Actually, no: you've missed some central things about GAs. In fact, given a sufficiently diverse 'gene pool', genetic algorithms work even _without_ mutation. Mutation is a sideshow: recombination ('crossover') is where the real action is. If 'organisms' wind up with roughly comparable ways of doing things, crossover has a much greater chance of producing a useful optimization than mutation does. Mutation is only significant if you're starting from a monoculture in which, at first, crossover would do nothing.
Okay, show of hands, who has heard of a 'pointer to a pointer' and why you might want to have one? Concept could be used in roll-your-own memory management like you have to do in C and C++:)
Processor cost consideration
on
V2 OS
·
· Score: 2
One thing to bear in mind is that old slower processors can actually cost more than the new faster ones under certain conditions. I learned this when Powermacs were starting to take over and I thought a ultra-cheap onepiece Mac would be neat. Turned out 68040s actually cost more at the time than PPC chips, because the latter was where the volume was. Maybe you might think about embedded PPC or MIPS or something? Is there genuinely a market for, and producers of, 386 chips now?
I mean, go ahead, but that's very much like saying "Anyone with a 486 and Red Hat can write the next Q3test and do serious work on the kernel". To some extent it's the equipment...
(here's a tip just for slashdotters: you can take a large sheet of plexiglass or better yet a big sheet of steel suspended from springs (plate reverb) and cut a little hole in the middle just big enough to expose the membrane of a cheap radio shack electret condenser mic (to really win you need to trim away the aluminum shielding the membrane and possibly cover the exposed bit with a little spot of windscreen) to produce a very large condenser PZM mic that will do vocals more clearly than junkware sm58s due to the lack of comb filtering effects- also, the resulting apparent distance of the sound is a good foot or two closer than the actual distance)
Now. Seemed to be a lot going on in that hack, wasn't there? Why would I, a soundengineer hacker with his own music, be giving away secrets like that?
Because that WASN'T enough. It'd help. But if you don't know how to make cables that will get as much of your signal to the A/D converters as possible... if you make the mistake of doing a lot of destructive digital changes on your data (it can be better to re-digitize a track than to normalize it over a simple gain change)... if you're using a crap digital mixer working in only 16 bits and not dithering properly, it's not going to be enough! You'll need help.
This is normal for any skilled profession, craft, or art. You wouldn't go out and get some random guy off the street for your drummer- it's no different for the sound engineer. The difference is, the audio geeks and equipment tweekers and snotty audiophile types now have a market value- and there are just as many of them out there as there are unsigned musicians.
Not everybody gets to be George Piros or Wilma Cozart or even Bob Ludwig or (shudder) Bob Clearmountain (if you don't know those names you're not a sound engineer geek. Show of hands? I bet some slashdotters know _all_ of them and why I shuddered at Clearmountain;) ) but it's just the same as linux hacking- there are countless things to learn, it's a tremendously deep field, and you _can_ put together a 'garage' operation that competes with the big boys just the same as Linux competes with Windows NT.
It might involve a lot of geeky work. My mixing board had over 100$ of capacitors alone put into it. Not wizzy 'audiophile' caps of matching values- I increased values radically, now my board will put out bass on the order of 2 hz >:) there are most definitely audio hacks that can be done with equipment, it's a whole subculture.
I guess the long and the short of it is, at home with your PC or Mac you can _top_ the results of your average industry studio- if you're really willing to spend some years being a mad scientist audio hacker, or know someone who is. I always figure, I've been doing it for nearly 20 years, I can afford to give away everything and I'll still top ya in execution;) so, here's a list of things to do/use/remember...
Get serious monitoring. Learn how to place speakers in the right places from audiophiles. Get classical music or classic rock sounding great over the system- the stuff that requires the system to reproduce a soundspace (not synthesise one tho!)
If you're using bass reflex speakers, stuff a sock in the ports. To mix very deep bass you've gotta be able to hear it. A sock will also provide resistive damping. You can enjoy port thunder later, now you need to hear what's actually happening.
If you're using speaker wire try cutting the ends off AC extension cords (heavy duty indoor) and using those. If you're using those, try separating them all the way down into two individual wires, getting rid of capacitance effects. Crude but effective in getting more control of your highs.
Litz wire is better than stranded or solidcore wire at transmitting analog audio in such a way that it digitizes nicely. You can make amazing cables from using all the wires in outdoor phone cable (four little wires rattling around loose in a big plastic sheath) because they'll work as sorta-litzwire and any shielding around the wire has a serious air gap for zilch capacitance over even long runs.
Running cheap digital multitrack (like an old ADAT) into a reasonably decent analog mixer and then digitizing the result 'dithers' better than crap digital mixing software. (this is actually my next audio move- ADAT used in a hopped-up analog-uber-alles studio)
Pitch shifting is destructive. Digital EQ is destructive. Normalizing is destructive. As with something like JPEG or MP3 these are cumulative. Don't ever edit things around like mad without backup- keep master digital copies in case you want to do a neatly executed series of digital edits on a clean copy. Screw up on this one, and your tracks will have all the life and interest of Radio Shack Casiotone demo tunes. Don't get track rot.
Typically you can only afford to have a few sounds be 'big' or glossy or fancy- different genres approach this in different ways. Some great MODs have all the sounds big, but there's only about 2 or 3 tracks going at any one time! Conversely, some great rock mixes have, for instance, really big guitar sounds against a simple direct, dense bass tone without a lot of detail, and a very 'dry' drum tone. Even John Bonham's Led Zeppelin drum tone tended to be pretty 'dry': it just tended to sustain. When it was really wet and dense, the _guitar_ tended to be dry.
Drum tones only sound right in context. You'll only hear what a bassdrum is doing when coupled with a bass guitar- the weight of it can be way more than you expected combined with the bass's transient attack. A snare can sound clear but boring until it's in context and sounds great- if you add stuff to make the unaccompanied snare sound hot and exciting, when it's in context you might lose most of the actual impact because everything gets muddy and confused. Think of the backbeat as a composite of drums and instruments- as if the guitar or whatever is _part_ of the snaredrum. Mix it as such.
Never mix over headphones. Any headphones. It's a totally different presentation from speakers- your body needs to feel the sounds (even at a low volume, subliminally). Headphones are for tracking not for mixing over.
Build monster speakers if you want to make club or house or rock music. You need to monitor over something comparable to what you'll be played on, only more accurate. I run towers with 12" 10" 8" and 6.5" woofers for (infinite baffle) bassbins. No one resonance dominates and the low end is understandably huge when required. Makes it easier to mix really serious bass content, you can hear exactly what it's doing. I also make my own tweeter elements. Audio geekiness is fun:) also, using them as computer speakers makes games and such more fun. Big explosions, and I've occasionally encountered stuff like this one alpha-quality game in which the guy had made the sounds carry _major_ subsonic rumble for the explosions. Very neat.
Geek out on it. It doesn't take that much money, you just end up very well known at Radio Shack and very familiar with what you can get from MCM, Mouser catalogs et al. It's kind of like Linux in a way- you can DIY, and even beat hell out of the industry's approach in certain areas.
I mean, read it. You retain copyright. Did you know how many mainstream artists lose copyright to their own songs to the record companies? The rights granted are NON-exclusive: try that with the mainstream industry. "Uh, Atlantic, mind if I also sell CDs through Sony?;P" You get to terminate at any time.
This is actually so much better than the record company deals it's not even funny. (It's not funny, actually. It's tragic.) Sure, Joe Schmoe isn't guaranteed income as a Rock And Roll Star- when has he ever been? Sure, you have to do ALL your own promotion: guess what, you would in the mainstream industry as well. Record companies _don't_ promote new signed bands. They promote the Spice Girls. In the case of something like Hootie or whatever, you're talking about a local band that had a _killer_ marketing machine and network and did a lot of work without asking the record company to do it for them (which ain't happening).
Lastly, nothing is stopping you from setting up shop as an indie label. Press your own CDs- it'll be about a dollar a CD including labels and sometimes also throwing in cassettes, you supply artwork, you pay shipping and fill orders yourself and keep _all_ the profit after taxes. Nothing in the mp3.com agreement is forbidding you from doing all the work, not to mention you can make higher quality CDs from stuff that hasn't been mp3ed first.
The one thing mp3.com takes is this: everything you put up, they can spread around in any way they choose. Well, duh- that's the point! If they were not going to let people download it, splash your little graphics around, drop your name in their email mailings, do a web-radio station and play your songs all they want, what would you be there for, why would they be giving you lots of HD space? You can back out (unlike the regular industry), it's nonexclusive (unlike the regular industry), it's clear that you aren't losing copyright to your own work... (unlike the regular industry... are we seeing a pattern here? Ever wonder why certain artists 'seem' to be really into selling their songs to crappy compilation album producers?)
Thanks for the link. I made a copy on my HD of the agreement to go over at my leisure- but I gotta tell you, I'm impressed with it. I don't see much of a gripe, compared to existing industry practices. They were supposed to provide MP3-sized HD space to zillions of unsigned musicians for nothing maybe?;)
I'm picturing a checkbox labelled "Allow untrusted users to plant Trojan horse programs":) of course, it defaults to off except for when you set Office to 'Active Content':)
*turns on Royal DS2260 organizer* (bip!) *hits backlight* (bip!) *turns it back off, closes cover* (bip! clik)
This was $30. I once wanted an old Powerbook- just something that could take text notes and be carried around. B/W screen, some tiny amount of ram- never could get one, too expensive.
Now I have a toy with 256k of ram (that's an awful lot of little memos!). It has a backlight, which is more than the old Powerbooks had. It's got chiclet keys that go (bip!). There's a fourway arrow key thing that is occasionally relevant. The text editing is rather like vi or something- hell, the whole thing is extremely modal, and yet it's such a little thing that it doesn't matter- I don't need a trackball or color for this. So where a few years ago I could only wish for an old Powerbook (which wouldn't fit in my pocket without _serious_, er, percussive redesign;) ), now I can actually have a little toy 'laptop' that goes (bip!) and takes notes and shows what time it is in Hong Kong or Berlin.
And this is totally cool- and I can't wait until I get to have one for $50 that's the same size, the same lcd screen with blue backlighting- but has 4M, bash, and vi;)
That would just feel so good tucked into your pocket. Imagine. Hip-nix;) (poc-nix?)
I was awakened this morning by a telephone marketing call, and talked for over a half an hour.
Why is this? Because it was Pixels: the little company that makes Pixels3d, a modeller/renderer etc for the Mac. I already knew them- in fact I keep a copy of Pixels3D 2.1.4 around because they are one of the many Mac vendors who have taken to releasing their last-year's model at no cost, and I grabbed it. Being allowed to fully use something like that left me with a good feeling about them, and they are the antithesis of a big vague corporation- it's a bunch of computer geeks running a company, and their product kicks butt (except that I hate the Lightwave-like interface:) )
I wasn't able to take 'em up on their promo, won't be buying anything today- but, you know, I am _very_ used to shutting off telemarketers. I don't give them three seconds. I interrupt, I firmly say I'm not interested and then hang right up. Yet in this case these people were able to keep my attention and get my sympathies- why?
Partly because they were ready to put some serious effort towards getting me what I wanted. I learned that the scripting language was like Renderman shaders. I learned people write plugins in REALbasic- hey, I have that! I even ended up talking to the main programmer for about ten minutes on how many semitransparent layers you could stack to simulate volumetric clouds (a POV trick I've been playing with), and he had all the techie details. It was so deeply about what _I_ wanted to know, rather than about what they wanted to sell.
I hear people saying IBM is also taking this approach. Well, good for them! The predatory thing only works when you have a lock on people. Pixels doesn't- they do Mac software, and compete with everything from Metacreations (Poser, Bryce, Infini-D) to Lightwave itself and who knows what else? Seeing this glimpse of how they work with their customers gave me a bit of insight into why they're still around at all. Did you know that you can go to their community page and they will put up _pictures_ of their users? (one wonders if it's pictures of _most_ of their users!;) ) Looking down the row of faces, next to banner links to the respective websites, was a lesson in PR.
It doesn't always stay- that language I use, REALbasic, is very neat but the marketing people have taken to trying to get people to link to the RB site by offering space on the CD in exchange for posting 'Made With Realbasic' logos on things. That's a 'what can we get' approach, not a 'what do you want?' approach. I don't mind it but I'm not doing it. Pixels is smarter- or wiser. As, apparently, is IBM...
Cinema: A Whole New Chapter In The Revolutionary World
Letters From Virtual Impeachment
The Web: A Whole New Chapter In Online Buffy
Giving Thanks For Virtual Cinema
The Power Of The Techno Stock Market And Interactive Dying Babies
Murder: A Whole New Chapter In Hidden Being Different
Is The Stock Market About To Transform Slashdot?
Can Hidden Anonymous Cowards Stop Post-Columbine Minimum Wage?
The scary part is, this _reeks_ of Markov Chaining. I suspect these are made up of actual genuine interactive dying baby Katz headlines, shredded and rearranged.
One envisions doing this to the whole Katz on Slashdot collection, thus getting not only terrific headlines, but entire articles. For anybody interested in doing this, Katz is a _very_ good subject for travesty as he is disjointed and covers the same subjects over and over. To markov chain this stuff it's best to work at the word level- pairs of words end up chaining very abstractedly, and word triples can sometimes spit out overlarge chunks of unaltered text (though I suspect with Katz the triples would work perfectly).
I'm a musician. Mind explaining to me why I shouldn't have control of the means of production? I'm not talking about distribution, I'm not suggesting I should get free promotion: I'm just saying that the current approach seems to be reliance on say 40 special encryption keys, and you have to basically be a multinational to get one.
I have no problem with all this if equvalent _unencrypted_ data can be played back in consumer decks- if I can burn a DVD and distribute it using my own limited resources. I am getting the feeling that this is a 'Why would you want to do that?' situation to the big conglomerates- and that it's not necessarily going to be happening.
I really don't _care_ if it's easy or hard to copy off major-label releases. I'm not even interested. I'm more interested in what all this means to me- and if we're looking at a world where you CAN'T MAKE the 'records' of the future, scant years after it became seriously easy to do just that with the once-highly-techie audio CDs, well, I have a problem with that.
Again: give me ability to record unencrypted and release product, or give me ability to use some default encryption if you _must_ have it, but if the ability to produce these disks will be limited to the industry only, with 'indies' finally legislated out of existence by technology itself, then that is a very ugly world for an artist. It's bad enough being a musician without such crap. This is the thin end of the wedge- it _won't_ stop there. The logical direction this is going is to stomp out support for old formats that can be easily ripped from (and, coicidentally, easily released on), and then for the indie musician, not even cassette tapes or CDs would remain, and it'd be "Oh, please, mister consumer man, won't you buy a worn-out old CD player to mount under your _new_ wonderful encrypted CD player so that you can listen to _my_ music too? It won't take up too much room.";P
I can't even begin to explain how bad that would suck, and I'm already boycotting the industry as hard as I can. I think us indie musician types need help. The odds were already heavily against us, but now it's just getting ridiculous, and there's no reason to expect it'll level off by itself. Hopefully there will always be some sort of mp3 underground, as that might end up the only option for a lot of us. We could be watching all of the following:
the final takeover of the entertainment industries by multinational conglomerates
the greatest level of control and highest barriers to entry of any industry ever- comparable to if you had to run all floppy and CD-Rom software distributions through Microsoft because they built encryption into the drives and became the gatekeepers
possibly the most receptive audience for stuff like underground, 'indie' music not coming from the machine. Normally people don't care to pay attention to acts with low budgets. If the industry stuff gets plastic enough, and the barriers to crossover get high enough, it almost by default produces a market for underground stuff- acts that normally would have found a niche within the machine, but can't anymore because things are so well controlled.
really severe skepticism towards the entertainment industry fantasies- some people these days still believe you can play music and have that facilitated by the industry, and make as much as 99 cents on each CD, rather than pumping large amounts of money into the fantasy to get to see your name in lights. This belief will fade, to be replaced by the anger of betrayal.
probably a lot of absolutely phenomenal music by indie artists who will be blacklisted by the industry for practices such as releasing mp3s. Such artists might be barred from playing certain clubs or gigs for this (if the gig is in the pocket of the industry, that is).
Maybe ten percent of the absolutely phenomenal indie artists will actually make a lot of money at it. These might not be the greatest artists, but the one common factor will be this: they will be the artists who run their bands/record labels/whatever like a serious business. They'll plow most of their profits back into their music (as Frank Zappa said, "Some rock stars take their money and snort it up their nose. I stick mine in my ear") and they'll put out product that's markedly better than the average industry offering, and they'll do well even when forced to handle all the promo and distribution themselves, typically with a heavy Internet emphasis, and a heavy emphasis on mp3 freebies.
The money is in merchandising and in releasing versions of the music that sound better than mp3s- as the industry pushes newer and newer technology, the back-end stuff such as amps and speakers will start to perform really well at good price points. Klipsch has a computer speaker setup that is flatly astonishing. Information will be out there helping people to maximize their 'stereos' (which will tend to merge with the computer). This gets back to the need for access to tech like the new audio DVDs- but honestly, if you know what you're doing and try hard you can get extremely impressive sound out of the humble CD, even, and the indie people will try all the harder if they're barred from putting out DVDs through lack of access to the encryption keys.
Nobody is likely to argue for indie musicians' access to such technology- it's all about the industry versus pirates. There _are_ no artists except those owned by major labels. Right?;P
Hey, I didn't once say I supported the WTO. In my opinion it is a freaking nightmare, and the most likely outcome is to solidify global power in corporate hands, substantially bolstering it until these corporations are directly comparable to governments.
That said, just where do you get off suggesting that _protest_ is going to change this? (I'm not even going to get into 'suggesting that looting Starbucks is going to change this')
I would suggest, in all seriousness, that protest, already meaningless except in a (dangerously open to misinterpretation) publicity sense when used against governments, is totally useless against corporations. Corporations can replace ALL the beef in your supermarket with genetically enhanced hormone pumped beef, by economic leverage. Corporations can make things like this happen and you can't complain to an elected representative as there isn't one, and you can't complain to the corporation even if they are sympathetic because they do not have, shall we say, lots of legal support for going against the interests of the stockholders- I question if the stockholders can even get together and say 'do X, which will cost an extra 2%' and have it override the corporation's obligation to go after that 2%.
So protest becomes irrelevant. I would suggest that you'd better learn to fight, instead. I don't know exactly how this would be done: I don't claim to have all the answers. Possibly IT fighting, cyber-attacking a corporation, would hurt it. Very likely continued physical attacks on the corporate executives would make it difficult for the corporation to conduct its business, though you cannot hurt the corporate entity itself, and anyhow corporations can and do hire bodyguards, chauffeurs who are taught anti-hijacking by Bob Bondurant etc. so a corporate exec is a _tough_ target, as tough as any government official. Finally, the corporation is a creature of law so it can be attacked by law- maybe. The trick is, they are increasingly calling all the shots in the legal sphere- we wouldn't want to have to overthrow the fscking _government_ just to get control of the legal system back. All in all, it's a very nasty puzzle.
I do acknowledge the dangers of the WTO, against which all these groups from Greenpeace to loggers to environmentalists are aligning. I think maybe I consider it more dangerous than you do... but I doubt it is remotely helpful to have sit-ins, or seize Seattle by a _mob_ (a militia, maybe, but a mob??) or to loot Starbucks.
Anybody who's read some history on the 60s knows it was a time of utter, total lusers on both sides. Some noticed this, and some didn't.
Remember the Mothers of Invention? Frank Zappa noticed- and produced a lot of music to try (and fail) to shake up a lot of people who, quote, "mindlessly accepted everything they were given, without questioning it". That was the hippies, to Zappa.
Woodstock was paid for entirely by John Roberts. Remember that name, or Joel Rosenman's? John Roberts financed the whole festival, having been conned to believe it was a moneymaking venture, originally put forth as a 'press party' for a recording studio to be built. Roberts pretty well had a breakdown while the festival was happening- left totally responsible, far in debt, being asked to sign checks for which there was no money anymore, Roberts and a few other people took the whole load of the Woodstock Festival upon themselves, a festival that was declared free after it had already eaten through Roberts' entire inheritance.
The whole hippie concept is a story of rip-off, stealing, lying, and destruction, painted to appear as virtue and freedom. Virtually nothing was accomplished- the end of the war in Vietnam, for instance, owes much more to the fact that eventually Middle America was sick of it and wanted it stopped, and to Nixon's bid for re-election.
Having hippie idealism on the Net is a bad thing. It's spelled out quite obviously: "Let's march on Website X and stomp it with DOS attacks!". No thought is given to other services that may be hosted on that computer, no interest is taken in the additional load produced by X many lusers running a Javascript program and monopolizing all the network links to the target. The idea of responsibility is seriously lacking here.
It's like a riot in cyberspace: riots were seen as civil disobedience in the 60s. My generation saw them more clearly: "Tomorrow you're homeless- tonight it's a gas" -The Dead Kennedys
Rioting is not freedom. Rioting is collectively throwing a fit. If you want civil disobedience, "smash the right windows" (Lee Felsenstein)- get smart, intrude, change their web page, don't just riot in the cyberstreet smashing everything. Stupidity is not insulation, it will not protect you.
If you want FREEDOM, then write fscking software! This is the most annoying aspect of all this. DOSing a site that happens to contain something you don't like is freedom? Write software, GPL it (or BSD license it depending on if you don't want to _enforce_ the availability of the code), put it out there. That does more for freedom than any twelve hippie web pages. If you are a blackhat at heart, learn how to pull off intrusions into whatever's out there, get good at the surgical strike, be smart enough to spare the environment you're in while punishing your enemies. That's harder, of course: a lot harder, in fact. But there's no excuse for the hippie approach. It's a disaster, a mess! Fight smart or go do bong hits, if you can't get a clue then get out of the way.
Otherwise you might well find that the GenXers (_my_ generation, thank you) have very much their own opinions on what activism is. You might find that they take a dim view of mindless destroying to prove some vague point. You might find some GenXer who's done his or her homework sneaking onto your precious target site and setting up some sort of viral attack from Javascript on the site itself- which itself attacks the site, but also whacks all the idiot DoSers in the bargain.
Hippies are the Commodore 64s of making change happen. It's time to move on. It's time to get _serious_. For example, the GNU GPL uses the law and copyright to attack what copyright is _normally_ used for, and Linux takes the GPL and proliferates it wildly- now there is a huge amount of Linux out there, and it's got the legal backing to fight attempts to subvert what it stands for. Now that's change. That's _significant_ and it matters and it's constructive but uncompromising.
Forget the hippie approach. Go with the Linux approach. Build something good and be ready to protect it.
I don't agree that it's that guaranteed. Perhaps what I'm saying is that I don't have the faith you have that this is the case- especially when we're talking about a community-used license going up against the lawyers of big corporations. You do realise that if (to use the most extreme example I can think of) somebody bribed the judge to make an exception and find the GPL 'not really a license', in other words defining it by usage rather than by the language, that would be the legal precedent? I really, really don't have the faith in the virtuous and incorruptible judicial system you do. I'm damn glad it works as advertised a lot of the time, but I'm simply not ready to assume that people can use the GPL differently than the wording of the license and not raise the question of later consequences. I think we will have to disagree there.
I'm seeing quite a bit of sentiment indicating that Linux users and GPL-using authors should ask Corel to apologize, to say 'oops', and then should sit back and trust them.
Well... of course somebody at Corel will be told to say 'oops'. That's his job. Its name is 'Public Relations'.
There's also people there whose job it is to figure out the next testing of the outside of the envelope. They are called 'lawyers'.
Forget that they have PR people for a minute, and also forget about their lawyers, and let's just look at their ACTIONS, okay? Their actions have consistently been causing problems. It doesn't matter that there's somebody to say 'oops'. That is a PR person, that's their job, it does not mean as much as people would like to believe.
I personally have a lot more respect for Bruce Perens for momentarily _refusing_ to listen to the PR flacks and looking at what Corel is repeatedly doing. I'm not sure if people realise that permitting such abuses of the GPL is very similar to letting people talk about xeroxing documents...
If you do that, the Xerox people _will_ speak to you and make sure you are aware that you are using their registered trademark as if it were a generic term, that other companies have lost their trademarks over just such weakening of the mark, and that Xerox (r) does not intend to weaken its mark in such a manner.
Well, the behavior by Corel obviously weakens the GPL- to the extent that it goes unchallenged. Anybody picking a fight with the GPL will be able to point at the Corel actions involving it, and the community responses to it. They will be able to truthfully say to a judge or jury, "This is not actually binding, because it is not obeyed or respected by even the people who use it. Look, here are 500 slashdot posts saying that Corel should not be penalized in any way for their alterations of the license, because Corel will benefit the community. Isn't it true that the GPL is a legal fiction that is not actually intended to be followed- as seen by these posts effectively waiving it in the court of public opinion... and therefore, shouldn't it be declared null and void?"
That's a perfectly legitimate approach for a lawyer to take, to invalidate the GPL. It's Corel above all who are making that argument possible, by repeatedly adding and altering conditions- and it's Linux users who are helping, by assenting to this alteration of the license, and making the argument that Corel should not be held responsible to the terms of the license.
I'm sorry, but I don't assent to the alterations of the license. I have software out there under the GPL, and I don't remember giving Corel permission to alter, challenge and weaken the license for their own convenience. I'm not concerned with how quickly a PR flack can be found and made to say 'oops'. These changes need to stop, or Bruce Perens' hasty reaction will be the only sensible course of action in the long run.
I _realize_ people want to be nice and not seem threatening or combative, but this is not fantasyland, or high school: this is the real world, there are consequences, and allowing a widely used agreement like the GPL to be defined differently in practice than it is in writing is bad. It makes an argument possible that the written version isn't really the applicable one- just as we need to look at Corel's _action_, so the courts will assuredly look at the Linux community's _action_ to ascertain how valid the GPL is when challenged. Do we, seriously, want to establish the precedent that you get to do whatever you want as long as you don't keep doing it for so long that it gets annoying? More relevantly, how many people advocating that Corel be let off easy have themselves written GPLed software- which of these people don't actually have anything to lose?
I think if attention is paid to the Linux community's wishes as to how closely the GPL must be followed, the attention needs to go to the people who have written software that uses it. I don't think it's at all helpful for people who aren't actually using the license on their own work to say 'Take it easy, maybe they don't have to obey this license that closely'. I think most people who do use the GPL on their own work would not wish legal precedent to be set that the license isn't really real, that in practice it's customarily altered and bent to fit situations. I certainly do not. From Bruce Perens' initial reaction, I don't think he wishes that conclusion to be drawn either.
It's all very well to be calm and rational and reasonable. However, in the event of someone calmly and rationally taking actions to basically weaken the GPL by setting a precedent to alter pre-existing licenses on the grounds of convenience, I would calmly and rationally have their head for it:) More accurately, since I _have_ read the GPL, do have a reasonable idea of what it requires, and have released software under it based on my desire for my licensing terms to be _just_ _like_ _that_, I will calmly and rationally suggest that Corel not only has to either recant quickly or lose their permission to redistribute GPLed software, I would also suggest that this 'worsening of the signal to noise ratio' is harmful. It is harmful because the more this is allowed, the more will happen. I would suggest (calmly and rationally) that it is very much a bad thing that Corel is permitted to blunder in this manner. Their mistakes make it appear that the GPL isn't really a license, or doesn't really bind, or doesn't really mean what it says. Never mind that no commercial vendor would _want_ licenses in general to be questioned- that their own EULAs and agreements depend on the same respect for the wording of licenses- the reality is that for Corel's legal people and many other people in their position, 'some are more equal than others' if they can get away with it. And that's not something that can be encouraged. In other words: yes, somebody at Corel would have no problem with simultaneously arguing that the GPL doesn't really need to be taken literally, yet the Corel EULA _does_ need to be taken literally. Hence this confusion, and this problem. The central irony is this: in order for Linux to be properly protected, Linux has to _demand_ equal protection for its intellectual property under the law, and equal respect for its licenses. On the other hand, the people responsible for endangering this IP protection are the people who _most_ depend on it. The logical consequence is that Microsoft ought to donate huge sums to the FSF in order to fight Corel in court over the GPL, because then Microsoft is rid of Corel, and the GPL is finally tested in court and given the same protection as EULAs and such. It's nuts, but it's logical. Who better to protect the rights to IP than corporations? And what is this but a challenge by Corel (intentionally or not) that risks weakening the GPL's right to impose terms on IP? Calmly, rationally, this nonsense has to stop.
I wish my question had made the cut, upon reading this mess of answers and flames. I was the one who asked "In the ideal KDE environment, what percentage of user input is expected to be:
Selecting menu entries, popup or root or Start
Entering text in text entry areas
Clicking a small button with an icon on it
Clicking a larger button with text on it
Radio buttons, checkboxes, listboxes
Keyboard shortcuts (combinations like meta-*)
Editing a dotfile
Which of these control types need to be made the sole control for a type of functionality? Which if any do you feel represents the preferred KDE method of doing things, and which if any are discouraged?" This was a trick question. It didn't get an answer, either trick or otherwise, but I can guess at an answer. I'd say the KDE people like chiclet-button-bars best, like menus and radio buttons and checkboxes pretty well, consider keyboard shortcuts a low priority, deprecate entering text wherever possible, and wish editing of dotfiles would just die;) The reason it's a trick question is because all these things and their resulting priority levels are copied off Windows, expecially buttonbars. There is no evidence to suggest these are in fact any easier for newbies to cope with- instead of mysterious invisible incantations, they become mysterious _visible_ graphic pictures. This is thought to be an improvement. The fact that this isn't significantly simpler doesn't make the mysterious invisible incantations any easier- it's true that the classic Unix approach isn't intuitive (whatever that means) all by itself. Once the various (and many) little tools all with different args are learned, _then_ the _building_ of larger tools out of the little tools _can_ be intuitive. People who have learned to do this tend to forget how tough the initial stages are- tough and tiresome. On the other hand, and I now know two KDE developers with this point of view, the idea that making a Windows-like desktop magically makes things easy for newbies to use is rubbish! I would hope these people were asking 'OK, so HOW is this easier then?' and analysing their work and looking at their GUI vocabulary to see what parts can easily generalise across the whole system- unwritten 'rules' that hopefully are learned by experimentation and generalised across all GUI-using programs, successfully. I don't see any evidence of this. At best the KDE people are choosing to NOT FOLLOW some of the more ugly Windows GUI mistakes and will end up with a nice inoffensive vanilla GUI with themes to conceal its basic blandness. At worst, they could take even less effort than Microsoft and end up with even more twisted and unobvious GUIs, all the while angrily claiming any critics obviously have never TRIED their masterpieces and so can't possibly know a thing. Because obviously you have to try a thing to know if it's good, right? Because obviously there is no right or wrong other than what people are used to, right? Because obviously if there ARE different sorts of people, then presumably 90% of them are all ONE sort, the consumer Windows-using plebian sort, and they couldn't possibly want other than the most obsequious handholding simplified GUI interface possible, right? Because they all went out and CHOSE Windows, therefore proving that concept of GUI accurately represented what most people wanted, right? ...
I think there are some major holes in these quiet assumptions. I would really like to get a better technical summary of exactly where KDE thinks it is going. Not long geeky diatribes on object models: pretty much nobody cares about that unless they program and like OOP. Not "KDE will win the desktop!", that's empty hype. I'd be interested in things like the breakdown of graphic object usage, in terms of what % buttonbars, what % checkboxes etc is the goal. If they cannot answer this then they have nobody thinking about human interface at all- there's nobody at the wheel and a whole lot of engines and gears rushing frantically... where? There are answers and answers, and some answers are copouts. Nobody asked the KDE people true human interface questions, except for Tom Christiansen- and his weren't used and were quite hostile anyway. As a result, the KDE people have said absolutely nothing about human interface with this exception: over and over again is the suggestion that usability is no more or less than familiarity with a set of rules- "learn KDE, use KDE, be happy, there is no rule 4". Anyone with a background or even cursory familiarity to human interface design knows that's a crock. There are rules. It's as involved and pervasive as the 'rules' of traffic flow in a crowded building, and simple changes can have profound effects on smoothness of workflow. Tom Christiansen knows this because HIS rules just do not coincide with what KDE offers. My own experience with KDE (which was the means of my first linux dialup, no lie) was not much more encouraging- compared to MacOS (a tough competitor, to be sure) KDE didn't seem to have a focus. The only rule seemed to be 'click on stuff and do things!' and that's not enough. It was enough to get me online- I clicked on stuff and kept doing things enough to make PPP connect, but it was like learning disconnected tricks. The common points seemed to be the presence of buttonbars on things, and a strong emphasis on forcing mouse actions over keyboard actions. I would like to see better thought taken, both in the KDE and for that matter the classic Unix CLI camp, on what the unspoken assumptions of interface design are. It's just not enough to merely soak these things up by osmosis- soak them up from Windows and you soak up a lot of chaos and lossage along with them. WHAT about a button bar is easy for a newbie? Going 'click' is relatively easy for a total newbie. What is easy about little pictures? They symbolise things, arguably pictures are more easily remembered than words (maybe). How to get a translation for the pictures? Experiment randomly or look for words (tooltips) that are not always present. What to do with the information gained? MEMORIZE it. Just like reading man pages and memorizing args, or referring back to the manpages habitually. This is NOT different from that old way of doing things. It's just as opaque, it's just in pictures this time! If serious thought isn't given to the underlying structure (quick, what order are paste copy and cut in KDE pulldown menus! Is it always the same? _Why_ was that order picked? How rigorous is the whole menu structure?) then the result is going to be a morass no matter if it's CLI or GUI. Unix CLI is already a morass- learnable, but a mess. KDE looks to be headed for a similar mess if the people involved don't quit with the kneejerk dismissals of criticism, and start listening. Normally I'd be more diplomatic, as I've usually considered the "It's your fault for being unwilling to learn KDE which is just as good as anything else, by definition" attitude as one person's opinion, and people have a right to their opinions. However, it's not just one person's opinion. It's heard from a fair number of KDE supporters and developers. It doesn't seem to be contradicted by anybody working on KDE- and this tends to minimize my desire to be diplomatic about it. Hence this little diatribe: no sense in my not calling it a diatribe, as others will anyhow. I'm just not impressed with the KDE attitude towards human interface. It looks like the KDE attitude towards human interface is take whatever was there, add whatever you want, call it the interface and expect people to learn it and like it or lump it. If they don't like it, rather than try to fix it you make it more configurable so they can make random changes. It's interesting to observe that this is EXACTLY what happened to create the very same classic Unix CLI that the KDE folks are horrified of. Both sides (i.e. KDE people + Tom Christiansen;) ) should mark that well- Tom, to get some perspective and learn that the KDE people are simply creating a very different mess which people can learn the details of just like they learned classic Unix: and the KDE people, to remain aware that it still _sucks_ even if the same approach with classic Unix led to stuff some people really liked. Now having offended everybody, I depart, chortling mischieviously;)
Interestingly, this makes an ironclad case that Corel are not legally allowed to be distributing their product! It (the Linux GPLed stuff of it anyway) is _only_ available under a license that explicitly states that if you can't conform to the GPL you can't redistribute. Great effort was taken to make that clear: IANAL but I doubt practical issues take precedence over what the license actually SAYS, and the GPL is unusually clear and direct for a license. I think it would be hard to reach any conclusion other than:
If Corel must impose extra terms and restrictions onto GPLed software, then the GPL forbids them to be a channel for the distribution of said GPLed software,
Corel are doing this for reasons of their own, possibly even compelled to do this for reasons of corporate liability or some such compulsion,
therefore Corel do not have a right to ship a Linux under these conditions, because the GPL _requires_ that they not add or remove restrictions, and no other license is available for their use for much of the Linux distribution.
This is certainly a drag for Corel, particularly if their corporate rules didn't allow them any wiggle room and they were compelled to try and add the age restriction, but unfortunately they're stuck- no corporation would want to weaken the binding force of contracts and license agreements in general. Unless Corel can stick exactly 100% to the GPL license agreements, they are _hosed_ and it says outright they can't use GPLed stuff in a distorted form. They could distort and change the _software_ all they want but they cannot change the rules without weakening all contract law- if you can disregard the GPL, then it'd be just as legal to disregard their own EULA, or Microsoft's EULAs, etc. ad infinitum. Perhaps Microsoft would like to donate several billion to the FSF to fight Corel with, thus keeping contract and license law solid and binding? MS or any huge corp. would _hate_ to think that its licenses and binding agreements were subject to being overruled if inconvenient. It's interesting to think that on this point, commercial software would actually defend the GPL like rabid wolverines. They may not _use_ the GPL, but to question the ability of a software license to be binding and relevant? They'd back it to the limit, because if the GPL gets weakened, their OWN licenses would end up just as weak.
If you must have a civilian flightsim, you've no business citing MS Flight Simulator when 'X-Plane' exists. The latter uses blade element modelling to simulate all airplanes by actually simulating them- no lookup tables in this one- on a home computer! Ten years ago (never mind twenty) this was unimaginable. On the other hand, if you cite MS Flight Simulator you should really be citing the source it came from- SubLogic A2-FS1. At least that's how I knew it, I understand it was a crossplatform product. A graphical flight simulator on an Apple ][ was truly a great hack, and my understanding is that MS flight simulator began with a purchasing of the SubLogic product. Regardless, Bruce Artwick was there first. (apologies if I've got any facts wrong)
Slashdot Mirror
Oh, don't be so picayune and specific, Jon: try to cover some more ground! ;)
Actually, no: you've missed some central things about GAs. In fact, given a sufficiently diverse 'gene pool', genetic algorithms work even _without_ mutation. Mutation is a sideshow: recombination ('crossover') is where the real action is. If 'organisms' wind up with roughly comparable ways of doing things, crossover has a much greater chance of producing a useful optimization than mutation does. Mutation is only significant if you're starting from a monoculture in which, at first, crossover would do nothing.
Okay, show of hands, who has heard of a 'pointer to a pointer' and why you might want to have one? Concept could be used in roll-your-own memory management like you have to do in C and C++ :)
One thing to bear in mind is that old slower processors can actually cost more than the new faster ones under certain conditions. I learned this when Powermacs were starting to take over and I thought a ultra-cheap onepiece Mac would be neat. Turned out 68040s actually cost more at the time than PPC chips, because the latter was where the volume was. Maybe you might think about embedded PPC or MIPS or something? Is there genuinely a market for, and producers of, 386 chips now?
I mean, go ahead, but that's very much like saying "Anyone with a 486 and Red Hat can write the next Q3test and do serious work on the kernel". To some extent it's the equipment...
Now. Seemed to be a lot going on in that hack, wasn't there? Why would I, a soundengineer hacker with his own music, be giving away secrets like that?Because that WASN'T enough. It'd help. But if you don't know how to make cables that will get as much of your signal to the A/D converters as possible... if you make the mistake of doing a lot of destructive digital changes on your data (it can be better to re-digitize a track than to normalize it over a simple gain change)... if you're using a crap digital mixer working in only 16 bits and not dithering properly, it's not going to be enough! You'll need help.
This is normal for any skilled profession, craft, or art. You wouldn't go out and get some random guy off the street for your drummer- it's no different for the sound engineer. The difference is, the audio geeks and equipment tweekers and snotty audiophile types now have a market value- and there are just as many of them out there as there are unsigned musicians.
Not everybody gets to be George Piros or Wilma Cozart or even Bob Ludwig or (shudder) Bob Clearmountain (if you don't know those names you're not a sound engineer geek. Show of hands? I bet some slashdotters know _all_ of them and why I shuddered at Clearmountain ;) ) but it's just the same as linux hacking- there are countless things to learn, it's a tremendously deep field, and you _can_ put together a 'garage' operation that competes with the big boys just the same as Linux competes with Windows NT.
It might involve a lot of geeky work. My mixing board had over 100$ of capacitors alone put into it. Not wizzy 'audiophile' caps of matching values- I increased values radically, now my board will put out bass on the order of 2 hz >:) there are most definitely audio hacks that can be done with equipment, it's a whole subculture.
I guess the long and the short of it is, at home with your PC or Mac you can _top_ the results of your average industry studio- if you're really willing to spend some years being a mad scientist audio hacker, or know someone who is. I always figure, I've been doing it for nearly 20 years, I can afford to give away everything and I'll still top ya in execution ;) so, here's a list of things to do/use/remember...
- Get serious monitoring. Learn how to place speakers in the right places from audiophiles. Get classical music or classic rock sounding great over the system- the stuff that requires the system to reproduce a soundspace (not synthesise one tho!)
- If you're using bass reflex speakers, stuff a sock in the ports. To mix very deep bass you've gotta be able to hear it. A sock will also provide resistive damping. You can enjoy port thunder later, now you need to hear what's actually happening.
- If you're using speaker wire try cutting the ends off AC extension cords (heavy duty indoor) and using those. If you're using those, try separating them all the way down into two individual wires, getting rid of capacitance effects. Crude but effective in getting more control of your highs.
- Litz wire is better than stranded or solidcore wire at transmitting analog audio in such a way that it digitizes nicely. You can make amazing cables from using all the wires in outdoor phone cable (four little wires rattling around loose in a big plastic sheath) because they'll work as sorta-litzwire and any shielding around the wire has a serious air gap for zilch capacitance over even long runs.
- Running cheap digital multitrack (like an old ADAT) into a reasonably decent analog mixer and then digitizing the result 'dithers' better than crap digital mixing software. (this is actually my next audio move- ADAT used in a hopped-up analog-uber-alles studio)
- Pitch shifting is destructive. Digital EQ is destructive. Normalizing is destructive. As with something like JPEG or MP3 these are cumulative. Don't ever edit things around like mad without backup- keep master digital copies in case you want to do a neatly executed series of digital edits on a clean copy. Screw up on this one, and your tracks will have all the life and interest of Radio Shack Casiotone demo tunes. Don't get track rot.
- Typically you can only afford to have a few sounds be 'big' or glossy or fancy- different genres approach this in different ways. Some great MODs have all the sounds big, but there's only about 2 or 3 tracks going at any one time! Conversely, some great rock mixes have, for instance, really big guitar sounds against a simple direct, dense bass tone without a lot of detail, and a very 'dry' drum tone. Even John Bonham's Led Zeppelin drum tone tended to be pretty 'dry': it just tended to sustain. When it was really wet and dense, the _guitar_ tended to be dry.
- Drum tones only sound right in context. You'll only hear what a bassdrum is doing when coupled with a bass guitar- the weight of it can be way more than you expected combined with the bass's transient attack. A snare can sound clear but boring until it's in context and sounds great- if you add stuff to make the unaccompanied snare sound hot and exciting, when it's in context you might lose most of the actual impact because everything gets muddy and confused. Think of the backbeat as a composite of drums and instruments- as if the guitar or whatever is _part_ of the snaredrum. Mix it as such.
- Never mix over headphones. Any headphones. It's a totally different presentation from speakers- your body needs to feel the sounds (even at a low volume, subliminally). Headphones are for tracking not for mixing over.
- Build monster speakers if you want to make club or house or rock music. You need to monitor over something comparable to what you'll be played on, only more accurate. I run towers with 12" 10" 8" and 6.5" woofers for (infinite baffle) bassbins. No one resonance dominates and the low end is understandably huge when required. Makes it easier to mix really serious bass content, you can hear exactly what it's doing. I also make my own tweeter elements. Audio geekiness is fun
:) also, using them as computer speakers makes games and such more fun. Big explosions, and I've occasionally encountered stuff like this one alpha-quality game in which the guy had made the sounds carry _major_ subsonic rumble for the explosions. Very neat. - Geek out on it. It doesn't take that much money, you just end up very well known at Radio Shack and very familiar with what you can get from MCM, Mouser catalogs et al. It's kind of like Linux in a way- you can DIY, and even beat hell out of the industry's approach in certain areas.
Good luck!This is actually so much better than the record company deals it's not even funny. (It's not funny, actually. It's tragic.) Sure, Joe Schmoe isn't guaranteed income as a Rock And Roll Star- when has he ever been? Sure, you have to do ALL your own promotion: guess what, you would in the mainstream industry as well. Record companies _don't_ promote new signed bands. They promote the Spice Girls. In the case of something like Hootie or whatever, you're talking about a local band that had a _killer_ marketing machine and network and did a lot of work without asking the record company to do it for them (which ain't happening).
Lastly, nothing is stopping you from setting up shop as an indie label. Press your own CDs- it'll be about a dollar a CD including labels and sometimes also throwing in cassettes, you supply artwork, you pay shipping and fill orders yourself and keep _all_ the profit after taxes. Nothing in the mp3.com agreement is forbidding you from doing all the work, not to mention you can make higher quality CDs from stuff that hasn't been mp3ed first.
The one thing mp3.com takes is this: everything you put up, they can spread around in any way they choose. Well, duh- that's the point! If they were not going to let people download it, splash your little graphics around, drop your name in their email mailings, do a web-radio station and play your songs all they want, what would you be there for, why would they be giving you lots of HD space? You can back out (unlike the regular industry), it's nonexclusive (unlike the regular industry), it's clear that you aren't losing copyright to your own work... (unlike the regular industry... are we seeing a pattern here? Ever wonder why certain artists 'seem' to be really into selling their songs to crappy compilation album producers?)
Thanks for the link. I made a copy on my HD of the agreement to go over at my leisure- but I gotta tell you, I'm impressed with it. I don't see much of a gripe, compared to existing industry practices. They were supposed to provide MP3-sized HD space to zillions of unsigned musicians for nothing maybe? ;)
I'm picturing a checkbox labelled "Allow untrusted users to plant Trojan horse programs" :) of course, it defaults to off except for when you set Office to 'Active Content' :)
The Message from Seattle: Girding For The Fight Of The 21st Century
Is this really _Katzbot_? o_O
*hits backlight* (bip!)
*turns it back off, closes cover* (bip! clik)
This was $30. I once wanted an old Powerbook- just something that could take text notes and be carried around. B/W screen, some tiny amount of ram- never could get one, too expensive.
Now I have a toy with 256k of ram (that's an awful lot of little memos!). It has a backlight, which is more than the old Powerbooks had. It's got chiclet keys that go (bip!). There's a fourway arrow key thing that is occasionally relevant. The text editing is rather like vi or something- hell, the whole thing is extremely modal, and yet it's such a little thing that it doesn't matter- I don't need a trackball or color for this. So where a few years ago I could only wish for an old Powerbook (which wouldn't fit in my pocket without _serious_, er, percussive redesign ;) ), now I can actually have a little toy 'laptop' that goes (bip!) and takes notes and shows what time it is in Hong Kong or Berlin.
And this is totally cool- and I can't wait until I get to have one for $50 that's the same size, the same lcd screen with blue backlighting- but has 4M, bash, and vi ;)
That would just feel so good tucked into your pocket. Imagine. Hip-nix ;) (poc-nix?)
Why is this? Because it was Pixels: the little company that makes Pixels3d, a modeller/renderer etc for the Mac. I already knew them- in fact I keep a copy of Pixels3D 2.1.4 around because they are one of the many Mac vendors who have taken to releasing their last-year's model at no cost, and I grabbed it. Being allowed to fully use something like that left me with a good feeling about them, and they are the antithesis of a big vague corporation- it's a bunch of computer geeks running a company, and their product kicks butt (except that I hate the Lightwave-like interface :) )
I wasn't able to take 'em up on their promo, won't be buying anything today- but, you know, I am _very_ used to shutting off telemarketers. I don't give them three seconds. I interrupt, I firmly say I'm not interested and then hang right up. Yet in this case these people were able to keep my attention and get my sympathies- why?
Partly because they were ready to put some serious effort towards getting me what I wanted. I learned that the scripting language was like Renderman shaders. I learned people write plugins in REALbasic- hey, I have that! I even ended up talking to the main programmer for about ten minutes on how many semitransparent layers you could stack to simulate volumetric clouds (a POV trick I've been playing with), and he had all the techie details. It was so deeply about what _I_ wanted to know, rather than about what they wanted to sell.
I hear people saying IBM is also taking this approach. Well, good for them! The predatory thing only works when you have a lock on people. Pixels doesn't- they do Mac software, and compete with everything from Metacreations (Poser, Bryce, Infini-D) to Lightwave itself and who knows what else? Seeing this glimpse of how they work with their customers gave me a bit of insight into why they're still around at all. Did you know that you can go to their community page and they will put up _pictures_ of their users? (one wonders if it's pictures of _most_ of their users! ;) ) Looking down the row of faces, next to banner links to the respective websites, was a lesson in PR.
It doesn't always stay- that language I use, REALbasic, is very neat but the marketing people have taken to trying to get people to link to the RB site by offering space on the CD in exchange for posting 'Made With Realbasic' logos on things. That's a 'what can we get' approach, not a 'what do you want?' approach. I don't mind it but I'm not doing it. Pixels is smarter- or wiser. As, apparently, is IBM...
Cinema: A Whole New Chapter In The Revolutionary World
Letters From Virtual Impeachment
The Web: A Whole New Chapter In Online Buffy
Giving Thanks For Virtual Cinema
The Power Of The Techno Stock Market And Interactive Dying Babies
Murder: A Whole New Chapter In Hidden Being Different
Is The Stock Market About To Transform Slashdot?
Can Hidden Anonymous Cowards Stop Post-Columbine Minimum Wage?
The scary part is, this _reeks_ of Markov Chaining. I suspect these are made up of actual genuine interactive dying baby Katz headlines, shredded and rearranged.
One envisions doing this to the whole Katz on Slashdot collection, thus getting not only terrific headlines, but entire articles. For anybody interested in doing this, Katz is a _very_ good subject for travesty as he is disjointed and covers the same subjects over and over. To markov chain this stuff it's best to work at the word level- pairs of words end up chaining very abstractedly, and word triples can sometimes spit out overlarge chunks of unaltered text (though I suspect with Katz the triples would work perfectly).
Here's a sample of this sort of travesty: Speak Roughly To Your Evidence, an edited Alice In Wonderland travesty.
I'm a musician. Mind explaining to me why I shouldn't have control of the means of production? I'm not talking about distribution, I'm not suggesting I should get free promotion: I'm just saying that the current approach seems to be reliance on say 40 special encryption keys, and you have to basically be a multinational to get one.
I have no problem with all this if equvalent _unencrypted_ data can be played back in consumer decks- if I can burn a DVD and distribute it using my own limited resources. I am getting the feeling that this is a 'Why would you want to do that?' situation to the big conglomerates- and that it's not necessarily going to be happening.
I really don't _care_ if it's easy or hard to copy off major-label releases. I'm not even interested. I'm more interested in what all this means to me- and if we're looking at a world where you CAN'T MAKE the 'records' of the future, scant years after it became seriously easy to do just that with the once-highly-techie audio CDs, well, I have a problem with that.
Again: give me ability to record unencrypted and release product, or give me ability to use some default encryption if you _must_ have it, but if the ability to produce these disks will be limited to the industry only, with 'indies' finally legislated out of existence by technology itself, then that is a very ugly world for an artist. It's bad enough being a musician without such crap. This is the thin end of the wedge- it _won't_ stop there. The logical direction this is going is to stomp out support for old formats that can be easily ripped from (and, coicidentally, easily released on), and then for the indie musician, not even cassette tapes or CDs would remain, and it'd be "Oh, please, mister consumer man, won't you buy a worn-out old CD player to mount under your _new_ wonderful encrypted CD player so that you can listen to _my_ music too? It won't take up too much room." ;P
I can't even begin to explain how bad that would suck, and I'm already boycotting the industry as hard as I can. I think us indie musician types need help. The odds were already heavily against us, but now it's just getting ridiculous, and there's no reason to expect it'll level off by itself. Hopefully there will always be some sort of mp3 underground, as that might end up the only option for a lot of us. We could be watching all of the following:
That said, just where do you get off suggesting that _protest_ is going to change this? (I'm not even going to get into 'suggesting that looting Starbucks is going to change this')
I would suggest, in all seriousness, that protest, already meaningless except in a (dangerously open to misinterpretation) publicity sense when used against governments, is totally useless against corporations. Corporations can replace ALL the beef in your supermarket with genetically enhanced hormone pumped beef, by economic leverage. Corporations can make things like this happen and you can't complain to an elected representative as there isn't one, and you can't complain to the corporation even if they are sympathetic because they do not have, shall we say, lots of legal support for going against the interests of the stockholders- I question if the stockholders can even get together and say 'do X, which will cost an extra 2%' and have it override the corporation's obligation to go after that 2%.
So protest becomes irrelevant. I would suggest that you'd better learn to fight, instead. I don't know exactly how this would be done: I don't claim to have all the answers. Possibly IT fighting, cyber-attacking a corporation, would hurt it. Very likely continued physical attacks on the corporate executives would make it difficult for the corporation to conduct its business, though you cannot hurt the corporate entity itself, and anyhow corporations can and do hire bodyguards, chauffeurs who are taught anti-hijacking by Bob Bondurant etc. so a corporate exec is a _tough_ target, as tough as any government official. Finally, the corporation is a creature of law so it can be attacked by law- maybe. The trick is, they are increasingly calling all the shots in the legal sphere- we wouldn't want to have to overthrow the fscking _government_ just to get control of the legal system back. All in all, it's a very nasty puzzle.
I do acknowledge the dangers of the WTO, against which all these groups from Greenpeace to loggers to environmentalists are aligning. I think maybe I consider it more dangerous than you do... but I doubt it is remotely helpful to have sit-ins, or seize Seattle by a _mob_ (a militia, maybe, but a mob??) or to loot Starbucks.
doh!
Remember the Mothers of Invention? Frank Zappa noticed- and produced a lot of music to try (and fail) to shake up a lot of people who, quote, "mindlessly accepted everything they were given, without questioning it". That was the hippies, to Zappa.
Woodstock was paid for entirely by John Roberts. Remember that name, or Joel Rosenman's? John Roberts financed the whole festival, having been conned to believe it was a moneymaking venture, originally put forth as a 'press party' for a recording studio to be built. Roberts pretty well had a breakdown while the festival was happening- left totally responsible, far in debt, being asked to sign checks for which there was no money anymore, Roberts and a few other people took the whole load of the Woodstock Festival upon themselves, a festival that was declared free after it had already eaten through Roberts' entire inheritance.
The whole hippie concept is a story of rip-off, stealing, lying, and destruction, painted to appear as virtue and freedom. Virtually nothing was accomplished- the end of the war in Vietnam, for instance, owes much more to the fact that eventually Middle America was sick of it and wanted it stopped, and to Nixon's bid for re-election.
Having hippie idealism on the Net is a bad thing. It's spelled out quite obviously: "Let's march on Website X and stomp it with DOS attacks!". No thought is given to other services that may be hosted on that computer, no interest is taken in the additional load produced by X many lusers running a Javascript program and monopolizing all the network links to the target. The idea of responsibility is seriously lacking here.
It's like a riot in cyberspace: riots were seen as civil disobedience in the 60s. My generation saw them more clearly: "Tomorrow you're homeless- tonight it's a gas" -The Dead Kennedys
Rioting is not freedom. Rioting is collectively throwing a fit. If you want civil disobedience, "smash the right windows" (Lee Felsenstein)- get smart, intrude, change their web page, don't just riot in the cyberstreet smashing everything. Stupidity is not insulation, it will not protect you.
If you want FREEDOM, then write fscking software! This is the most annoying aspect of all this. DOSing a site that happens to contain something you don't like is freedom? Write software, GPL it (or BSD license it depending on if you don't want to _enforce_ the availability of the code), put it out there. That does more for freedom than any twelve hippie web pages. If you are a blackhat at heart, learn how to pull off intrusions into whatever's out there, get good at the surgical strike, be smart enough to spare the environment you're in while punishing your enemies. That's harder, of course: a lot harder, in fact. But there's no excuse for the hippie approach. It's a disaster, a mess! Fight smart or go do bong hits, if you can't get a clue then get out of the way.
Otherwise you might well find that the GenXers (_my_ generation, thank you) have very much their own opinions on what activism is. You might find that they take a dim view of mindless destroying to prove some vague point. You might find some GenXer who's done his or her homework sneaking onto your precious target site and setting up some sort of viral attack from Javascript on the site itself- which itself attacks the site, but also whacks all the idiot DoSers in the bargain.
Hippies are the Commodore 64s of making change happen. It's time to move on. It's time to get _serious_. For example, the GNU GPL uses the law and copyright to attack what copyright is _normally_ used for, and Linux takes the GPL and proliferates it wildly- now there is a huge amount of Linux out there, and it's got the legal backing to fight attempts to subvert what it stands for. Now that's change. That's _significant_ and it matters and it's constructive but uncompromising.
Forget the hippie approach. Go with the Linux approach. Build something good and be ready to protect it.
No. They need to stop that. Period.
I don't agree that it's that guaranteed. Perhaps what I'm saying is that I don't have the faith you have that this is the case- especially when we're talking about a community-used license going up against the lawyers of big corporations. You do realise that if (to use the most extreme example I can think of) somebody bribed the judge to make an exception and find the GPL 'not really a license', in other words defining it by usage rather than by the language, that would be the legal precedent? I really, really don't have the faith in the virtuous and incorruptible judicial system you do. I'm damn glad it works as advertised a lot of the time, but I'm simply not ready to assume that people can use the GPL differently than the wording of the license and not raise the question of later consequences. I think we will have to disagree there.
*takes it to email. yeesh.*
I haven't seen many slashdot posts, even my own, so important that they needed to be posted twice. ;)
:)
However, I shall take the point, and since I like brs instead of ps for linebreaks, I shall take to using two of 'em between paragraphs. HTH
I'm seeing quite a bit of sentiment indicating that Linux users and GPL-using authors should ask Corel to apologize, to say 'oops', and then should sit back and trust them.
Well... of course somebody at Corel will be told to say 'oops'. That's his job. Its name is 'Public Relations'.
There's also people there whose job it is to figure out the next testing of the outside of the envelope. They are called 'lawyers'.
Forget that they have PR people for a minute, and also forget about their lawyers, and let's just look at their ACTIONS, okay? Their actions have consistently been causing problems. It doesn't matter that there's somebody to say 'oops'. That is a PR person, that's their job, it does not mean as much as people would like to believe.
I personally have a lot more respect for Bruce Perens for momentarily _refusing_ to listen to the PR flacks and looking at what Corel is repeatedly doing. I'm not sure if people realise that permitting such abuses of the GPL is very similar to letting people talk about xeroxing documents...
If you do that, the Xerox people _will_ speak to you and make sure you are aware that you are using their registered trademark as if it were a generic term, that other companies have lost their trademarks over just such weakening of the mark, and that Xerox (r) does not intend to weaken its mark in such a manner.
Well, the behavior by Corel obviously weakens the GPL- to the extent that it goes unchallenged. Anybody picking a fight with the GPL will be able to point at the Corel actions involving it, and the community responses to it. They will be able to truthfully say to a judge or jury, "This is not actually binding, because it is not obeyed or respected by even the people who use it. Look, here are 500 slashdot posts saying that Corel should not be penalized in any way for their alterations of the license, because Corel will benefit the community. Isn't it true that the GPL is a legal fiction that is not actually intended to be followed- as seen by these posts effectively waiving it in the court of public opinion... and therefore, shouldn't it be declared null and void?"
That's a perfectly legitimate approach for a lawyer to take, to invalidate the GPL. It's Corel above all who are making that argument possible, by repeatedly adding and altering conditions- and it's Linux users who are helping, by assenting to this alteration of the license, and making the argument that Corel should not be held responsible to the terms of the license.
I'm sorry, but I don't assent to the alterations of the license. I have software out there under the GPL, and I don't remember giving Corel permission to alter, challenge and weaken the license for their own convenience. I'm not concerned with how quickly a PR flack can be found and made to say 'oops'. These changes need to stop, or Bruce Perens' hasty reaction will be the only sensible course of action in the long run.
I _realize_ people want to be nice and not seem threatening or combative, but this is not fantasyland, or high school: this is the real world, there are consequences, and allowing a widely used agreement like the GPL to be defined differently in practice than it is in writing is bad. It makes an argument possible that the written version isn't really the applicable one- just as we need to look at Corel's _action_, so the courts will assuredly look at the Linux community's _action_ to ascertain how valid the GPL is when challenged. Do we, seriously, want to establish the precedent that you get to do whatever you want as long as you don't keep doing it for so long that it gets annoying? More relevantly, how many people advocating that Corel be let off easy have themselves written GPLed software- which of these people don't actually have anything to lose?
I think if attention is paid to the Linux community's wishes as to how closely the GPL must be followed, the attention needs to go to the people who have written software that uses it. I don't think it's at all helpful for people who aren't actually using the license on their own work to say 'Take it easy, maybe they don't have to obey this license that closely'. I think most people who do use the GPL on their own work would not wish legal precedent to be set that the license isn't really real, that in practice it's customarily altered and bent to fit situations. I certainly do not. From Bruce Perens' initial reaction, I don't think he wishes that conclusion to be drawn either.
It's all very well to be calm and rational and reasonable. However, in the event of someone calmly and rationally taking actions to basically weaken the GPL by setting a precedent to alter pre-existing licenses on the grounds of convenience, I would calmly and rationally have their head for it :)
More accurately, since I _have_ read the GPL, do have a reasonable idea of what it requires, and have released software under it based on my desire for my licensing terms to be _just_ _like_ _that_, I will calmly and rationally suggest that Corel not only has to either recant quickly or lose their permission to redistribute GPLed software, I would also suggest that this 'worsening of the signal to noise ratio' is harmful. It is harmful because the more this is allowed, the more will happen. I would suggest (calmly and rationally) that it is very much a bad thing that Corel is permitted to blunder in this manner. Their mistakes make it appear that the GPL isn't really a license, or doesn't really bind, or doesn't really mean what it says. Never mind that no commercial vendor would _want_ licenses in general to be questioned- that their own EULAs and agreements depend on the same respect for the wording of licenses- the reality is that for Corel's legal people and many other people in their position, 'some are more equal than others' if they can get away with it. And that's not something that can be encouraged.
In other words: yes, somebody at Corel would have no problem with simultaneously arguing that the GPL doesn't really need to be taken literally, yet the Corel EULA _does_ need to be taken literally. Hence this confusion, and this problem. The central irony is this: in order for Linux to be properly protected, Linux has to _demand_ equal protection for its intellectual property under the law, and equal respect for its licenses. On the other hand, the people responsible for endangering this IP protection are the people who _most_ depend on it. The logical consequence is that Microsoft ought to donate huge sums to the FSF in order to fight Corel in court over the GPL, because then Microsoft is rid of Corel, and the GPL is finally tested in court and given the same protection as EULAs and such. It's nuts, but it's logical. Who better to protect the rights to IP than corporations? And what is this but a challenge by Corel (intentionally or not) that risks weakening the GPL's right to impose terms on IP? Calmly, rationally, this nonsense has to stop.
- Selecting menu entries, popup or root or Start
- Entering text in text entry areas
- Clicking a small button with an icon on it
- Clicking a larger button with text on it
- Radio buttons, checkboxes, listboxes
- Keyboard shortcuts (combinations like meta-*)
- Editing a dotfile
Which of these control types need to be made the sole control for a type of functionality? Which if any do you feel represents the preferred KDE method of doing things, and which if any are discouraged?"This was a trick question. It didn't get an answer, either trick or otherwise, but I can guess at an answer. I'd say the KDE people like chiclet-button-bars best, like menus and radio buttons and checkboxes pretty well, consider keyboard shortcuts a low priority, deprecate entering text wherever possible, and wish editing of dotfiles would just die
The reason it's a trick question is because all these things and their resulting priority levels are copied off Windows, expecially buttonbars. There is no evidence to suggest these are in fact any easier for newbies to cope with- instead of mysterious invisible incantations, they become mysterious _visible_ graphic pictures. This is thought to be an improvement.
The fact that this isn't significantly simpler doesn't make the mysterious invisible incantations any easier- it's true that the classic Unix approach isn't intuitive (whatever that means) all by itself. Once the various (and many) little tools all with different args are learned, _then_ the _building_ of larger tools out of the little tools _can_ be intuitive. People who have learned to do this tend to forget how tough the initial stages are- tough and tiresome.
On the other hand, and I now know two KDE developers with this point of view, the idea that making a Windows-like desktop magically makes things easy for newbies to use is rubbish! I would hope these people were asking 'OK, so HOW is this easier then?' and analysing their work and looking at their GUI vocabulary to see what parts can easily generalise across the whole system- unwritten 'rules' that hopefully are learned by experimentation and generalised across all GUI-using programs, successfully. I don't see any evidence of this. At best the KDE people are choosing to NOT FOLLOW some of the more ugly Windows GUI mistakes and will end up with a nice inoffensive vanilla GUI with themes to conceal its basic blandness. At worst, they could take even less effort than Microsoft and end up with even more twisted and unobvious GUIs, all the while angrily claiming any critics obviously have never TRIED their masterpieces and so can't possibly know a thing.
Because obviously you have to try a thing to know if it's good, right?
Because obviously there is no right or wrong other than what people are used to, right?
Because obviously if there ARE different sorts of people, then presumably 90% of them are all ONE sort, the consumer Windows-using plebian sort, and they couldn't possibly want other than the most obsequious handholding simplified GUI interface possible, right?
Because they all went out and CHOSE Windows, therefore proving that concept of GUI accurately represented what most people wanted, right?
I think there are some major holes in these quiet assumptions. I would really like to get a better technical summary of exactly where KDE thinks it is going. Not long geeky diatribes on object models: pretty much nobody cares about that unless they program and like OOP. Not "KDE will win the desktop!", that's empty hype. I'd be interested in things like the breakdown of graphic object usage, in terms of what % buttonbars, what % checkboxes etc is the goal. If they cannot answer this then they have nobody thinking about human interface at all- there's nobody at the wheel and a whole lot of engines and gears rushing frantically... where?
There are answers and answers, and some answers are copouts. Nobody asked the KDE people true human interface questions, except for Tom Christiansen- and his weren't used and were quite hostile anyway. As a result, the KDE people have said absolutely nothing about human interface with this exception: over and over again is the suggestion that usability is no more or less than familiarity with a set of rules- "learn KDE, use KDE, be happy, there is no rule 4".
Anyone with a background or even cursory familiarity to human interface design knows that's a crock. There are rules. It's as involved and pervasive as the 'rules' of traffic flow in a crowded building, and simple changes can have profound effects on smoothness of workflow. Tom Christiansen knows this because HIS rules just do not coincide with what KDE offers. My own experience with KDE (which was the means of my first linux dialup, no lie) was not much more encouraging- compared to MacOS (a tough competitor, to be sure) KDE didn't seem to have a focus. The only rule seemed to be 'click on stuff and do things!' and that's not enough. It was enough to get me online- I clicked on stuff and kept doing things enough to make PPP connect, but it was like learning disconnected tricks. The common points seemed to be the presence of buttonbars on things, and a strong emphasis on forcing mouse actions over keyboard actions.
I would like to see better thought taken, both in the KDE and for that matter the classic Unix CLI camp, on what the unspoken assumptions of interface design are. It's just not enough to merely soak these things up by osmosis- soak them up from Windows and you soak up a lot of chaos and lossage along with them. WHAT about a button bar is easy for a newbie? Going 'click' is relatively easy for a total newbie. What is easy about little pictures? They symbolise things, arguably pictures are more easily remembered than words (maybe). How to get a translation for the pictures? Experiment randomly or look for words (tooltips) that are not always present. What to do with the information gained? MEMORIZE it. Just like reading man pages and memorizing args, or referring back to the manpages habitually. This is NOT different from that old way of doing things. It's just as opaque, it's just in pictures this time! If serious thought isn't given to the underlying structure (quick, what order are paste copy and cut in KDE pulldown menus! Is it always the same? _Why_ was that order picked? How rigorous is the whole menu structure?) then the result is going to be a morass no matter if it's CLI or GUI. Unix CLI is already a morass- learnable, but a mess. KDE looks to be headed for a similar mess if the people involved don't quit with the kneejerk dismissals of criticism, and start listening. Normally I'd be more diplomatic, as I've usually considered the "It's your fault for being unwilling to learn KDE which is just as good as anything else, by definition" attitude as one person's opinion, and people have a right to their opinions. However, it's not just one person's opinion. It's heard from a fair number of KDE supporters and developers. It doesn't seem to be contradicted by anybody working on KDE- and this tends to minimize my desire to be diplomatic about it.
Hence this little diatribe: no sense in my not calling it a diatribe, as others will anyhow. I'm just not impressed with the KDE attitude towards human interface. It looks like the KDE attitude towards human interface is take whatever was there, add whatever you want, call it the interface and expect people to learn it and like it or lump it. If they don't like it, rather than try to fix it you make it more configurable so they can make random changes.
It's interesting to observe that this is EXACTLY what happened to create the very same classic Unix CLI that the KDE folks are horrified of. Both sides (i.e. KDE people + Tom Christiansen
Now having offended everybody, I depart, chortling mischieviously
- If Corel must impose extra terms and restrictions onto GPLed software, then the GPL forbids them to be a channel for the distribution of said GPLed software,
- Corel are doing this for reasons of their own, possibly even compelled to do this for reasons of corporate liability or some such compulsion,
- therefore Corel do not have a right to ship a Linux under these conditions, because the GPL _requires_ that they not add or remove restrictions, and no other license is available for their use for much of the Linux distribution.
This is certainly a drag for Corel, particularly if their corporate rules didn't allow them any wiggle room and they were compelled to try and add the age restriction, but unfortunately they're stuck- no corporation would want to weaken the binding force of contracts and license agreements in general. Unless Corel can stick exactly 100% to the GPL license agreements, they are _hosed_ and it says outright they can't use GPLed stuff in a distorted form. They could distort and change the _software_ all they want but they cannot change the rules without weakening all contract law- if you can disregard the GPL, then it'd be just as legal to disregard their own EULA, or Microsoft's EULAs, etc. ad infinitum.Perhaps Microsoft would like to donate several billion to the FSF to fight Corel with, thus keeping contract and license law solid and binding? MS or any huge corp. would _hate_ to think that its licenses and binding agreements were subject to being overruled if inconvenient. It's interesting to think that on this point, commercial software would actually defend the GPL like rabid wolverines. They may not _use_ the GPL, but to question the ability of a software license to be binding and relevant? They'd back it to the limit, because if the GPL gets weakened, their OWN licenses would end up just as weak.
If you must have a civilian flightsim, you've no business citing MS Flight Simulator when 'X-Plane' exists. The latter uses blade element modelling to simulate all airplanes by actually simulating them- no lookup tables in this one- on a home computer! Ten years ago (never mind twenty) this was unimaginable.
On the other hand, if you cite MS Flight Simulator you should really be citing the source it came from- SubLogic A2-FS1. At least that's how I knew it, I understand it was a crossplatform product. A graphical flight simulator on an Apple ][ was truly a great hack, and my understanding is that MS flight simulator began with a purchasing of the SubLogic product. Regardless, Bruce Artwick was there first. (apologies if I've got any facts wrong)
...well, that's what all this seems to be saying.