It'd be great now if he offered an implementation which we coudl all use.
I think a progressive, ever-going implementation is best. I also think its best to filter based on headers first, and not download any spam (to save bandwidth) and then filter based on message content (for the messages downloaded) and move any spam to a spam folder.
Then the user simply looks at the spam folder and looks for false-positives, and marks them as "legit". Then the Bayesian filter recalculates.
Same thing for false negatives, and for the messages not downloaded. The user can look at the headers of the messages not downloaded and say if they're spam. Then the Bayesian filter recalculates.
Another good thing to do is to give a "password" to your friends for them to put in headers sent to you. I.e., 13y4890dshfpljk2134y9073254y32p9ur. Any message with that in the header would be given a 0% probability of being spam, as only those you gave that to would know to put it in the header. Should it become compromised, you can change it (or just don't give it to people who might compromise it).
Back to the Bayesian filter, another good thing might be to have varying levels of "spam". I.e., if something is almost certailny spam (i.e., 99.99999999% likely to be spam, as would a message with the header "Get fucked for free and make lots of $$$$$"), it would be placed in a DEFINATELY SPAM FOLDER. Other things would be placed in a "PROBABLY SPAM FOLDER". Etc.
Anyways, Bayesian Analysis is a really great method.
If your interested in Bayesian Analysis, there's a great phylogeny program which gives you (basically) a bootstrapped maximum likelihood tree (calculated from millions of trees) via Bayesian Analysis: MrBayes.
LOL, the IRS uses computers to do your taxes? Ok, if you call those room-sized monstrosities in the IRS from the 1950's which still use tape-recorders, then maybe.
Anyways, you claim that if the software the IRS uses to determine audits were OpenSourced, people could scam the IRS, is absurd.
Think configuration files.
I.e., in psuedocode, a section of the program might say:
Conduct audit if:
income > x AND taxes paid y
Where the variables x, y, and z would be defined in a configuration file, which would not be released to the public.
Similar problems can be dealth with in a similar way. In either case, the problem isn't with the disclosure of the code itself, but rather certain key values. Its the difference between disclosind the encryption algorithm/software and disclosing one's personal key.
Alot of people here seem to be pouncing on the idea that such laws would require the gov't to use OSS / FS even when it is not the best solution. Though I'd argue that in most cases -- overall -- it is, it's possible that it may not be the best solution. So, yes, this law would force the government to occasionally use software which wasn't the best solution.
But there are other things at play here than the best solution to a problem: namely, a core value to any Democracy -- transparency. Proprietary software decreases transparency, OSS / FS / public domain software increases it. I'd also argue that the gov't can use software which is "source for free to dist/change/whatever, but pay to use". This still preserves that basic principal of transparency.
I know there are some of you here who insist that this is "affirmative actions for software". I disagree. This is mandating moral character in the software that our government uses. Don't laugh. A person without moral character -- i.e., a crook -- will have difficulty getting a job. Software can also have a moral character; in this case, it certainly does. In the government, software which is open preserves transparency has moral character, whereas that which does not is does not have moral character.
If you still have problems with this, it most likely comes back to the idea that "the best software should be chosen". While I think that the licensing should be considered in deciding what's the best software, lets ignore that for now. Lets say that a proprietary product is the best. So what? The government has lots of money, and can easily take an open piece of software and make it the best for the job. This would ultimately save the government money. Alternatively, it could buy the rights to the "best product" from the owner, then open that product up.
The question then, is which license should software the government creates/modifies/buys fall under? Well, if its modified GPL code, obviously it has to fall under the GPL. But if they create something from scratch, or buy out a proprietary product, what then? Well, I suggest in all cases -- until the desired license is decided on -- the government stick to the GPL. Why? Because you can change the license on the exact same product from the GPL to OSS (i.e., BSD) to public domain. But you can't change the license on the same piece of software from public domain to BSD to GPL.
Transparency is a key value in any Democratic government, and the more transparent a democracy is, the healthier it is. Conversely, the worst of governments and authorities aren't transparent at all. As an example government, take Iraq, and as an example authority, take ICANN.
"So if the paper copy is kept in a file cabinet, do we have the right to know how the lock works on the file cabinet?
Yes, actually we do have the right to know how the lock works. How else can be know that the information is actually safe?
Do we have a right to try to break into said file cabinet?
I don't know where you got that crazy idea. The right to know how something works (in other words, transparency in our government) does not imply the right to try to break into a locked cabinet or file.
Should we really lobby the government to outlaw the use of any file cabinet that's not home-brew with published blueprints?
Any time there is not total transparency in government, the basic principals of Democracy are threatened. Democracy itself is threatened. Any Democracy which does not have as much transparency in the government as possible is in danger. Obviously, there are some exceptions; no one's suggesting the F-117 should have been developed out in the open.
What about all the software a government might need that doesn't have a good open-source alternative?
Such as? I find it difficult to think of any significant software for which there is not an equivalent if not better OSS / FS option. And should an option not exist, the government can create one, which would be either public domain, OSS, or FS.
Should we drive out of business all the companies that make cheap, good, proprietary software for government use?
Who says they're going out of business? They're free to make money from other sourcees. They're also free (or I'd propose they should be, in most cases) to develop software openly, under a license which would grant the right to distribute the source code and any modifications, but not to run a copy of that software without paying the company for the license to do so.
Do we really need to publicly shame decenting voices within our own community, labelling them pro-Microsoft zealots with a hidden agenda?
That's your take on what Michael said, and, I'll grant, a reasonable one, but not the only one. He did not label O'Reilly a pro-MS zeal, nor did he say he had a hidden agenda. Michael said that he thinks that O'Reilly is promoting the Software Choice campaign. He said that perhaps O'Reilly has a hidden agenda.
Also, you seem to imply that I'm agreeing with Michael, which I'm not; other people who disagree with O'Reilly on this aren't necessarily agreeing with MIchael's opinions on O'Reilly promoting the Software Choice campaign.
I simply said that I think that O'Reilly is off on this one. I don't think that he's promoting the Software Choice campaign, nor do I think he has alterior motives.
Do you really think the founding father's wanted a government which operated secretly, behind the shadows, with the people not knowing hot it operates?
Just because the constitution does not say something does not mean it shouldn't be a guiding principal. The founding father's could not have envisioned the circumstances we face today; the constitution was designed to protect a certain set of values, but it could only protect those values from threats that existed at that time, or threats of a similar nature.
The founding father's never imagined computers, TV's, radio, etc etc.
I didn't say that would necessarily be the goal -- to keep the algorithms a secret.
However, there's some things which the government should keep a secret. Namely, I'm thinking of anything involved in stealth technology.
The software used to aid in developing the F-117, F-22, and B-2 should be kept secret, for example. My point was simply that the gov't could base such software on OSS / FS software, but keep the modifications secret.
his is not about OSS / FSS software on anyone. Its about transparency in the government -- about the people's right to know.
The people have the right to know exactly what source code the government is using to protect them. We have the right to know what code protects our privacy in, for example, records which are ruled sealed.
Lets say that your daughter's molested and a trial occurs, in which she testifies. For her protection, her testimony is sealed; if an electronic copy is made, it is cryptographically sealed. If this is done using proprietary software, we the citizens have no way of being assured that it is really secure. If the software used to do that is OSS / FS, then we can check and make sure.
This is a somewhat important example, but the same principal applies to even trivial things. We, the citizens, have the right to know exactly how the software our government is using works; at least where it pertains to us.
Obviously, military top secret stuff is different; though it certainly need not be based on proprietary technology -- nothing prevents the military from modifying OSS / FS software and then keeping those modifications secret within the division. As that doesn't really count as distribution; i.e., in house modifications are not considered "distributed". Its only "distribution" when you make it available to the general public.
That is why the government mustI use OSS / FS, because of our right to know.
An additional benefit is cost-effectiveness. Our tax dollars pay for this stuff, and in almost all cases, OSS / FS is a cheaper solution, both in terms of initial price and total cost of ownership.
As a recent convert from MS Office to OpenOffice, I'll admit, OpenOffice has problems.
Namely, its hard to do a lot of common things and it loads slowly. This is not just "conversion pains". I've become accustomed to OpenOffice rather quickly, but the ways in which it makes you do things are just too long. The shortest distancess between two points is a straight line: A --> Z. Not A --> D --> B --> E --> Q --> N --> S --> Z.
That said, most of the problems with OpenOffice can be fixed by the user, if one isn't too lazy. Its very customizable, so you can define your own shortcut functions, and toolbars, etc.
Another big problem with OpenOffice is the spell-checker. There needs to be a spell-checker and grammar checker.
There are also some very nice things about OpenOffice:
1. It generally doesn't fuck you up. Usually, it won't automatically change what you type. If Itype in nip7p at the beginning of a sentence, that's what I want, not Nip7p. A word processor should not second-guess the user.
2. Word completion. Nice.
3. Pinnable stuff. Alot of things are pinnable, like the color selection menu.
4. FREE PowerPoint modifier: Impress. Why should I waste 300 dollars on PowerPoint when Impress is free?!
5. Its not MS. Has a good, GPL'ed license.
6. Can read/open/save many different file-formats.
7. Metric! Inches are out, centimeters are in. Ok, at least among us scientists.
8. Available on many diff platforms: Apple, Intel, AMD, Sparc. This is great if you work with Apples and PC's.
That said, all these good things are no excuse for OpenOffice's deficiencies:
1. User interface. It needs to be smoother. Commonly used things should be easily accessible, and right clicking should always bring up something useful.
2. Load/run time. I have a 1.1GHz computer, 256Mb RAM, 7200rpm ATA100 hard drive, and it takes 15-30s for it to load. COME ON. That's CRAP. You'd think it was written in Java or something. Any program which doesn't open nearly instantaneously on my machine is crap in terms of load time.
So, my advice to OpenOffice: don't worry about features. The features in OpenOffice are sufficient to 99.99% of all the users. The problem is making those features easily accessible, and making the program load/run faster.
Idiot. A business is different than a private home. Businesses have stocks, hence are public entities (that is, the public owns a share in them). Businesses are protected by special laws. Citizens of Neveda pay extra taxes to help protect Casino's from frauds. Hence, businesses -- especially Casino's -- have to play by public rules.
You can't constitutionally exclude someone from your place of business (if its an otherwise public place, which Casino's are) because they're black. You also can't exclude them because you don't make as much money off of them.
Don't like it? Too bad. That's the way it is. The problem is these inconsistencies when it comes to Casino's. Casino's are for some inexplicable reason treated differnetly from other businesses. McDonalds can't prevent someone from coming there because they're black, or because they order cheaper meals; yet, Casino's can do such things.
If a company wants to have shares, and wants to obtain the advantages the government offers to corporations, they have to play by public rules. If they don't want any of that, then I'm fine with them having whatever assinite rules they want. But if your going to receive special benefits off of MY TAX DOLLARS, you should have to play by public rules. That's why I think the Church should either not get any tax breaks, or have to reform their internal rules to allow (for example) homosexual marriages.
Anyways, my main gripe isn't that Casino's exclude people for unjust reasons. Its that they're all run by and affiliated with crooks. Read the article. You win, the Casino's don't give you your money. They harass people. They assault people. They make people "disappear".
Unlike some idiots here have suggested, I'm not basing my impressions of Casino's off of B-rated movies. I can't remember a movie I watched where a Casino played a predominant part. I'm also not basing it off of personal experience -- I've never gambled, and I find nothing fun about losing all of your money. If I'm going to go to Las Vegas, its going to be for the good cheap stuff there (food, hotel, strippers, prostitutes) and not to gamble. If I'm going to spend 100 bucks, its going to be on a blow-job, not a fucking black-jack game.;-)
I'm basing my opinions of Casino's on what I've heard from those who do well in Casino's. Anyone who consistently does well in a Casino is harassed, violated, assaulted, murdered, etc etc. Again, read the article.
Btw, I find it rather funny that the ppl at/. can see that the likes of Hillary Rosen, Bill Gates, and Jack Valentini are dark shady characters, but somehow think that the people running Casino's in Las Vegas are angels.
For a group, you need a large number of people -- perhaps 100 or so -- to work together randomly and hit different casino's. That makes it difficult for Casino's to notice a "regular" winner.
Also, probably a good idea to bring a few lawyers along. The people who work in or own Casino's are all mob-affiliated crooks, anyways.
Casino's are the one's who are organized crooks
on
MIT vs. Las Vegas
·
· Score: 2, Flamebait
The Casino's are the entities that are involved with and/or organized crime, not card-counting groups.
Breaking into people's appartments, stealing their money, harassing them, killing people -- this is stuff that Casino's do, not card-counters. They are the one's involved with organized crime.
The rulings by courts that Casino's can exclude individuals for any reason are unconstitutional. McDonald's can't ban anyone from coming in there; restaurants can't ban people from coming there who come there and order cheap meals along with water to save money. In other words, they can't ban the people who aren't as profitable to them. So why should Casino's be able to?
Stories like this illustrate why gambling is illegal in most states. Casinos are run by crooks and mobsters, who will use illegal tactics to maintain their profitability (i.e., breaking/entering, harassment, murder, blackmail, etc).
I agree that gambling should be legal; however, it should be tightly regulated and controlled.
Casino's don't like card counters. Tough. That's not a good enough reason to ban them from your resort. Fast food places can't ban people for any reason, why should Casino's be able to?
If Casino's have a problem with card-counting, its up to them to come up with legitimate tactics to deal with it: cutting the deck, switching dealers, using large decks, mixing more thoroughly, etc. Plenty of tactics they can use which aren't illegal.
But quite frankly, I don't care if this ruins their business. They have billions of dollars to spend. If they aren't smart enough to catch on to card-counting schemes and develop counter-measures, they deserve to go out of business.
Bunch of big whiners. Waaah! Waaah! Keep on crying because your too fucking dumb.
Greedy fucks at Casino's. They rig the game against you so that you basically can't win (i.e., house has 90% advantage).
Then, when a few smart people (maybe one out of 1000) come along who can count cards and actually break even or better, they bitch and whine.
Its just a technique -- a legal one, as opposed to casino's illegal techniques of winning. Look in Hoyle's rule-books on cards. They won't mention anything about it being illegal to count cards. However, their rules for blackjack don't set it up so that the dealer has a 90% advantage.
By the way, does anyone here really think that the dealers don't count the cards? Bullshit. You know damn well they do.
Furthermore, casinos have a number of tactics to foil card-counting, such as cutting the deck, starting a new deck, or mixing in several decks.
Thus, the source code for any software it uses should be available.
The people have the right to know what code the government is using to protect confidential information, criminal records, driving records, manage taxes, etc etc. Closed-source software destroys the possibility of transparency in the government, and denies people that right.
Furthermore, its OUR tax dollars which are paying for this stuff. Thus, more cost-effective solutions -- hence open source software / free software / public domain software -- should be used by the government.
Furthermore, the government should not use any standards which lock/force people into using any particular kind of software. That means no proprietary standards (like MS.doc files). Proprietary standards force people into using particular programs (like MS Word). Open Source / Free standards should be used, as they don't lock the people into any particular program. A standard like OpenPGP can be incorporated by anyone into their program, be it the FSF, the OSI, or MS.
Now, regarding government development of software. In all cases, government-funded projects should produce something which is freely available to the public. That means public domain, GPL, or Open Source Licenses. These licenses (or lack thereof) make the results of government-funded projects available to the public. In regards to the GPL, it requires that you GPL any modifications. But this is a good thing. It is good that the government promote recipricol relationship communities, as the GPL does. This is in the public interest. It is in the public's interest that any software produced or funded or supported by the US Government become public domain, GPL, or covered under any of the OSI certified licenses. It is not in the public's interest that such fall under a proprietary license: that means that citizens pay TWICE for a product. Once to support its development, then again to buy it.
The simple fact that in 99.99% of cases using Open Source/GPL software saves money should be enough to justify its use. In the few cases where it doesn't, that's b/c its not as good as the proprietary equivalent, but that can easily be fixed by government-funded development.
Even in the very few (0.01%) of cases where you save money by using proprietary software, that still doesn't justify using it in those cases. Because the public has the right to know what the code is the govenrment is using (as this affects their lives), any code the government uses should be transparent.
Computer screens are 2D. TWO DIMENSIONAL. Thus, they're best suited to displaying TWO DIMENSIONAL things, not 3D things (games are an exception).
3D file browsers, web pages, word processors, and whatever else are cool, but not nearly as useful as the plain old 2D ones. FACT -- its hard to read stuff at an angle.
Lets give the 3D stuff a rest. It may be cool, but its completely useless. I am not a believer that people can always handle things better in 3D. Try finding some file on your desk. You can do it, but it'll take you awhile. Certainly not as convenient as the way an OS displays files, organized in folders and whatnot.
3D interfaces will never be useful on a 2D screen. When VR helmuts and suits become common, then maybe 3D interfaces will find their use, but I'm still doubtful.
The fact is, any thing which is supposed to be productive and is in 3D is just a publicity stunt. Make it look cool. Who cares if it works or not. 3DOS -- useless. Fsn -- useless. Fsv -- useless. 3DWM -- useless. Need I go on? I've used all the 3D interfaces just because they're cool. That doesn't mean that I'm going to use them on a regular basis to do actual work.
As for web-sites, they should convey INFORMATION, not show off the latest new flashy useless technique which hogs up all of our bandwidth.
There is a very good reason why UNIX is case-sensitive, or at least I can think of a very good reason to keep it case-sensitive.
In biology, writing out a protein name in all uppercase letters indicates the wild-type GENE (i.e., VAC8), while writing it in lowercase letters indicates a mutant of the wild-type gene (i.e., vac8). This is just a specific example. In other words, there are many instances where case matters.
VAC8 is not the same as vac8 (one refers to the wild-type gene, the other to a mutant).
Federal is not the same as federal (one, I believe, refers to specific federal entities, the other to the idea there-of).
Bush is not the same as bush (one is the President, the other is a plant).
Banks is not the same as banks (one is a last name, the other is a financial institution).
Come on. Its not that difficult for people to grasp the concept that a file named dave is not the same as one named DAVE. If certain users really don't like that, then they should disable cap-sensitivity (I'm sure there are options to do such).
There is a lot of talk about making Linux easier to use for the average person. Just as easy -- and an effort which would not require programming skills -- would be to make the average user more Linux-savvy. This is what a large part of the Linux community (known as gurus) does. It means you don't say rtfm to every question, though it is good to try to help people figure things out for themselves, rather than telling them the answer.
Its not a matter of me liking this future or not liking it (though the apparent consensus on/. that a program and set of algorithms can completely eliminate the need for human beings in acting is disturbing), its simply a matter of what's realistic.
Firstly, to be able to realistically have a computer CGI character emulate what a real-life person would do, we'd need to know alot about real-life people. Fact is, we don't. We simply do not know enough about human beings to accurately emulate them in a CGI world. Its doubtful that we ever will, but certainly no "profound understanding of human beings" psychologically, physically, or biologically will be come about in the next century or two. Sure, we're making leaps and bounds; but, all things considered, of what there is to understand about human beings, we understand 1x10^-6% of it. In other words, very little at all.
Because of this, CGI characters will not (unless you naively think that we will soon learn 50% or more of all there is to know about human beings) accurately emulate human beings, no matter how complex the algorithms, subroutiines, and routines may be.
Your insistence that CGI characters will be able to perfectly emulate human beings, or so close that the differences will be imperceptible, shows a rather large ignorance of the complexity of life, specifically human life.
There are problems out there alot less complicated than human behavior patterns which could bring the most poweful supercomputers in the world to a crawling halt. All the computers in the world working as one computer couldn't give you the phylogeny (by Bayesian Analysis) of 10,000 species. That's a simple problem. Human behavior is a much more complicated problem than maximizing local optima of probability.
That said, there's no way that CGI characters will accurately emulate human behavior in the near future.
On a ideological note, I find this apparent desire among/.ers to replace actors with computer programs somewhat disturbing. Today, its we replace actors with computer programs. The trend continues. In a few centuries, human beings will be doing nothing but eating food provided by machines, sleeping, having sex, and lying on the beach.
Sorry, but no matter how complex algorithms get, they'll never be quite the same as real actors.
Computers cannot convey emotion. Period.
Despite the rave about AI, it will never be anywhere near what humans are. Why? Well, that's obvious. There is no way any combination of computer architecture and programming could come anywhere near the complexity of the human brain -- ever.
CGI is not going to make actors, costume designers, score composers (like John Williams), or directors obsolete. Its simply going to be a tool to supplement and aid.
Even when computers graphics, sound, and physics get so good that we could design exacting realism via CGI, it would still be painstaking, consuming too much time. Think about all of the things that real-life actors do and real-life scenarios do, which would have to be emulated. All of the little habbits, motions, etc etc -- not to mention voice and emotion. Sorry, but there's no way that one guy is going to be able to sit at his computer and create a complicated movie with several characters, and accurately express emotion in their appearances and voices.
Ultimately, it will still be much cheaper to higher real actors for major parts -- they won't be necessary for background parts, like crowds, armies, etc; but for the main parts, completely necessary.
CGI will, of course, be very useful in many movies (don't count on it being used for Soap Operas, though). It will be used to eliminate flaws, or even to place characters in a virtual or modified world (as was done in Jurassic Park 1/2). CGI will also be useful for things which simply aren't possible in the real world -- like dinosaurs, for example; or space-ships, aliens, etc etc.
But real-world models will also still be used. Though computer CGI is evolving at an exponential rate, so is animatronics. 10, 20 years down the road, it may be possible to do a movie like Jurassic Park using life-sized robotic recreations. What's the advantage to this? Well, in terms of the creature, very little. But in terms of the actors, alot. Its hard for an actor to seriously act terrified when some head on a stick representing a T-rex is chasing them.
Of course, if such is used, CGI will also be used to supplement it. Animatronic models may be able to walk and look like dinosaurs, for example, but don't count on them steaming up a window with their breath, or many other things which real animals would do. So CGI will be used to add that.
CGI will (already has been) very useful. But it does not completely eliminate the need for traditional approaches. I'm sorry, but a person created entirely on a computer will never have the same emotion as a real character.
No, its more likely that the article mis-stated what occured.
There's no way that one could be able to calculate something in 11 minutes (which before took 17 hours) just by switching to Linux. Linux may be good, but this would imply that the Unix code was so crappy that it was hundreds of times less efficient than it could have been. Doubtful.
No court in the nation is going to enforce any restricitons which prevent one from publishing an evaluation of a product -- certainly not benchmarks, which are one of the most important tools for evaluation. The public has the right to know how well a product works.
As for the automatic update worry, one should note that the whole issue is moot so long as the user can disable automatic updating (in which case, they'd update manually, and only select the things for which they wanted and presumably knew what they contained).
For UI's, whats need is -- like the author says, transparency. Complete transparency would be best.
That's why most UI's are crap. MacOS's has the right idea with the universal file menu, as that saves alot of screen space -- but they should take it one step further. The universal file menu should hide away unless you move your mouse to the top of the screen. Better yet, it should "pop up" whenver you "Apple" click the mouse: why make the user go to the menu, when the menu can come to the user?
In that regard, MS has the right idea with making the start bar hide-away (as an option). Again, we want to see what we're doing, not joe-genuis' idea of a "cool UI".
On to Mozilla. Absolutely terrible use of screen space. File menu, buttons, search bar (i.e., UltraBar), and address box should all be on ONE LINE. They should also be hide-away. Or better yet, pop-up. Again, why make the user go to the UI when the UI can come to the user? Better yet would be both hide away and pop up.
IE's interface isn't much better -- still crap. You can get it down to one line (as I have): (1) File menu; (2) Buttons (moderate amount); (3) UltraBar; (4) Address bar. This might not be possible on everyone's computer, but I work at 1600x1200, so it damn well better be possible on mine. Even so, its still clunky and could be hide-away, or pop-up; i.e., pressing "Ctrl + right click" would bring up the file menu, buttons, search bar, and ultra-bar. Or maybe a "middle click" would. Btw, that feature where u hold down the scroll wheel and drag the screen -- useless: screen's scroll by too fast. The scroll wheel, however, when set to scroll one page at a time, is good.
So, what's my criteria for the ideal browser UI? Well, I think its obvious from what I've said. It is also my criteria for the ideal UI period. We shouldn't see it unless we want to. It should be hide-away when the mouse isn't moved to the sides of the screen, or pop-up, or both. So, does any browser succeed in doing that? Well, I was about to say Lynx, but not even Lynx does it. Lynx is the text-base equivalent of Mozilla, where all the UI is basically many line. If you look, you'll notice that not one, not two, but thre entire lines are taken up by the UI; four, if you include the header. Don't get me wrong, Lynx is a great browser for efficiency, but its UI is clearly crap. Four lines taken up by UI? Two of which ("press space for next page" and "My Lynx Page (p1 of 2)", could have been placed on the same line, one aligned left, the other right? One of which was largely unnecessary due to the obviousness of what it said? What crap.
Why not use an interface like Vi's? The UI only apears when you press "Shift +:".
That said -- that basically all UI's to date are not transparent enough -- the engine behind Mozilla, Gecko, is great. Just needs a better UI. Does not need a "haxr633Rt" UI which looks techy and complicated. Does not need a sickeningly sweet pretty (read Aqua) UI. Needs a UI which you don't see when you don't want to see it. Or at least which provides that option to users with moderate skill.
Why not set up a WiFi web between various broadband users, even if they use different services. This way, one can potentially get more bandwidth when others are at work or out. It also provides complete annonymity.
For ANY broadband connection, your paying for unlimited 24/7 time at a pretty high bandwidth (100-200KB/s). They sell you unlimited access, and that's what they should expect you to do.
All four of these options on how to use one's unlimited access are essentially the same, and users should have the right to do any of them:
1. Stay online 24/7. This may be done by avid file-sharers and/or downloaders. Some Linux people might want to download the latest ISO for every different distro.
2. Though not using the connection 24/7 one's self, allowing other's in one's house-hold to use it when one isn't. If different people work different shifts, this is essentially 24/7 usage.
3. Allow any of your neighbors to come in your home at any time and use your internet connection. Again, essentially the same as 24/7.
4. Set up a WiFi network. Same as #3, but avoids security issues such as one's computer being stolen.
ISP's are selling you unlimited bandwidth, and they should expect you to use it. They advertise 24/7 then whine when people actually do use their connection 24/7. Waaah.
So, if a car company sells u a car with a contract that endangers u, that's legit? This is like saying Ford can sell you a car but deny you the right to notify others of problems with it. Its invalid.
Slashdot Mirror
It'd be great now if he offered an implementation which we coudl all use.
I think a progressive, ever-going implementation is best. I also think its best to filter based on headers first, and not download any spam (to save bandwidth) and then filter based on message content (for the messages downloaded) and move any spam to a spam folder.
Then the user simply looks at the spam folder and looks for false-positives, and marks them as "legit". Then the Bayesian filter recalculates.
Same thing for false negatives, and for the messages not downloaded. The user can look at the headers of the messages not downloaded and say if they're spam. Then the Bayesian filter recalculates.
Another good thing to do is to give a "password" to your friends for them to put in headers sent to you. I.e., 13y4890dshfpljk2134y9073254y32p9ur. Any message with that in the header would be given a 0% probability of being spam, as only those you gave that to would know to put it in the header. Should it become compromised, you can change it (or just don't give it to people who might compromise it).
Back to the Bayesian filter, another good thing might be to have varying levels of "spam". I.e., if something is almost certailny spam (i.e., 99.99999999% likely to be spam, as would a message with the header "Get fucked for free and make lots of $$$$$"), it would be placed in a DEFINATELY SPAM FOLDER. Other things would be placed in a "PROBABLY SPAM FOLDER". Etc.
Anyways, Bayesian Analysis is a really great method.
If your interested in Bayesian Analysis, there's a great phylogeny program which gives you (basically) a bootstrapped maximum likelihood tree (calculated from millions of trees) via Bayesian Analysis: MrBayes.
LOL, the IRS uses computers to do your taxes? Ok, if you call those room-sized monstrosities in the IRS from the 1950's which still use tape-recorders, then maybe.
Anyways, you claim that if the software the IRS uses to determine audits were OpenSourced, people could scam the IRS, is absurd.
Think configuration files.
I.e., in psuedocode, a section of the program might say:
Conduct audit if:
income > x
AND
taxes paid y
Where the variables x, y, and z would be defined in a configuration file, which would not be released to the public.
Similar problems can be dealth with in a similar way. In either case, the problem isn't with the disclosure of the code itself, but rather certain key values. Its the difference between disclosind the encryption algorithm/software and disclosing one's personal key.
Alot of people here seem to be pouncing on the idea that such laws would require the gov't to use OSS / FS even when it is not the best solution. Though I'd argue that in most cases -- overall -- it is, it's possible that it may not be the best solution. So, yes, this law would force the government to occasionally use software which wasn't the best solution.
But there are other things at play here than the best solution to a problem: namely, a core value to any Democracy -- transparency. Proprietary software decreases transparency, OSS / FS / public domain software increases it. I'd also argue that the gov't can use software which is "source for free to dist/change/whatever, but pay to use". This still preserves that basic principal of transparency.
I know there are some of you here who insist that this is "affirmative actions for software". I disagree. This is mandating moral character in the software that our government uses. Don't laugh. A person without moral character -- i.e., a crook -- will have difficulty getting a job. Software can also have a moral character; in this case, it certainly does. In the government, software which is open preserves transparency has moral character, whereas that which does not is does not have moral character.
If you still have problems with this, it most likely comes back to the idea that "the best software should be chosen". While I think that the licensing should be considered in deciding what's the best software, lets ignore that for now. Lets say that a proprietary product is the best. So what? The government has lots of money, and can easily take an open piece of software and make it the best for the job. This would ultimately save the government money. Alternatively, it could buy the rights to the "best product" from the owner, then open that product up.
The question then, is which license should software the government creates/modifies/buys fall under? Well, if its modified GPL code, obviously it has to fall under the GPL. But if they create something from scratch, or buy out a proprietary product, what then? Well, I suggest in all cases -- until the desired license is decided on -- the government stick to the GPL. Why? Because you can change the license on the exact same product from the GPL to OSS (i.e., BSD) to public domain. But you can't change the license on the same piece of software from public domain to BSD to GPL.
Transparency is a key value in any Democratic government, and the more transparent a democracy is, the healthier it is. Conversely, the worst of governments and authorities aren't transparent at all. As an example government, take Iraq, and as an example authority, take ICANN.
"So if the paper copy is kept in a file cabinet, do we have the right to know how the lock works on the file cabinet?
Yes, actually we do have the right to know how the lock works. How else can be know that the information is actually safe?
Do we have a right to try to break into said file cabinet?
I don't know where you got that crazy idea. The right to know how something works (in other words, transparency in our government) does not imply the right to try to break into a locked cabinet or file.
Should we really lobby the government to outlaw the use of any file cabinet that's not home-brew with published blueprints?
Any time there is not total transparency in government, the basic principals of Democracy are threatened. Democracy itself is threatened. Any Democracy which does not have as much transparency in the government as possible is in danger. Obviously, there are some exceptions; no one's suggesting the F-117 should have been developed out in the open.
What about all the software a government might need that doesn't have a good open-source alternative?
Such as? I find it difficult to think of any significant software for which there is not an equivalent if not better OSS / FS option. And should an option not exist, the government can create one, which would be either public domain, OSS, or FS.
Should we drive out of business all the companies that make cheap, good, proprietary software for government use?
Who says they're going out of business? They're free to make money from other sourcees. They're also free (or I'd propose they should be, in most cases) to develop software openly, under a license which would grant the right to distribute the source code and any modifications, but not to run a copy of that software without paying the company for the license to do so.
Do we really need to publicly shame decenting voices within our own community, labelling them pro-Microsoft zealots with a hidden agenda?
That's your take on what Michael said, and, I'll grant, a reasonable one, but not the only one. He did not label O'Reilly a pro-MS zeal, nor did he say he had a hidden agenda. Michael said that he thinks that O'Reilly is promoting the Software Choice campaign. He said that perhaps O'Reilly has a hidden agenda.
Also, you seem to imply that I'm agreeing with Michael, which I'm not; other people who disagree with O'Reilly on this aren't necessarily agreeing with MIchael's opinions on O'Reilly promoting the Software Choice campaign.
I simply said that I think that O'Reilly is off on this one. I don't think that he's promoting the Software Choice campaign, nor do I think he has alterior motives.
Do you really think the founding father's wanted a government which operated secretly, behind the shadows, with the people not knowing hot it operates?
Just because the constitution does not say something does not mean it shouldn't be a guiding principal. The founding father's could not have envisioned the circumstances we face today; the constitution was designed to protect a certain set of values, but it could only protect those values from threats that existed at that time, or threats of a similar nature.
The founding father's never imagined computers, TV's, radio, etc etc.
I didn't say that would necessarily be the goal -- to keep the algorithms a secret.
However, there's some things which the government should keep a secret. Namely, I'm thinking of anything involved in stealth technology.
The software used to aid in developing the F-117, F-22, and B-2 should be kept secret, for example. My point was simply that the gov't could base such software on OSS / FS software, but keep the modifications secret.
Again, O'Reilly has missed the point.
his is not about OSS / FSS software on anyone. Its about transparency in the government -- about the people's right to know.
The people have the right to know exactly what source code the government is using to protect them. We have the right to know what code protects our privacy in, for example, records which are ruled sealed.
Lets say that your daughter's molested and a trial occurs, in which she testifies. For her protection, her testimony is sealed; if an electronic copy is made, it is cryptographically sealed. If this is done using proprietary software, we the citizens have no way of being assured that it is really secure. If the software used to do that is OSS / FS, then we can check and make sure.
This is a somewhat important example, but the same principal applies to even trivial things. We, the citizens, have the right to know exactly how the software our government is using works; at least where it pertains to us.
Obviously, military top secret stuff is different; though it certainly need not be based on proprietary technology -- nothing prevents the military from modifying OSS / FS software and then keeping those modifications secret within the division. As that doesn't really count as distribution; i.e., in house modifications are not considered "distributed". Its only "distribution" when you make it available to the general public.
That is why the government mustI use OSS / FS, because of our right to know.
An additional benefit is cost-effectiveness. Our tax dollars pay for this stuff, and in almost all cases, OSS / FS is a cheaper solution, both in terms of initial price and total cost of ownership.
It also has promise.
As a recent convert from MS Office to OpenOffice, I'll admit, OpenOffice has problems.
Namely, its hard to do a lot of common things and it loads slowly. This is not just "conversion pains". I've become accustomed to OpenOffice rather quickly, but the ways in which it makes you do things are just too long. The shortest distancess between two points is a straight line: A --> Z. Not A --> D --> B --> E --> Q --> N --> S --> Z.
That said, most of the problems with OpenOffice can be fixed by the user, if one isn't too lazy. Its very customizable, so you can define your own shortcut functions, and toolbars, etc.
Another big problem with OpenOffice is the spell-checker. There needs to be a spell-checker and grammar checker.
There are also some very nice things about OpenOffice:
1. It generally doesn't fuck you up. Usually, it won't automatically change what you type. If Itype in nip7p at the beginning of a sentence, that's what I want, not Nip7p. A word processor should not second-guess the user.
2. Word completion. Nice.
3. Pinnable stuff. Alot of things are pinnable, like the color selection menu.
4. FREE PowerPoint modifier: Impress. Why should I waste 300 dollars on PowerPoint when Impress is free?!
5. Its not MS. Has a good, GPL'ed license.
6. Can read/open/save many different file-formats.
7. Metric! Inches are out, centimeters are in. Ok, at least among us scientists.
8. Available on many diff platforms: Apple, Intel, AMD, Sparc. This is great if you work with Apples and PC's.
That said, all these good things are no excuse for OpenOffice's deficiencies:
1. User interface. It needs to be smoother. Commonly used things should be easily accessible, and right clicking should always bring up something useful.
2. Load/run time. I have a 1.1GHz computer, 256Mb RAM, 7200rpm ATA100 hard drive, and it takes 15-30s for it to load. COME ON. That's CRAP. You'd think it was written in Java or something. Any program which doesn't open nearly instantaneously on my machine is crap in terms of load time.
So, my advice to OpenOffice: don't worry about features. The features in OpenOffice are sufficient to 99.99% of all the users. The problem is making those features easily accessible, and making the program load/run faster.
Idiot. A business is different than a private home. Businesses have stocks, hence are public entities (that is, the public owns a share in them). Businesses are protected by special laws. Citizens of Neveda pay extra taxes to help protect Casino's from frauds. Hence, businesses -- especially Casino's -- have to play by public rules.
;-)
/. can see that the likes of Hillary Rosen, Bill Gates, and Jack Valentini are dark shady characters, but somehow think that the people running Casino's in Las Vegas are angels.
You can't constitutionally exclude someone from your place of business (if its an otherwise public place, which Casino's are) because they're black. You also can't exclude them because you don't make as much money off of them.
Don't like it? Too bad. That's the way it is. The problem is these inconsistencies when it comes to Casino's. Casino's are for some inexplicable reason treated differnetly from other businesses. McDonalds can't prevent someone from coming there because they're black, or because they order cheaper meals; yet, Casino's can do such things.
If a company wants to have shares, and wants to obtain the advantages the government offers to corporations, they have to play by public rules. If they don't want any of that, then I'm fine with them having whatever assinite rules they want. But if your going to receive special benefits off of MY TAX DOLLARS, you should have to play by public rules. That's why I think the Church should either not get any tax breaks, or have to reform their internal rules to allow (for example) homosexual marriages.
Anyways, my main gripe isn't that Casino's exclude people for unjust reasons. Its that they're all run by and affiliated with crooks. Read the article. You win, the Casino's don't give you your money. They harass people. They assault people. They make people "disappear".
Unlike some idiots here have suggested, I'm not basing my impressions of Casino's off of B-rated movies. I can't remember a movie I watched where a Casino played a predominant part. I'm also not basing it off of personal experience -- I've never gambled, and I find nothing fun about losing all of your money. If I'm going to go to Las Vegas, its going to be for the good cheap stuff there (food, hotel, strippers, prostitutes) and not to gamble. If I'm going to spend 100 bucks, its going to be on a blow-job, not a fucking black-jack game.
I'm basing my opinions of Casino's on what I've heard from those who do well in Casino's. Anyone who consistently does well in a Casino is harassed, violated, assaulted, murdered, etc etc. Again, read the article.
Btw, I find it rather funny that the ppl at
For a group, you need a large number of people -- perhaps 100 or so -- to work together randomly and hit different casino's. That makes it difficult for Casino's to notice a "regular" winner.
Also, probably a good idea to bring a few lawyers along. The people who work in or own Casino's are all mob-affiliated crooks, anyways.
The Casino's are the entities that are involved with and/or organized crime, not card-counting groups.
Breaking into people's appartments, stealing their money, harassing them, killing people -- this is stuff that Casino's do, not card-counters. They are the one's involved with organized crime.
The rulings by courts that Casino's can exclude individuals for any reason are unconstitutional. McDonald's can't ban anyone from coming in there; restaurants can't ban people from coming there who come there and order cheap meals along with water to save money. In other words, they can't ban the people who aren't as profitable to them. So why should Casino's be able to?
Stories like this illustrate why gambling is illegal in most states. Casinos are run by crooks and mobsters, who will use illegal tactics to maintain their profitability (i.e., breaking/entering, harassment, murder, blackmail, etc).
I agree that gambling should be legal; however, it should be tightly regulated and controlled.
Casino's don't like card counters. Tough. That's not a good enough reason to ban them from your resort. Fast food places can't ban people for any reason, why should Casino's be able to?
If Casino's have a problem with card-counting, its up to them to come up with legitimate tactics to deal with it: cutting the deck, switching dealers, using large decks, mixing more thoroughly, etc. Plenty of tactics they can use which aren't illegal.
But quite frankly, I don't care if this ruins their business. They have billions of dollars to spend. If they aren't smart enough to catch on to card-counting schemes and develop counter-measures, they deserve to go out of business.
Bunch of big whiners. Waaah! Waaah! Keep on crying because your too fucking dumb.
Greedy fucks at Casino's. They rig the game against you so that you basically can't win (i.e., house has 90% advantage).
Then, when a few smart people (maybe one out of 1000) come along who can count cards and actually break even or better, they bitch and whine.
Its just a technique -- a legal one, as opposed to casino's illegal techniques of winning. Look in Hoyle's rule-books on cards. They won't mention anything about it being illegal to count cards. However, their rules for blackjack don't set it up so that the dealer has a 90% advantage.
By the way, does anyone here really think that the dealers don't count the cards? Bullshit. You know damn well they do.
Furthermore, casinos have a number of tactics to foil card-counting, such as cutting the deck, starting a new deck, or mixing in several decks.
Thus, the source code for any software it uses should be available.
.doc files). Proprietary standards force people into using particular programs (like MS Word). Open Source / Free standards should be used, as they don't lock the people into any particular program. A standard like OpenPGP can be incorporated by anyone into their program, be it the FSF, the OSI, or MS.
The people have the right to know what code the government is using to protect confidential information, criminal records, driving records, manage taxes, etc etc. Closed-source software destroys the possibility of transparency in the government, and denies people that right.
Furthermore, its OUR tax dollars which are paying for this stuff. Thus, more cost-effective solutions -- hence open source software / free software / public domain software -- should be used by the government.
Furthermore, the government should not use any standards which lock/force people into using any particular kind of software. That means no proprietary standards (like MS
Now, regarding government development of software. In all cases, government-funded projects should produce something which is freely available to the public. That means public domain, GPL, or Open Source Licenses. These licenses (or lack thereof) make the results of government-funded projects available to the public. In regards to the GPL, it requires that you GPL any modifications. But this is a good thing. It is good that the government promote recipricol relationship communities, as the GPL does. This is in the public interest. It is in the public's interest that any software produced or funded or supported by the US Government become public domain, GPL, or covered under any of the OSI certified licenses. It is not in the public's interest that such fall under a proprietary license: that means that citizens pay TWICE for a product. Once to support its development, then again to buy it.
The simple fact that in 99.99% of cases using Open Source/GPL software saves money should be enough to justify its use. In the few cases where it doesn't, that's b/c its not as good as the proprietary equivalent, but that can easily be fixed by government-funded development.
Even in the very few (0.01%) of cases where you save money by using proprietary software, that still doesn't justify using it in those cases. Because the public has the right to know what the code is the govenrment is using (as this affects their lives), any code the government uses should be transparent.
Useless and impractical as that standard may be.
Computer screens are 2D. TWO DIMENSIONAL. Thus, they're best suited to displaying TWO DIMENSIONAL things, not 3D things (games are an exception).
3D file browsers, web pages, word processors, and whatever else are cool, but not nearly as useful as the plain old 2D ones. FACT -- its hard to read stuff at an angle.
Lets give the 3D stuff a rest. It may be cool, but its completely useless. I am not a believer that people can always handle things better in 3D. Try finding some file on your desk. You can do it, but it'll take you awhile. Certainly not as convenient as the way an OS displays files, organized in folders and whatnot.
3D interfaces will never be useful on a 2D screen. When VR helmuts and suits become common, then maybe 3D interfaces will find their use, but I'm still doubtful.
The fact is, any thing which is supposed to be productive and is in 3D is just a publicity stunt. Make it look cool. Who cares if it works or not. 3DOS -- useless. Fsn -- useless. Fsv -- useless. 3DWM -- useless. Need I go on? I've used all the 3D interfaces just because they're cool. That doesn't mean that I'm going to use them on a regular basis to do actual work.
As for web-sites, they should convey INFORMATION, not show off the latest new flashy useless technique which hogs up all of our bandwidth.
NO!
B != b
There is a very good reason why UNIX is case-sensitive, or at least I can think of a very good reason to keep it case-sensitive.
In biology, writing out a protein name in all uppercase letters indicates the wild-type GENE (i.e., VAC8), while writing it in lowercase letters indicates a mutant of the wild-type gene (i.e., vac8). This is just a specific example. In other words, there are many instances where case matters.
VAC8 is not the same as vac8 (one refers to the wild-type gene, the other to a mutant).
Federal is not the same as federal (one, I believe, refers to specific federal entities, the other to the idea there-of).
Bush is not the same as bush (one is the President, the other is a plant).
Banks is not the same as banks (one is a last name, the other is a financial institution).
Come on. Its not that difficult for people to grasp the concept that a file named dave is not the same as one named DAVE. If certain users really don't like that, then they should disable cap-sensitivity (I'm sure there are options to do such).
There is a lot of talk about making Linux easier to use for the average person. Just as easy -- and an effort which would not require programming skills -- would be to make the average user more Linux-savvy. This is what a large part of the Linux community (known as gurus) does. It means you don't say rtfm to every question, though it is good to try to help people figure things out for themselves, rather than telling them the answer.
Its not a matter of me liking this future or not liking it (though the apparent consensus on /. that a program and set of algorithms can completely eliminate the need for human beings in acting is disturbing), its simply a matter of what's realistic.
/.ers to replace actors with computer programs somewhat disturbing. Today, its we replace actors with computer programs. The trend continues. In a few centuries, human beings will be doing nothing but eating food provided by machines, sleeping, having sex, and lying on the beach.
Firstly, to be able to realistically have a computer CGI character emulate what a real-life person would do, we'd need to know alot about real-life people. Fact is, we don't. We simply do not know enough about human beings to accurately emulate them in a CGI world. Its doubtful that we ever will, but certainly no "profound understanding of human beings" psychologically, physically, or biologically will be come about in the next century or two. Sure, we're making leaps and bounds; but, all things considered, of what there is to understand about human beings, we understand 1x10^-6% of it. In other words, very little at all.
Because of this, CGI characters will not (unless you naively think that we will soon learn 50% or more of all there is to know about human beings) accurately emulate human beings, no matter how complex the algorithms, subroutiines, and routines may be.
Your insistence that CGI characters will be able to perfectly emulate human beings, or so close that the differences will be imperceptible, shows a rather large ignorance of the complexity of life, specifically human life.
There are problems out there alot less complicated than human behavior patterns which could bring the most poweful supercomputers in the world to a crawling halt. All the computers in the world working as one computer couldn't give you the phylogeny (by Bayesian Analysis) of 10,000 species. That's a simple problem. Human behavior is a much more complicated problem than maximizing local optima of probability.
That said, there's no way that CGI characters will accurately emulate human behavior in the near future.
On a ideological note, I find this apparent desire among
No matter how good the actor, it still isn't quite the same thing when they're acting terrified, as when they really are.
Proof of point, The Blair Witch Project.
Sorry, but no matter how complex algorithms get, they'll never be quite the same as real actors.
Computers cannot convey emotion. Period.
Despite the rave about AI, it will never be anywhere near what humans are. Why? Well, that's obvious. There is no way any combination of computer architecture and programming could come anywhere near the complexity of the human brain -- ever.
CGI is not going to make actors, costume designers, score composers (like John Williams), or directors obsolete. Its simply going to be a tool to supplement and aid.
Even when computers graphics, sound, and physics get so good that we could design exacting realism via CGI, it would still be painstaking, consuming too much time. Think about all of the things that real-life actors do and real-life scenarios do, which would have to be emulated. All of the little habbits, motions, etc etc -- not to mention voice and emotion. Sorry, but there's no way that one guy is going to be able to sit at his computer and create a complicated movie with several characters, and accurately express emotion in their appearances and voices.
Ultimately, it will still be much cheaper to higher real actors for major parts -- they won't be necessary for background parts, like crowds, armies, etc; but for the main parts, completely necessary.
CGI will, of course, be very useful in many movies (don't count on it being used for Soap Operas, though). It will be used to eliminate flaws, or even to place characters in a virtual or modified world (as was done in Jurassic Park 1/2). CGI will also be useful for things which simply aren't possible in the real world -- like dinosaurs, for example; or space-ships, aliens, etc etc.
But real-world models will also still be used. Though computer CGI is evolving at an exponential rate, so is animatronics. 10, 20 years down the road, it may be possible to do a movie like Jurassic Park using life-sized robotic recreations. What's the advantage to this? Well, in terms of the creature, very little. But in terms of the actors, alot. Its hard for an actor to seriously act terrified when some head on a stick representing a T-rex is chasing them.
Of course, if such is used, CGI will also be used to supplement it. Animatronic models may be able to walk and look like dinosaurs, for example, but don't count on them steaming up a window with their breath, or many other things which real animals would do. So CGI will be used to add that.
CGI will (already has been) very useful. But it does not completely eliminate the need for traditional approaches. I'm sorry, but a person created entirely on a computer will never have the same emotion as a real character.
No, its more likely that the article mis-stated what occured.
There's no way that one could be able to calculate something in 11 minutes (which before took 17 hours) just by switching to Linux. Linux may be good, but this would imply that the Unix code was so crappy that it was hundreds of times less efficient than it could have been. Doubtful.
More likely that it went from 17 hrs to 11 hrs.
No court in the nation is going to enforce any restricitons which prevent one from publishing an evaluation of a product -- certainly not benchmarks, which are one of the most important tools for evaluation. The public has the right to know how well a product works.
As for the automatic update worry, one should note that the whole issue is moot so long as the user can disable automatic updating (in which case, they'd update manually, and only select the things for which they wanted and presumably knew what they contained).
For UI's, whats need is -- like the author says, transparency. Complete transparency would be best.
:".
That's why most UI's are crap. MacOS's has the right idea with the universal file menu, as that saves alot of screen space -- but they should take it one step further. The universal file menu should hide away unless you move your mouse to the top of the screen. Better yet, it should "pop up" whenver you "Apple" click the mouse: why make the user go to the menu, when the menu can come to the user?
In that regard, MS has the right idea with making the start bar hide-away (as an option). Again, we want to see what we're doing, not joe-genuis' idea of a "cool UI".
On to Mozilla. Absolutely terrible use of screen space. File menu, buttons, search bar (i.e., UltraBar), and address box should all be on ONE LINE. They should also be hide-away. Or better yet, pop-up. Again, why make the user go to the UI when the UI can come to the user? Better yet would be both hide away and pop up.
IE's interface isn't much better -- still crap. You can get it down to one line (as I have): (1) File menu; (2) Buttons (moderate amount); (3) UltraBar; (4) Address bar. This might not be possible on everyone's computer, but I work at 1600x1200, so it damn well better be possible on mine. Even so, its still clunky and could be hide-away, or pop-up; i.e., pressing "Ctrl + right click" would bring up the file menu, buttons, search bar, and ultra-bar. Or maybe a "middle click" would. Btw, that feature where u hold down the scroll wheel and drag the screen -- useless: screen's scroll by too fast. The scroll wheel, however, when set to scroll one page at a time, is good.
So, what's my criteria for the ideal browser UI? Well, I think its obvious from what I've said. It is also my criteria for the ideal UI period. We shouldn't see it unless we want to. It should be hide-away when the mouse isn't moved to the sides of the screen, or pop-up, or both. So, does any browser succeed in doing that? Well, I was about to say Lynx, but not even Lynx does it. Lynx is the text-base equivalent of Mozilla, where all the UI is basically many line. If you look, you'll notice that not one, not two, but thre entire lines are taken up by the UI; four, if you include the header. Don't get me wrong, Lynx is a great browser for efficiency, but its UI is clearly crap. Four lines taken up by UI? Two of which ("press space for next page" and "My Lynx Page (p1 of 2)", could have been placed on the same line, one aligned left, the other right? One of which was largely unnecessary due to the obviousness of what it said? What crap.
Why not use an interface like Vi's? The UI only apears when you press "Shift +
That said -- that basically all UI's to date are not transparent enough -- the engine behind Mozilla, Gecko, is great. Just needs a better UI. Does not need a "haxr633Rt" UI which looks techy and complicated. Does not need a sickeningly sweet pretty (read Aqua) UI. Needs a UI which you don't see when you don't want to see it. Or at least which provides that option to users with moderate skill.
Why not set up a WiFi web between various broadband users, even if they use different services. This way, one can potentially get more bandwidth when others are at work or out. It also provides complete annonymity.
For ANY broadband connection, your paying for unlimited 24/7 time at a pretty high bandwidth (100-200KB/s). They sell you unlimited access, and that's what they should expect you to do.
All four of these options on how to use one's unlimited access are essentially the same, and users should have the right to do any of them:
1. Stay online 24/7. This may be done by avid file-sharers and/or downloaders. Some Linux people might want to download the latest ISO for every different distro.
2. Though not using the connection 24/7 one's self, allowing other's in one's house-hold to use it when one isn't. If different people work different shifts, this is essentially 24/7 usage.
3. Allow any of your neighbors to come in your home at any time and use your internet connection. Again, essentially the same as 24/7.
4. Set up a WiFi network. Same as #3, but avoids security issues such as one's computer being stolen.
ISP's are selling you unlimited bandwidth, and they should expect you to use it. They advertise 24/7 then whine when people actually do use their connection 24/7. Waaah.
So, if a car company sells u a car with a contract that endangers u, that's legit? This is like saying Ford can sell you a car but deny you the right to notify others of problems with it. Its invalid.