" Internet TV device as a "QNX box." It's not that it's inaccurate, but it's somewhat misleading. Neither iToaster or MyWeb are meant to be PCs, even cheap ones. It's a different market.
BeOS is $70 from Be directly, although you can get it online at some other places more cheaply (PC Mall, I think, sells it for $53). It should be in retail boxes next month, I think, too, so stores may have it lower. It's not dirt cheap, but it's not horrid, either. (Red Hat 6.0 is more expensive if you get the "official" release.)
From what I've read about the iToaster, getting BeOS off it and putting Linux out it would be a challenge--it has no CD-ROM and only dial-up networking, and is about as expandable as a styrofoam brick. It's more "PC-like" than a real PC. It might be interesting to try to get Linux on it for someone up to the challenge, but success probably wouldn't bring you much useful--it'd have to be done for the "because it is there" feeling.