Because the GOVERNMENT taxes us like an inmate fucks his bitches. Stop stealing people's money and they can spend it elsewhere, creating jobs, giving to charity, etc.
Alright, but it certainly isn't causing the poverty. And do you really think that people would fix the problem if they weren't taxed so much?
Suppose you didn't have to pay a dime in taxes next year. How much of that would you spend dealing with poverty?
Now, suppose you got a raise to a job that pays twice as much next year. Same question.
Fact is, if people aren't solving the problem now, taxing them less isn't going to fix it.
"But without those regulations, people would die from snake oil and Dr. Bob-Earl-Smith's toolshed clinic!!!" you cry. I ask you, what about the myriad souls who lose their life every year because they don't have access to that new life-saving drug that *might* be harmful?
What about them?
You make an interesting point, but don't you think there would be far more of us dying from Dr. Bob-Earl-Smith if we didn't have a decent amount of regulations? You may want to fix the regulations, but I hope you don't actually want to remove them.
25% of the worlds prison population
Because the GOVERNMENT is throwing people in jail for smoking joints, while throwing out the violent offenders to make more room for more non-violent hippies.
No disagreement here. However, wouldn't you rather your capslock GOVERNMENT would talk about this stupidity, rather than the "issue" of "obscenity"?
46800 car deaths in 2005
And this is in government's field because... ? I had a loaded semi tractor trailer rear-end me earlier this year, nearly killing me and my father; government wouldn't have been able to do a damn about that.
Well, since you asked: Better free public transportation, tighter restrictions on licenses -- maybe the truck driver shouldn't have been driving?
I don't have all the answers, but wouldn't you rather they actually talk about this, and take this on? It's not like it's any more hopeless than the "war on terror", and terrorism actually costs us far fewer lives.
Higher education costs and arm and a leg and your first born.
Ah, this would be because the GOVERNMENT has a monopoly on education.
What. The. Fuck?
Do the words "private school" mean anything to you?
Destroy the monopoly, and the free market will provide better quality education at a lower price
So you're saying that in a free market, a kid would walk into school and get paid? I mean, you do realize that public school is free, right?
I mean, that's the whole point of public school in the first place. It's free.
You may have more success with the gay marriage thing if you stop insisting on calling it marriage.
We might also have more success with the Linux thing if we stop calling it Linux. We might also have more success with the Free Software thing if we stopped calling ourselves hackers.
If you want to simply cause the state to recognize this unions in the same manner as unions between heterosexuals, you will probably win over a lot more people.
That's actually most of what I want. The problem is, the state currently calls such unions "marriages". You strike the word "marriage" from every law, and then we'll talk about calling it something other than gay marriage.
Or, the religious people can stop getting in such a righteous fury over this kind of language, and accept that "marriage" as the state is concerned doesn't dictate what "marriage" means to their religion. They can allow gay marriage to be legal, and still preach that it's not really marriage.
Except that this really isn't usually their goal. One I talked to recently basically said that marriage is between a man and a woman, and he wouldn't mind if the gays up the street got a civil union and pretended it was a marriage -- but he also felt that marriage should be recognized by the state, and should receive special benefits not given to civil unions. So, the whole argument about language was just a smoke screen.
An unfortunate consequence for your cause-- if you wish to prove that you are truly interested in equality and not just an agenda-- is that any two (or more!) people who live together will be claiming social partnership benefits.
And what's wrong with that?
No, really, I'm waiting for your answer, because chances are, anything you have to say could be made to work as part of the definition of "civil union".
... it is not lying... it is, and has been for many years called marketing.
That is because marketing, for many years, has been institutionalized deception when it isn't outright lying.
the beer commercial shows you that when you open up one of their beers you get 20 naked women to show up at your party, but the "other" beer brings balding middle aged men.
And you know it's a beer commercial, and it's also ludicrous enough that you know it's a joke. How is this related to something plausible that deliberately tries to hide the fact that it's an advertisement?
Hell, Some drug ads never say what they do (so they don't have to give side effects) similar to Sony not saying this is an ad.
How is that similar? Some drug ads never say what they do. It's at least obvious that they're drug ads -- or at least, that they're ads for something!
You'd have a point if the drug ads pretended to be documentaries, but they don't. Thank you for reminding us that even drug companies are more ethical than Sony.
And based on your ethics, I strongly suspect that you work for Sony. If you do, go listen to the Bill Hicks rant, and kill yourself.
While you have a valid point, and one I've been trying to make, you also have a bad example. My Mac laptop is missing a lot of features I take for granted on Linux, even ones with decently simple interfaces. Just to start with, there's no package manager.
Funny you should mention that. The Gimp, in its default configuration, reminds me very much of MSPaint. It has a few more tools, but they all look the same as the Paint ones, and it's easy to find the ones you need.
There's also tons of complexity hidden in menus, but if you know how to use Paint, chances are you already know how to use the Gimp, except you can now type in an extension (other than.bmp), and it will almost certainly support whatever image type you specify.
Paint doesn't crop. You must select a section and cut it out, then paste it into a new workspace.
Thus, cropping.
This is an important point -- you don't need a "crop" feature, because you can easily acheive it from the other features. I'm not saying there shouldn't be a crop button, but this is an example of how you can make a program simpler and yet still have it be as powerful.
It also cannot select anything outside the visible part of the window because the scroll bars won't move, and you can't resize the view.
From what I remember, you're just wrong. I could easily click inside the window (rectangular selection), then drag it outside the window, and the scrollbars would follow. You can resize the window or the image, and you can do it all with simple mouse movements, though it may not be that intuitive.
Also, you can zoom in and out, and the tools work at all zoom levels.
But if you think simplicity means "not very many features" or "does one thing and does it well," then I applaud your integrity but you can't go that far with a product that deliberately leaves features out.
And yet, it can still be made simpler, and he's completely missing the point of "does one thing and does it well".
Consider: Both gzip and bzip2 do one thing, and do it well. Yet, by combining these with tools like dd, cp, tar, nc, ntfsclone, or less, I can create all kinds of features. Consider: tar has support for gzip and bzip2, but it does it by calling the gzip and bzip2 programs, and it's the exception.
This guy would rather have every single commandline utility link against zlib.
Now, that doesn't prevent you from making a big, bloaty app, and still having it be "simple". You just have to make individual features of that app combine in the way that gzip and tar combine. A simple example is word processing -- notice how there's a bold, italic, and underline button, but no bolditalic button. If you want bolditalic, you click the bold button, then the italic button. If someone wanted to spend a lot of effort on design, they could probably make a very simple word processor, with most of the functionality organized such that it feels like you're using AbiWord or TextEdit, but it's actually more powerful than OpenOffice.
And he misses the point with Google. Consider that most of the features on Yahoo's homepage are available through Google's services, they just don't clutter up the homepage.
Given all of this, your point:
even if 80% of your users only use 20% of the features, it's probably not the same 20%.
only applies if, as you say:
only an idiot thinks simplicity is equivalent to fewer features.
And you could even have "fewer features" and yet have it be more powerful. LISP is one of the most powerful languages we have, and it also has incredibly simple syntax and builtin stuff. If you need a missing feature, you implement it yourself, naturally -- as naturally as I implement a bolditalic button by clicking the bold and italic buttons.
That works... except then everyone will try to find some way to mis-interpret laws in order to be able to break them.
Your way is almost as bad. Mis-interpret just about every finable offense ever, and simply avoid the fines.
The trick is to come up with some good, solid language for obviousness, that itself passes an obviousness test. I'm not sure we're capable of that as a species yet, but it might help to have that as a goal. (Or find a completely different approach that would achieve the same result.)
I suppose one could claim that I'm trying to devise an idiot-proof Constitution, which of course would cause even greater idiots to be elected to Congress.
Still a decent goal.
The trouble is, even if we all had a perfect understanding of the 18th century mind, and so we knew intimately how the Constitution was intended to be interpreted, there are parts of it that are simply obsolete -- that 3/5ths rule is a perfect example.
Rewriting from scratch is a scary prospect, but it may be easier than trying to reform a document that's over two hundred yeras old.
It's to prevent tyranny by the government, by restricting the amount of force that can be used against the people.
What we need is a way to not cripple legitimate things, like SWAT teams, and yet still prevent a police state. Having a civilian be Commander in Chief is a step in the right direction.
After all, if the government is really going to be tyrannical that way, they can simply declare martial law, or ammend this limitation away -- so we need another solution anyway. However, SWAT teams do seem legitimate to me -- although if I go back and read your original proposal, it might cover that (let us buy all the equipment the SWAT teams use).
In other words, if the law isn't clear to a reasonable person (and hasn't been made clear by a previous precedent-setting court case), then if you're found guilty of breaking that law you can't be punished for it, beyond corrective action (so, in an example pertinent to this topic, if the law isn't clear about owning fully-automatic weapons, and you have one, they can take it away or modify it to be semi-automatic, but they can't fine you or throw you in prison).
Not good enough. If the law isn't clear enough, it should be void, which means I get to keep my fully-automatic weapon until they fix the law.
If your rule is implemented, politicians would pass laws that no one wants except their own special interests, but that are confusing enough that no one knows just what they do until they're directly affected. It's not good enough to just say that you can't be punished -- the law could've sat around for years until they had just the person to spring it upon, and it could easily be devastating even without direct, blatant punishment.
A person should be able to read everything their elected officials do. They should be able to read it in the morning newspaper, over breakfast or coffee, and understand it.
The only problem is, how do we word this properly? Or is there some other angle we can attack this from? Because the problem with what I just said is, anyone can bring in a retarded person and say "This person can't understand your Constitution; therefore, it's void.
And your language about preventing riders doesn't help a lot, because they can always be separate bills, the deals will just have to be more explicit. Nothing to stop them from telling each other (behind closed doors) that they will only let bill A pass if bill B does.
Why should it only apply to citizens?
Why should it apply to non-citizens?
As far as I'm concerned, it's not necessary that we grant rights to non-citizens. It's often the decent thing to do, yes, but really, if you want the rights of a US citizen, become a US citizen.
if it's their job to interpret the law, then my point is that the other branches of gov't should be able to ask them for an interpretation.
Perhaps, but consider:
Suppose the Supreme Court "interprets" the Constitution to still have the 3/5ths rule (or whatever it was) in effect, or something bizarre like that.
Or suppose they simply decide that any new laws, especially laws which attempt to reform the Supreme Court, are "unconstitutional" or "illegal". For that matter, you said "illegal" -- are you honestly saying a new law can't break any existing laws? We could never get rid of laws, then -- any new law attempting to replace them would be "illegal".
Do net labels such as Magnatune manage promotion to commuters using commercial FM radio?
Not that I know of. However, this isn't entirely their fault -- ClearChannel owns the airwaves, so they'd have to deal with the RIAA in order to do that.
Do they provide clearance services for songwriters to make sure that they haven't accidentally copied someone else's song (see the "My Sweet Lord" controversy)?
That I really don't know about. You could ask them.
However, even given that possible risk, and the lack of promotion, it still seems likely that a decent band could make far more money that they actually get to keep by selling their stuff online. Is the commercial FM radio promotion actually worth it?
Magnatune allows their music to be podcasted for free, and also gives free copies to music reviewers. I don't see any information about radio usage, but I imagine our local open source radio station may be able to get music from them without much effort.
That's right, we have an open source FM radio station. Check out their website, or flip to 100.1 FM if you're in the area. Most of their machines run Ubuntu, and they've completely replaced all use of any proprietary software, including Windows and ProTools. They use XMMS, Audacity, Ardour, and others, and it's actually relatively easy to get a show right now, although it will likely become harder as they get popular enough to actually have scheduling competition.
Windows is asking you for your network password before it starts mapping drives.
No, it's not. You're thinking of a separate password.
If you create multiple Windows 98 accounts, which is possible, and password-protect them -- also possible, you can still get into the default account by pressing esc at the login prompt. It then gives you full access to the system, because after all, FAT32 doesn't support permissions -- meaning if you really wanted to login as someone else (and gain access to their unique theme and Start Menu layout)... I don't remember if you can change their password directly, but you could certainly install a keylogger anywhere on the system.
As far as I know, this login prompt is completely useless. You may be right about it being intended to connect to an NT domain or something, but there's a different password prompt for when you're actually mounting the network drive (unless you tell it to save your password), and this seems like it's intended to be part of the whole "personalization" idea.
There is one legit hack I know of, that I'm surprised no one has mentioned it yet.
They have.
In one of the Matrix sequels,
Reloaded.
Trinity uses an SSH exploit to get into a powerstation control network. Apparently nmap is involved as well.
That wasn't explained. It was there on the screen if you were looking for it, and I thought it was pretty damned cool, but the fact remains that when any movie, including The Matrix, decides to explain something like that, they dumb it down so it can be understood by most intelligent non-geeks, then they dumb it down more, until by the time they think it's simple enough, it doesn't mean the same thing anymore. Or they give up and replace it with generic technobabble, with the intent that no one will understand it -- but it often ends up sounding close enough that geeks will understand it and find it ludicrous.
For instance, when freeing Neo in the first Matrix movie, they talk about a "carrier signal". I looked it up -- the carrier signal is that static-y sound that dialup internet uses to carry your data. That's right -- The Matrix runs on dialup -- and you can hear dialup-like sounds as the metal slime pours down his throat and he starts to wake up.
Before I knew that, it was simply technobabble, and I understood what the rest of the audience understood. "We're losing his <technobabble> signal!" I assumed that they knew what they were talking about.
But if they had just said "We're losing his signal," it wouldn't have sounded as stupid as it does today, when we're all connected via DSL, Cable, and Fiber, and none of these makes a phone-like sound.
I'd guess that your main advantage is training, not tech. Tech may eventually give you a significant advantage in this situation, but we're not there yet. Looks like body armor is the best advantage you've got.
I would like very much to have a torrent download as quickly and easily as a normal HTTP download. As it is, BitTorrent takes awhile to get started, and I have to upload as well for no good reason. Or at least, it'd be no good reason if multicast happened.
Besides, at first glance, it looks as though BT actually uses more resources than multicast would. Are routing tables so hard to manage between some 20 or 30 users that we'd rather deal with 200 or 300 open connections and 2-3 times the abound of bandwidth, if we're lucky?
so are unicast solutions to streaming media that prove that multicast is inessential.
What kind of solutions? Show me one that isn't wasting as much bandwidth as BitTorrent.
if you're on the same LAN as your other players, the network is already plenty fast for gaming even with unicast.
That much is true, although I believe some games use broadcast, which is why a LAN StarCraft game doesn't seem to need a server. But yes, for most games, I would rather have the positional updates and such be unicast. But I would rather the game content be multicast (think WoW patches).
Forget about multicast.
Give me something better, then. And BitTorrent isn't better.
I have trouble defending Java here, but I think PHP is actually worse. I don't know how, but it is.
The only reason for writing anything in PHP is because "worse is better". You write in PHP because everyone thinks (server side software == PHP), and every hosting provider has it... Everyone seems to think PHP is the P in LAMP...
But really, everything "good" that PHP has promotes bad design. Embedding in HTML: Bad design, please separate code from content. Variables don't have to be declared? Great, now you can spend three hours hunting down a typo in a variable. Really, can you name anything that PHP has that Java doesn't that is actually a good thing?
I'd prefer perl/python/ruby, and Java used to be the worst programming language I'd ever used. But PHP is even worse -- a lot worse.
And by the way, Java is a lot higher-level than PHP, for a lot longer. Java was object-oriented from the beginning, and this was sort of tacked on to PHP...
The one reason I use PHP is Drupal. Drupal manages to implement all kinds of ridiculously good design practices which technically aren't supported by the language. It does object-oriented programming without dealing with PHP objects, so it can run on older versions of PHP. It even does some things you'd expect to require LISP...
I've learned to give up the language holy wars unless I'm prepared to start from scratch. I don't like that Linux is written in C, but if I do kernel hacking, it's C. I don't like that Drupal is in PHP, but if I want to mess with a CMS, I use PHP. When I'm writing from scratch, at least for now, I use Perl, but that's just because it's the best that I know reasonably well.
I essentially agree with this, as a bill of rights. I'm not sure we need a constitution, just different levels of law. Some laws may be changed quickly and easily, some laws require a long ratification process. These should be essential rights, and should be as difficult to change as it currently is to get a constitutional amendment, but there should be other degrees of that between "constitutional" and "tax code tweak".
Regarding #5: I don't agree that these should be limited this way. Instead, we should limit them by age. The President has to be 35 or so, right? We should also put a maximum age. No senile congressmen.
Regarding #6, I'm not quite sure about the rationale behind this.
Regarding #7, the declaration should be formal, and we should know about it. "War on Terror" does not count. And what about special ops?
Regarding #9: Needs better language, but yes. It might be simpler to have all bills or articles, anywhere, be limited by some arbitrary measure like a word count. In any case, any document which has the force of law should be possible to understand by any person it affects. So, for instance, I should not have to be a lawyer to understand the EULA on my software, and it shouldn't take more than a couple of minutes to read through it. Creative Commons has a great example of a human-readable summary -- there's legalese to back it up, but the summary is understandable.
I'd call this the anti-legalese right.
#10 should only apply to citizens -- and it should be possible to abridge this without convicting someone. If you're holding a suspect, obviously they should not be allowed to have a sword in their jail cell.
#11: Careful here. The Supreme Court currently has an insane amount of power to arbitrarily interpret stuff.
It may be time for us to not only re-examine our laws and constitution, but our whole system of government. Of course we should be a republic, but that covers a lot of ground. It doesn't help at all to have these three branches of government if, at any given time, every single branch is corrupt, and barely keeping the corruption of the other branches from bringing down the whole tree.
When you do it this way, it's possible you'll find yourself re-inventing the wheel a bunch. The trick is to choose a system that is flexible enough and portable enough that you're alright with having your application married to that system.
After all, we all have to choose a programming language to start with.
So, for instance, if your concern is that tying yourself to a graphics library, even a nice cross-platform one like wxwindows or opengl, will make your program forever a GUI program, you might design your code for a system that has both wx and ncurses backends. You might write a configuration management system that can be operated through config files, commandline, curses, or wx, and have your program tied heavily to that system.
But yes, we want to be portable. The important thing to remember is that the only way to make an idea portable from an application is to wrap that idea in a library. I prefer to have the majority of the application logic be in CPAN modules, and have the app itself be around a few hundred lines of Perl code. In fact, more than a hundred lines or so, and I probably haven't abstracted enough.
A simple example - office furniture could likely be described algorithmically, but doing so would be a huge waste of programmer time. An artist can create a large library of furniture for a higher-level algorithm to place in office buildings, and the furniture can be tagged with logical descriptions for these algorithms to use.
True enough. I do believe that there could be tools which would help with this kind of thing, though. For instance, an artist (not a programmer) could create a generic-looking chair by starting with something that kind of looks like a chair, and storing the modifications between that and a chair. It's possible that storing a table + instructions to generate a chair from a table is cheaper than storing the chair and table separately -- but either way, the software could choose the cheaper one automatically.
Probably, you'd want both, so that people who have already downloaded the chair can download the chair-to-table "patch", and vice versa, assuming the "patch" is smaller than the resulting object. Such a "patch" could be a set of steps you'd take in a modelling program to change one into the other, so it could essentially be a recording, but I'm thinking you'd want much more intelligent tools -- given any two objects, try to generate a patch between them.
Furthermore, the artist would define a generic chair -- for instance, an office chair -- wheels, armrests, could be cloth or leather -- and allow some permutations. It'd take some experimentation, and it'd probably take more time than just creating a bunch of prefab chairs, at least until the artist really got a knack for how to do it this way. But you can imagine -- some office chairs are lower, some are higher, and depending on how profitable the business, some might be missing an arm or two...
The idea is, at least: Create all chairs by starting with other chairs, or with a generic chair, and for that matter, create all objects by starting with other objects. Ideally, even if you have a finite set of chairs, the system should be able to distribute any one of those chairs as a "patch" to another one, and store them as the difference between that chair and some "base" chair.
So, you have prefab chairs, just like you describe, but they take less space, and your tools to generate small variations may not make it into the game (at least not for chairs), but they may mean your artist has to do a lot less work to generate a lot more kinds of chairs.
And of course, do this for your textures, also. You might generate some textures algorithmically, or allow some kind of materials to have algorithmic deviations -- my chair is made of a cloth that looks like a very fine checkerboard (probably the weave), but it's a little wavy, partly dependent on the way it's wrapped around the chair, partly sheer randomness. That way, the same exact chair model your artist created above could look slightly different in places that count for realism, but not in any way that throws it completely off.
And, it would certainly look much better than anything an algorithm could generate.
I'm not convinced of that, but the payoff of doing the chair algorithmically -- actually having one true "chair formula", rather than what I've described above -- wouldn't be worth the insane amount of effort it'd take.
The lower-level the algorithm, the more generic looking the final outcome, but the broader the range of possible variants. Higher-level algorithms have the advantage of creating more unique and creative-looking results, but typically draw on pre-fabricated components in order to achieve this, and thus have a limited number of permutations.
I see what you're saying here, and I partly agree. In the same way, a very generic "wood" texture could be re-used for all the wood in the game, but then it would all look the same.
I'm just thinking you want permutations everywhere, even when they don't seem to be doing
Just think one day the artists and the fans might connect directly on the internet with no middle man in between to screw the artists and sue the fans.
There are a few artists that dothat, but really, what we need is a middleman (or two) that doesn't screw the artists and sue the fans. Take a good, hard look at MagnaTune -- even if you pay the lowest possible price ($5/album), 50% of it goes straight to the artist (and $2.50 is more than the RIAA will pay them), and you are legally allowed to share it with 5 people. You get to listen to the whole thing in mp3 before you buy, and you can download the whole thing in FLAC once you do pay for it. (FLAC is lossless, so you can burn it straight to CD as if you'd bought it at the music store, only it probably cost you less, and you know the artist got more.)
Start supporting these guys now. They might not have the bands you like now, but you'll find music to like, and you won't be supporting the RIAA. Get even your non-techie friends doing this, and soon enough, we could actually make the RIAA irrelevant.
Just checked my spam folder since yesterday: 151 spams. This is just overnight, and since probably 10 PM, so I'm probably going to get quite a few more.
Also checked the "unsure" folder: 10 more spams. I actually read through these ones to make sure, whereas I generally only skim the subject lines in the actual spam folder.
I get maybe 1-5 spams that slip through the filter completely, on occasion. I drag those, along with most of "unsure", to the retrain folder. I also occasionally (maybe once a week) find a legitimate email in the "unsure" folder, and I can't remember the last time I found any false positives in the "spam" folder -- so I probably could actually redirect "spam" to/dev/null.
I only read my email from Thunderbird, on Linux and OS X, so I'm not afraid of viruses. I simply train my spamfilter on them. Thus, it would annoy me if the "birthday cards", as well as actual viruses, were silently dropped.
I also run this on my own server, and I'm the only one who gets email through it. My spamfilter of choice is BogoFilter, and I run BincIMAP as my only access other than ssh. I have a custom script that uses inotify to automatically retrain mail and re-deliver it as soon as I drop it in the retrain boxes -- thus, as soon as I mark something as "innocent", it gets retrained and re-sorted into my mailing list folders and such.
I haven't deleted any email completely in something like six months, just thrown it in the trash, figuring it's nice to be able to search it. And yet, even though I'm using maildir to store this, I'm also running the server on a very stable Reiser4, which makes it very space efficient. The home directory of my email user -- which includes all of the mail (and attachments), the training information, the cron jobs, and so on -- is 684 megs, and it takes it about a second to do a "du -sh" on it. That's almost 31 thousand files.
Gmail may not store things as efficiently, but still, they give you over 2 gigs. So, like I said, you really shouldn't have to drop stuff on the floor.
It's possible that I don't get much legit mail, but I'm on several mailing lists, and I do have friends and family email me occasionally, and all of this is handled properly. Worst case is someone just sends me a YouTube URL and ends up in the unsure folder.
Now, this would be difficult to get off the ground for someone who gets thousands of spams per day -- bogofilter must be trained first, so even if you train it with the spamassassin repositories, or borrow my training database, it's going to be awhile before it's absolutely perfect. But mine is pretty close now, and I can definitely say it was worth it.
an entire world could be nothing more than a transmitted seed number (plus the content to build it from).
What do you mean by "the content to build it from"?
I'll refer you to.kkreiger, an FPS in 96 kilobytes, as well as many of the other fun things from demoscene -- if by "content" you mean "reusable textures/materials and models", realize that those, too, can be generated.
The only content you should have to send is where you're deliberately changing a world generated from a scene. For instance, generate a city, then tweak a building. And almost any "tweak" you make could be a seed + a re-usable algorithm. You could have a "person" algorithm, which includes a hand algorithm, an arm algorithm, and so on...
Sequenced music, generated animations...
Of course, at a certain point, you have to decide that it's worth it to just let people devevlop content the way they're used to, and figure that it's OK to sacrifice a meg or two for a 100% hand-crafted final boss or something. But even here, you'd want tools to help it be more procedural -- use curves instead of triangles (and generate the triangles from the curves), use texture generation algorithms rather than re-using textures (every plank of wood should look different)...
I'm actually wanting to get deep into this stuff sometime next year. Having content look every bit as good, but fit inside a few kilobytes, opens up a few more possibilities that I'd like to keep to myself for now, but are probably obvious to anyone ambitious enough.
So let's say you're about to implement MathML in Word and find that MathML doesn't support a feature that Word has, like change tracking. Do you extend MathML and to support your feature and risk being called nasty names for making your implementation incompatible with anyone else's?
Well, it could be that you could do change tracking in such a way that it doesn't break said MathML, or provide simple XSLT to extract the MathML from your MathML + change tracking.
And, if you have to, you implement your custom changes in a very straightforward way and submit them to the MathML group to be included as a standard.
I'd say LGPL or freer. So long as it doesn't restrict what links to it, we should be fine.
But even this doesn't help much. It does help in that the reference implementation can be the spec in areas where the spec is wrong or incomplete. But if it's a full app, it could be just as useless as the specs (OpenOffice is HUGE), and if it's a library, it's problematic because now you've got a whole new spec to describe: how to talk to the library.
I actually was all ready to embrace MS's Office format as something to read from directly to get stuff out of word. I was thinking some simple XSLT or XML parsing.
I was wrong. Actually, OfficeXML is not "open" at all. It is to XML what BinHex is to ASCII -- sure as hell not human readable.
OpenDocument, I can actually read and understand, as a human, without documentation. OfficeXML, I'd be lucky to write a program that could decypher it with documentation. And these are on simple text documents -- I don't care what else the format supports, but it should at least look good for unformatted test text documents.
There is considerable evidence however that committees for single industry wide open standards hold back technical progress. This is fine if the standard is simple and fexible enough not to warrant much change (such as comms protocols for example), but for anything else can be deadly (CSS for example).
I'm really sorry you believe this, because open standards do far more for technical progress than they do against it.
The Internet would not have gotten to where it is today without standards. In fact, the #1 thing about CSS that limits web development is not what's missing from the spec, but what's completely broken in various implementations because someone thought they could do better than the spec.
Even people working for MS realize that the IE CSS Bug Suite is a Bad Thing.
To my knowledge it has never been proved that MS withheld information about an API for anti-competitive purposes.
Not any more than it's been "proved" that things fall when you drop them. We can only go on our observation.
Or do you have another reason MS has withheld information about APIs or file formats? Why they didn't even make their half-assed "standard" attempt until Mass started showing interest in OpenDocument?
This is the age of XML - we should embrace multiple formats, implementations and the conversion between each, rather than stifling progress.
Ooh! A buzzword and a false dichotomy!
It does not stifle progress to standardize on something like OpenDocument. It stifles progress to expect every word processor out there to have to read and understand a dozen different kinds of XML, some of which (OfficeXML) are retardedly impossible to read.
Throwing <xml> tags around unreadable crap doesn't hide the fact that it's inreadable crap. Stop propagating these myths that XML is some magic bullet that automatically makes everything transparent and understandable, and a breeze to write a parser for. Show me one thing OfficeXML can do that OpenDocument can't do, and we'll talk. Until then, you're a poorly disguised Microsoft lackey.
Although some minor functionality in our world editor is automated / algorithmic, too much still requires explicit hand-crafting by a level designer.
Build it into your engine, not your world editor. Of course both will have to support it, but the filler content should be generated on the end-user's machine, allowing it all to scale up more.
And it's not all "filler". Do it for your main content, too. Hire people who make interesting WinAmp visualisations and such, get shader people, have them think of ways to, for instance, generate a wood grain fractal, instead of having 16 different textures for a wooden plank.
I'd actually love to be developing tools like this, and I'm glad to hear it's needed in the real world. Unfortunately, I flunked out of college and have no experience with this, so I'm working in web development and trying to find spare time to play with it.
Now it's like this: drop spam as cheaply as possible (and lose half the legit email) or just give up on email entirely.
I'm sorry your spamfilters suck so much that you have to do that.
Mine is actually quite good. I don't lose mail, and I don't have to spend much of my day dealing with spam. However, if I wanted to do it more cheaply (save some bandwidth), I could start building a blacklist based on what sends me spam.
Modern email is an extremely unreliable communication method. Get used to it.
Actually, Earthlink is just that bad, and I wish people wouldn't get used to it, because then scum like Earthlink wouldn't be allowed to exist. We have solved the spam problem already.
Slashdot Mirror
Alright, but it certainly isn't causing the poverty. And do you really think that people would fix the problem if they weren't taxed so much?
Suppose you didn't have to pay a dime in taxes next year. How much of that would you spend dealing with poverty?
Now, suppose you got a raise to a job that pays twice as much next year. Same question.
Fact is, if people aren't solving the problem now, taxing them less isn't going to fix it.
What about them?
You make an interesting point, but don't you think there would be far more of us dying from Dr. Bob-Earl-Smith if we didn't have a decent amount of regulations? You may want to fix the regulations, but I hope you don't actually want to remove them.
No disagreement here. However, wouldn't you rather your capslock GOVERNMENT would talk about this stupidity, rather than the "issue" of "obscenity"?
Well, since you asked: Better free public transportation, tighter restrictions on licenses -- maybe the truck driver shouldn't have been driving?
I don't have all the answers, but wouldn't you rather they actually talk about this, and take this on? It's not like it's any more hopeless than the "war on terror", and terrorism actually costs us far fewer lives.
What. The. Fuck?
Do the words "private school" mean anything to you?
So you're saying that in a free market, a kid would walk into school and get paid? I mean, you do realize that public school is free, right?
I mean, that's the whole point of public school in the first place. It's free.
We might also have more success with the Linux thing if we stop calling it Linux. We might also have more success with the Free Software thing if we stopped calling ourselves hackers.
That's actually most of what I want. The problem is, the state currently calls such unions "marriages". You strike the word "marriage" from every law, and then we'll talk about calling it something other than gay marriage.
Or, the religious people can stop getting in such a righteous fury over this kind of language, and accept that "marriage" as the state is concerned doesn't dictate what "marriage" means to their religion. They can allow gay marriage to be legal, and still preach that it's not really marriage.
Except that this really isn't usually their goal. One I talked to recently basically said that marriage is between a man and a woman, and he wouldn't mind if the gays up the street got a civil union and pretended it was a marriage -- but he also felt that marriage should be recognized by the state, and should receive special benefits not given to civil unions. So, the whole argument about language was just a smoke screen.
And what's wrong with that?
No, really, I'm waiting for your answer, because chances are, anything you have to say could be made to work as part of the definition of "civil union".
That is because marketing, for many years, has been institutionalized deception when it isn't outright lying.
And you know it's a beer commercial, and it's also ludicrous enough that you know it's a joke. How is this related to something plausible that deliberately tries to hide the fact that it's an advertisement?
How is that similar? Some drug ads never say what they do. It's at least obvious that they're drug ads -- or at least, that they're ads for something!
You'd have a point if the drug ads pretended to be documentaries, but they don't. Thank you for reminding us that even drug companies are more ethical than Sony.
And based on your ethics, I strongly suspect that you work for Sony. If you do, go listen to the Bill Hicks rant, and kill yourself.
While you have a valid point, and one I've been trying to make, you also have a bad example. My Mac laptop is missing a lot of features I take for granted on Linux, even ones with decently simple interfaces. Just to start with, there's no package manager.
Funny you should mention that. The Gimp, in its default configuration, reminds me very much of MSPaint. It has a few more tools, but they all look the same as the Paint ones, and it's easy to find the ones you need.
.bmp), and it will almost certainly support whatever image type you specify.
There's also tons of complexity hidden in menus, but if you know how to use Paint, chances are you already know how to use the Gimp, except you can now type in an extension (other than
Thus, cropping.
This is an important point -- you don't need a "crop" feature, because you can easily acheive it from the other features. I'm not saying there shouldn't be a crop button, but this is an example of how you can make a program simpler and yet still have it be as powerful.
From what I remember, you're just wrong. I could easily click inside the window (rectangular selection), then drag it outside the window, and the scrollbars would follow. You can resize the window or the image, and you can do it all with simple mouse movements, though it may not be that intuitive.
Also, you can zoom in and out, and the tools work at all zoom levels.
And yet, it can still be made simpler, and he's completely missing the point of "does one thing and does it well".
Consider: Both gzip and bzip2 do one thing, and do it well. Yet, by combining these with tools like dd, cp, tar, nc, ntfsclone, or less, I can create all kinds of features. Consider: tar has support for gzip and bzip2, but it does it by calling the gzip and bzip2 programs, and it's the exception.
This guy would rather have every single commandline utility link against zlib.
Now, that doesn't prevent you from making a big, bloaty app, and still having it be "simple". You just have to make individual features of that app combine in the way that gzip and tar combine. A simple example is word processing -- notice how there's a bold, italic, and underline button, but no bolditalic button. If you want bolditalic, you click the bold button, then the italic button. If someone wanted to spend a lot of effort on design, they could probably make a very simple word processor, with most of the functionality organized such that it feels like you're using AbiWord or TextEdit, but it's actually more powerful than OpenOffice.
And he misses the point with Google. Consider that most of the features on Yahoo's homepage are available through Google's services, they just don't clutter up the homepage.
Given all of this, your point:
only applies if, as you say:
And you could even have "fewer features" and yet have it be more powerful. LISP is one of the most powerful languages we have, and it also has incredibly simple syntax and builtin stuff. If you need a missing feature, you implement it yourself, naturally -- as naturally as I implement a bolditalic button by clicking the bold and italic buttons.
Your way is almost as bad. Mis-interpret just about every finable offense ever, and simply avoid the fines.
The trick is to come up with some good, solid language for obviousness, that itself passes an obviousness test. I'm not sure we're capable of that as a species yet, but it might help to have that as a goal. (Or find a completely different approach that would achieve the same result.)
Still a decent goal.
The trouble is, even if we all had a perfect understanding of the 18th century mind, and so we knew intimately how the Constitution was intended to be interpreted, there are parts of it that are simply obsolete -- that 3/5ths rule is a perfect example.
Rewriting from scratch is a scary prospect, but it may be easier than trying to reform a document that's over two hundred yeras old.
What we need is a way to not cripple legitimate things, like SWAT teams, and yet still prevent a police state. Having a civilian be Commander in Chief is a step in the right direction.
After all, if the government is really going to be tyrannical that way, they can simply declare martial law, or ammend this limitation away -- so we need another solution anyway. However, SWAT teams do seem legitimate to me -- although if I go back and read your original proposal, it might cover that (let us buy all the equipment the SWAT teams use).
Not good enough. If the law isn't clear enough, it should be void, which means I get to keep my fully-automatic weapon until they fix the law.
If your rule is implemented, politicians would pass laws that no one wants except their own special interests, but that are confusing enough that no one knows just what they do until they're directly affected. It's not good enough to just say that you can't be punished -- the law could've sat around for years until they had just the person to spring it upon, and it could easily be devastating even without direct, blatant punishment.
A person should be able to read everything their elected officials do. They should be able to read it in the morning newspaper, over breakfast or coffee, and understand it.
The only problem is, how do we word this properly? Or is there some other angle we can attack this from? Because the problem with what I just said is, anyone can bring in a retarded person and say "This person can't understand your Constitution; therefore, it's void.
And your language about preventing riders doesn't help a lot, because they can always be separate bills, the deals will just have to be more explicit. Nothing to stop them from telling each other (behind closed doors) that they will only let bill A pass if bill B does.
Why should it apply to non-citizens?
As far as I'm concerned, it's not necessary that we grant rights to non-citizens. It's often the decent thing to do, yes, but really, if you want the rights of a US citizen, become a US citizen.
Perhaps, but consider:
Suppose the Supreme Court "interprets" the Constitution to still have the 3/5ths rule (or whatever it was) in effect, or something bizarre like that.
Or suppose they simply decide that any new laws, especially laws which attempt to reform the Supreme Court, are "unconstitutional" or "illegal". For that matter, you said "illegal" -- are you honestly saying a new law can't break any existing laws? We could never get rid of laws, then -- any new law attempting to replace them would be "illegal".
Not that I know of. However, this isn't entirely their fault -- ClearChannel owns the airwaves, so they'd have to deal with the RIAA in order to do that.
That I really don't know about. You could ask them.
However, even given that possible risk, and the lack of promotion, it still seems likely that a decent band could make far more money that they actually get to keep by selling their stuff online. Is the commercial FM radio promotion actually worth it?
Magnatune allows their music to be podcasted for free, and also gives free copies to music reviewers. I don't see any information about radio usage, but I imagine our local open source radio station may be able to get music from them without much effort.
That's right, we have an open source FM radio station. Check out their website, or flip to 100.1 FM if you're in the area. Most of their machines run Ubuntu, and they've completely replaced all use of any proprietary software, including Windows and ProTools. They use XMMS, Audacity, Ardour, and others, and it's actually relatively easy to get a show right now, although it will likely become harder as they get popular enough to actually have scheduling competition.
Well, I suppose you're right. It's cracking.
No, it's not. You're thinking of a separate password.
If you create multiple Windows 98 accounts, which is possible, and password-protect them -- also possible, you can still get into the default account by pressing esc at the login prompt. It then gives you full access to the system, because after all, FAT32 doesn't support permissions -- meaning if you really wanted to login as someone else (and gain access to their unique theme and Start Menu layout)... I don't remember if you can change their password directly, but you could certainly install a keylogger anywhere on the system.
As far as I know, this login prompt is completely useless. You may be right about it being intended to connect to an NT domain or something, but there's a different password prompt for when you're actually mounting the network drive (unless you tell it to save your password), and this seems like it's intended to be part of the whole "personalization" idea.
They have.
Reloaded.
That wasn't explained. It was there on the screen if you were looking for it, and I thought it was pretty damned cool, but the fact remains that when any movie, including The Matrix, decides to explain something like that, they dumb it down so it can be understood by most intelligent non-geeks, then they dumb it down more, until by the time they think it's simple enough, it doesn't mean the same thing anymore. Or they give up and replace it with generic technobabble, with the intent that no one will understand it -- but it often ends up sounding close enough that geeks will understand it and find it ludicrous.
For instance, when freeing Neo in the first Matrix movie, they talk about a "carrier signal". I looked it up -- the carrier signal is that static-y sound that dialup internet uses to carry your data. That's right -- The Matrix runs on dialup -- and you can hear dialup-like sounds as the metal slime pours down his throat and he starts to wake up.
Before I knew that, it was simply technobabble, and I understood what the rest of the audience understood. "We're losing his <technobabble> signal!" I assumed that they knew what they were talking about.
But if they had just said "We're losing his signal," it wouldn't have sounded as stupid as it does today, when we're all connected via DSL, Cable, and Fiber, and none of these makes a phone-like sound.
I'd guess that your main advantage is training, not tech. Tech may eventually give you a significant advantage in this situation, but we're not there yet. Looks like body armor is the best advantage you've got.
I would like very much to have a torrent download as quickly and easily as a normal HTTP download. As it is, BitTorrent takes awhile to get started, and I have to upload as well for no good reason. Or at least, it'd be no good reason if multicast happened.
Besides, at first glance, it looks as though BT actually uses more resources than multicast would. Are routing tables so hard to manage between some 20 or 30 users that we'd rather deal with 200 or 300 open connections and 2-3 times the abound of bandwidth, if we're lucky?
What kind of solutions? Show me one that isn't wasting as much bandwidth as BitTorrent.
That much is true, although I believe some games use broadcast, which is why a LAN StarCraft game doesn't seem to need a server. But yes, for most games, I would rather have the positional updates and such be unicast. But I would rather the game content be multicast (think WoW patches).
Give me something better, then. And BitTorrent isn't better.
I have trouble defending Java here, but I think PHP is actually worse. I don't know how, but it is.
The only reason for writing anything in PHP is because "worse is better". You write in PHP because everyone thinks (server side software == PHP), and every hosting provider has it... Everyone seems to think PHP is the P in LAMP...
But really, everything "good" that PHP has promotes bad design. Embedding in HTML: Bad design, please separate code from content. Variables don't have to be declared? Great, now you can spend three hours hunting down a typo in a variable. Really, can you name anything that PHP has that Java doesn't that is actually a good thing?
I'd prefer perl/python/ruby, and Java used to be the worst programming language I'd ever used. But PHP is even worse -- a lot worse.
And by the way, Java is a lot higher-level than PHP, for a lot longer. Java was object-oriented from the beginning, and this was sort of tacked on to PHP...
The one reason I use PHP is Drupal. Drupal manages to implement all kinds of ridiculously good design practices which technically aren't supported by the language. It does object-oriented programming without dealing with PHP objects, so it can run on older versions of PHP. It even does some things you'd expect to require LISP...
I've learned to give up the language holy wars unless I'm prepared to start from scratch. I don't like that Linux is written in C, but if I do kernel hacking, it's C. I don't like that Drupal is in PHP, but if I want to mess with a CMS, I use PHP. When I'm writing from scratch, at least for now, I use Perl, but that's just because it's the best that I know reasonably well.
I essentially agree with this, as a bill of rights. I'm not sure we need a constitution, just different levels of law. Some laws may be changed quickly and easily, some laws require a long ratification process. These should be essential rights, and should be as difficult to change as it currently is to get a constitutional amendment, but there should be other degrees of that between "constitutional" and "tax code tweak".
Regarding #5: I don't agree that these should be limited this way. Instead, we should limit them by age. The President has to be 35 or so, right? We should also put a maximum age. No senile congressmen.
Regarding #6, I'm not quite sure about the rationale behind this.
Regarding #7, the declaration should be formal, and we should know about it. "War on Terror" does not count. And what about special ops?
Regarding #9: Needs better language, but yes. It might be simpler to have all bills or articles, anywhere, be limited by some arbitrary measure like a word count. In any case, any document which has the force of law should be possible to understand by any person it affects. So, for instance, I should not have to be a lawyer to understand the EULA on my software, and it shouldn't take more than a couple of minutes to read through it. Creative Commons has a great example of a human-readable summary -- there's legalese to back it up, but the summary is understandable.
I'd call this the anti-legalese right.
#10 should only apply to citizens -- and it should be possible to abridge this without convicting someone. If you're holding a suspect, obviously they should not be allowed to have a sword in their jail cell.
#11: Careful here. The Supreme Court currently has an insane amount of power to arbitrarily interpret stuff.
It may be time for us to not only re-examine our laws and constitution, but our whole system of government. Of course we should be a republic, but that covers a lot of ground. It doesn't help at all to have these three branches of government if, at any given time, every single branch is corrupt, and barely keeping the corruption of the other branches from bringing down the whole tree.
When you do it this way, it's possible you'll find yourself re-inventing the wheel a bunch. The trick is to choose a system that is flexible enough and portable enough that you're alright with having your application married to that system.
After all, we all have to choose a programming language to start with.
So, for instance, if your concern is that tying yourself to a graphics library, even a nice cross-platform one like wxwindows or opengl, will make your program forever a GUI program, you might design your code for a system that has both wx and ncurses backends. You might write a configuration management system that can be operated through config files, commandline, curses, or wx, and have your program tied heavily to that system.
But yes, we want to be portable. The important thing to remember is that the only way to make an idea portable from an application is to wrap that idea in a library. I prefer to have the majority of the application logic be in CPAN modules, and have the app itself be around a few hundred lines of Perl code. In fact, more than a hundred lines or so, and I probably haven't abstracted enough.
True enough. I do believe that there could be tools which would help with this kind of thing, though. For instance, an artist (not a programmer) could create a generic-looking chair by starting with something that kind of looks like a chair, and storing the modifications between that and a chair. It's possible that storing a table + instructions to generate a chair from a table is cheaper than storing the chair and table separately -- but either way, the software could choose the cheaper one automatically.
Probably, you'd want both, so that people who have already downloaded the chair can download the chair-to-table "patch", and vice versa, assuming the "patch" is smaller than the resulting object. Such a "patch" could be a set of steps you'd take in a modelling program to change one into the other, so it could essentially be a recording, but I'm thinking you'd want much more intelligent tools -- given any two objects, try to generate a patch between them.
Furthermore, the artist would define a generic chair -- for instance, an office chair -- wheels, armrests, could be cloth or leather -- and allow some permutations. It'd take some experimentation, and it'd probably take more time than just creating a bunch of prefab chairs, at least until the artist really got a knack for how to do it this way. But you can imagine -- some office chairs are lower, some are higher, and depending on how profitable the business, some might be missing an arm or two...
The idea is, at least: Create all chairs by starting with other chairs, or with a generic chair, and for that matter, create all objects by starting with other objects. Ideally, even if you have a finite set of chairs, the system should be able to distribute any one of those chairs as a "patch" to another one, and store them as the difference between that chair and some "base" chair.
So, you have prefab chairs, just like you describe, but they take less space, and your tools to generate small variations may not make it into the game (at least not for chairs), but they may mean your artist has to do a lot less work to generate a lot more kinds of chairs.
And of course, do this for your textures, also. You might generate some textures algorithmically, or allow some kind of materials to have algorithmic deviations -- my chair is made of a cloth that looks like a very fine checkerboard (probably the weave), but it's a little wavy, partly dependent on the way it's wrapped around the chair, partly sheer randomness. That way, the same exact chair model your artist created above could look slightly different in places that count for realism, but not in any way that throws it completely off.
I'm not convinced of that, but the payoff of doing the chair algorithmically -- actually having one true "chair formula", rather than what I've described above -- wouldn't be worth the insane amount of effort it'd take.
I see what you're saying here, and I partly agree. In the same way, a very generic "wood" texture could be re-used for all the wood in the game, but then it would all look the same.
I'm just thinking you want permutations everywhere, even when they don't seem to be doing
There are a few artists that do that, but really, what we need is a middleman (or two) that doesn't screw the artists and sue the fans. Take a good, hard look at MagnaTune -- even if you pay the lowest possible price ($5/album), 50% of it goes straight to the artist (and $2.50 is more than the RIAA will pay them), and you are legally allowed to share it with 5 people. You get to listen to the whole thing in mp3 before you buy, and you can download the whole thing in FLAC once you do pay for it. (FLAC is lossless, so you can burn it straight to CD as if you'd bought it at the music store, only it probably cost you less, and you know the artist got more.)
Start supporting these guys now. They might not have the bands you like now, but you'll find music to like, and you won't be supporting the RIAA. Get even your non-techie friends doing this, and soon enough, we could actually make the RIAA irrelevant.
Just checked my spam folder since yesterday: 151 spams. This is just overnight, and since probably 10 PM, so I'm probably going to get quite a few more.
/dev/null.
Also checked the "unsure" folder: 10 more spams. I actually read through these ones to make sure, whereas I generally only skim the subject lines in the actual spam folder.
I get maybe 1-5 spams that slip through the filter completely, on occasion. I drag those, along with most of "unsure", to the retrain folder. I also occasionally (maybe once a week) find a legitimate email in the "unsure" folder, and I can't remember the last time I found any false positives in the "spam" folder -- so I probably could actually redirect "spam" to
I only read my email from Thunderbird, on Linux and OS X, so I'm not afraid of viruses. I simply train my spamfilter on them. Thus, it would annoy me if the "birthday cards", as well as actual viruses, were silently dropped.
I also run this on my own server, and I'm the only one who gets email through it. My spamfilter of choice is BogoFilter, and I run BincIMAP as my only access other than ssh. I have a custom script that uses inotify to automatically retrain mail and re-deliver it as soon as I drop it in the retrain boxes -- thus, as soon as I mark something as "innocent", it gets retrained and re-sorted into my mailing list folders and such.
I haven't deleted any email completely in something like six months, just thrown it in the trash, figuring it's nice to be able to search it. And yet, even though I'm using maildir to store this, I'm also running the server on a very stable Reiser4, which makes it very space efficient. The home directory of my email user -- which includes all of the mail (and attachments), the training information, the cron jobs, and so on -- is 684 megs, and it takes it about a second to do a "du -sh" on it. That's almost 31 thousand files.
Gmail may not store things as efficiently, but still, they give you over 2 gigs. So, like I said, you really shouldn't have to drop stuff on the floor.
It's possible that I don't get much legit mail, but I'm on several mailing lists, and I do have friends and family email me occasionally, and all of this is handled properly. Worst case is someone just sends me a YouTube URL and ends up in the unsure folder.
Now, this would be difficult to get off the ground for someone who gets thousands of spams per day -- bogofilter must be trained first, so even if you train it with the spamassassin repositories, or borrow my training database, it's going to be awhile before it's absolutely perfect. But mine is pretty close now, and I can definitely say it was worth it.
What do you mean by "the content to build it from"?
I'll refer you to .kkreiger, an FPS in 96 kilobytes, as well as many of the other fun things from demoscene -- if by "content" you mean "reusable textures/materials and models", realize that those, too, can be generated.
The only content you should have to send is where you're deliberately changing a world generated from a scene. For instance, generate a city, then tweak a building. And almost any "tweak" you make could be a seed + a re-usable algorithm. You could have a "person" algorithm, which includes a hand algorithm, an arm algorithm, and so on...
Sequenced music, generated animations...
Of course, at a certain point, you have to decide that it's worth it to just let people devevlop content the way they're used to, and figure that it's OK to sacrifice a meg or two for a 100% hand-crafted final boss or something. But even here, you'd want tools to help it be more procedural -- use curves instead of triangles (and generate the triangles from the curves), use texture generation algorithms rather than re-using textures (every plank of wood should look different)...
I'm actually wanting to get deep into this stuff sometime next year. Having content look every bit as good, but fit inside a few kilobytes, opens up a few more possibilities that I'd like to keep to myself for now, but are probably obvious to anyone ambitious enough.
Well, it could be that you could do change tracking in such a way that it doesn't break said MathML, or provide simple XSLT to extract the MathML from your MathML + change tracking.
And, if you have to, you implement your custom changes in a very straightforward way and submit them to the MathML group to be included as a standard.
I'd say LGPL or freer. So long as it doesn't restrict what links to it, we should be fine.
But even this doesn't help much. It does help in that the reference implementation can be the spec in areas where the spec is wrong or incomplete. But if it's a full app, it could be just as useless as the specs (OpenOffice is HUGE), and if it's a library, it's problematic because now you've got a whole new spec to describe: how to talk to the library.
I actually was all ready to embrace MS's Office format as something to read from directly to get stuff out of word. I was thinking some simple XSLT or XML parsing.
I was wrong. Actually, OfficeXML is not "open" at all. It is to XML what BinHex is to ASCII -- sure as hell not human readable.
OpenDocument, I can actually read and understand, as a human, without documentation. OfficeXML, I'd be lucky to write a program that could decypher it with documentation. And these are on simple text documents -- I don't care what else the format supports, but it should at least look good for unformatted test text documents.
I'm really sorry you believe this, because open standards do far more for technical progress than they do against it.
The Internet would not have gotten to where it is today without standards. In fact, the #1 thing about CSS that limits web development is not what's missing from the spec, but what's completely broken in various implementations because someone thought they could do better than the spec.
Even people working for MS realize that the IE CSS Bug Suite is a Bad Thing.
Not any more than it's been "proved" that things fall when you drop them. We can only go on our observation.
Or do you have another reason MS has withheld information about APIs or file formats? Why they didn't even make their half-assed "standard" attempt until Mass started showing interest in OpenDocument?
Ooh! A buzzword and a false dichotomy!
It does not stifle progress to standardize on something like OpenDocument. It stifles progress to expect every word processor out there to have to read and understand a dozen different kinds of XML, some of which (OfficeXML) are retardedly impossible to read.
Throwing <xml> tags around unreadable crap doesn't hide the fact that it's inreadable crap. Stop propagating these myths that XML is some magic bullet that automatically makes everything transparent and understandable, and a breeze to write a parser for. Show me one thing OfficeXML can do that OpenDocument can't do, and we'll talk. Until then, you're a poorly disguised Microsoft lackey.
Build it into your engine, not your world editor. Of course both will have to support it, but the filler content should be generated on the end-user's machine, allowing it all to scale up more.
And it's not all "filler". Do it for your main content, too. Hire people who make interesting WinAmp visualisations and such, get shader people, have them think of ways to, for instance, generate a wood grain fractal, instead of having 16 different textures for a wooden plank.
I'd actually love to be developing tools like this, and I'm glad to hear it's needed in the real world. Unfortunately, I flunked out of college and have no experience with this, so I'm working in web development and trying to find spare time to play with it.
I'm sorry your spamfilters suck so much that you have to do that.
Mine is actually quite good. I don't lose mail, and I don't have to spend much of my day dealing with spam. However, if I wanted to do it more cheaply (save some bandwidth), I could start building a blacklist based on what sends me spam.
Actually, Earthlink is just that bad, and I wish people wouldn't get used to it, because then scum like Earthlink wouldn't be allowed to exist. We have solved the spam problem already.