I've actually RTFA, and I'm still at a loss about exactly _what_ is better about OOo's XML Schema, or wrong about MS's.
OK, first things first; let's have a little lesson on what XML is. XML is not really that big a deal. All it means really is that the less-than and more-than signs are reserved symbols; one writes constructs such as <foo>..... </foo> to indicate a bounded block of type foo, or <bar/> to indicate a single instance of type bar. The meanings of different foo and bar are what constitute a schema. The HTML used for web pages is actually just a bastardised dialect of XML. End of lesson.
The issues are nothing to do with the schema itself, but rather to do with openness. The OpenOffice.org data format was conceived so that anybody who cares can write applications that speak it, as a right. By contrast, the Microsoft format is closed. If you want to write an application that speaks it, you have to ask Microsoft; they can charge you money for telling you, withhold bits if they see fit, and withdraw the privilege anytime. And if you do anything that Microsoft told you not to do, they can punish you.
What really interests me is exactly which concrete problems should I expect with MS's, that supposedly aren't there if I use OOo's format. If I try to retrieve that data in 5, 10 or 100 years, as in his answer, exactly in which way is OOo's format better?
You can expect the problem with Microsoft's format that only Microsoft -- and a chosen few appointed by Microsoft -- are allowed to write programs that can retrieve your data once it has been saved in Microsoft's proprietary format. OpenOffice.org's format is better because any competent programmer can help you to retrieve that data, without being beholden to anyone.
Exactly _what_ kind of data gets more benefits from his schema than from MS's in that context?
Any data that belongs to you rather than to Microsoft.
In which way, and for what concrete reasons does he foresee that MS's own converters (which so far still import Word 6 documents with no problems) will break down and cry like little girls if fed a Word 12 document some 10 years from now?
That is not the problem. The problem is if, five or ten years down the line, you decide for some reason to move away from Microsoft. There are any number of reasons why you might want to do that: for argument's sake, let's say MS have kept cranking up the cost of Office to the point where you now have to decide whether to try to save money on software licences or lay off staff. Now someone else's document converter may well not be able to handle Microsoft's proprietary format correctly. Your data might become inaccessible! There is also a very real possibility that Microsoft may not exist 10 years from now, and they may take their proprietary formats to the grave with them.
In five, fifty or a hundred years, any competent programmer will still be able to obtain the schema which will enable them to make sense of an OpenOffice.org document, because no one person or organisation controls that schema. No such guarantee can be made in respect of Microsoft's schema.
Or, let me put it this way. Imagine you buy a new car. The bonnet is fastened shut with a tamperproof seal, so only authorised dealers can make repairs -- and they have to use the manufacturer's original specified parts and procedures. You have to buy petrol from the manufacturer's specified filling stations {who will check from time to time that you haven't been tampering with things that do not concern you}. When the car reaches the end of its life {which may come sooner than you think, since the manufacturer can order their service centres not to repair it on a whim} you have to replace it with another one from that same manufacturer; otherwise everything and everybody you ever carried in that car will be left in limbo somewhere, and not fit properly in your new car.
pay a software vendor for a worthless piece of paper {"This may or may not do what you wanted, and we couldn't give a flying toss one way or the other; but if you try to poke about inside it to find out, we will rip you limb from limb and feed you to the hounds"}; or
pay a competent programmer, independent of the original authors and having nothing to lose whatever they say, to determine the suitability of a product for a particular application by inspecting the source code?
In effect, the above is already what most EULAs say anyway. It's just a matter of time before someone gets burned by closed software, tries to sue a major software vendor and calls the whole setup into question. I'm surprised it has not already happened, but maybe it has and they have been hushed up {bought off and given a stern warning not to talk about it ever again..... or just killed}.
My guess is that source code would stand up much better as a defence in court than an EULA, because source code does not actively seek to diminish your statutory rights. And if a bug is obvious to anyone in the courtroom from looking at the source, then it should have been obvious to the plaintiff or their agent in the first place.
I block TV advertisements -- when I'm not watching the Beeb, of course -- by getting up and taking a leak, or brewing a cuppa, or skinning a d00b, or doing anything else I can fit into five minutes. And muting the sound for good measure. {If I'm recording the show I'm watching, I'll stick a chapter marker after the break, so that whoever watches it later can just skip straight past the crap.}
I block magazine advertisements by skipping the pages with the adverts on. Though if the truth be told, I am highly selective with buying magazines. I generally prefer just to rotate myself through several newsagents, browsing without purchase though never for long enough to elicit unwanted attention -- it's a skill I taught myself many years ago. The publishers can't be too worried about it anyway; they are obviously losing less money through people like me than the couple of pence per copy it would take just to seal their magazines inside an envelope.
I block internet advertisements using Squid. It's probably overkill if the truth be told, but it's not hard to set up.
As for why I do it, that's easy. I am the sole judge of what I view. I have already decided that if you feel the need to shove your product in my face, then I will do my damnedest to avoid buying it. I do not owe advertisers anything.
Actually, yes. If you give away candy (Sweets.. as you call it) with poison in it, you're liable even if you mark them as such. The product is intended for consumption, and using the product as intended will cause harm.
But exactly this does happen in real life. You often get chocolate bars labelled with vague warnings like "On rare occasions this product may contain traces of peanuts or other nuts". If you have a serious allergy problem then it might just as well say "On rare occasions this product may contain lethal doses of cyanide or other poisons". They get away with it because it's only a tiny minority of people who are going to be affected. We'll probably have portable matter analysers soon enough, that will be able to tell you instantly at a glance whether or not a particular item actually does contain anything you are allergic to.
I think this -- randomly-present allergens rather than deliberately-placed poisons -- is really more like the software situation. If we're being kind to the closed-source developers, the bugs are there by accident, not design. But even so, the closed source vendors are still effectively saying "This chocolate bar categorically does not contain any kind of nuts whatsoever and you'd better believe us, we will punish you if you try to check with a matter analyser".
Why do you think the Cigarette companies lost BILLIONS of dollars in lawsuits, despite clearly marking their product that it will kill you and everyone around you?
Because someone managed to persuade a judge that two wrongs make a right. Still, it opened up the way for the families of all those people who had been killed by gu..... Oh, it didn't. Never mind.
Further, the source code is no guarantee either, since it's very easy for problems to hide in plain sight, even after extensive analysis.
It is a guarantee. If the source code says "subtract a penny from every transaction involving a person with an 'e' somewhere after an 'a' in their name every odd Tuesday as long as the printer is not switched on" and the software does just that, it's doing what the source code says it's supposed to do. That may not be what you wanted it to do, but it's there in black and white for someone to find. If that sort of logic trap could go unnoticed, then the program must be quite sloppily written. Whoever was carrying out the source audit should have said that there could be nasties lurking in there.
By shifting the onus of responsibility on to the (most likely) unskilled end user, it's just a cop out, and any judge would see that.
I don't buy that argument one bit. As the sole controller of your destiny, you are ultimately responsible for everything that ever happens to you. If you are unskilled in a particular field, then you have two options; either take the time to become skilled in that field yourself, or take the money to pay a skilled person to do it for you. It may not be your fault that you aren't a sufficiently competent programmer, but it sure as hell isn't anybody else's fault either.
Any idiot can go to B&Q and buy some copper pipe, fittings, solder, flux and a blowtorch. If their self-installed pipes leak, or they manage to set their bathroom on fire, or they drip hot solder on themself..... tough titty! Some things are just inherently so complex, that they are not easy to use straight away and you can make nasty mistakes. And programming is much harder than soldering pipes.....
Open Source software is not necessarily free as in gratis, because you may well have to pay to get the code checked over. But that's still not a service you are ever likely to get at any price from the vendors of closed-source software; they have their own reasons for keeping their secrets which exclude all possibility of independent audit.
Science began by making apparently unrelated observations, and later filled in the gaps to create unifying theories. Nature's apparent horror of a vacuum, and a whole lot of other phenomena, are explained by the pressure in a fluid acting equally in all directions. Many phenomena suddenly made sense when it was discovered that matter attracts other matter.
Now, we still have a few gaps, including that small things appear not to behave the same way as big things. No doubt, if we can quantify the differences* -- or explain why that would be impossible -- we can take a stab at a single Grand Unifying Theory which would underpin all of Physics.
It's also possible that there could be another possible set of laws of physics which would be mutually consistent, even consistent with the G.U.T., just contrary to all our observations. If there existed a parallel universe which obeyed this set of laws, one of four things could happen:
It would collapse to a single point in our space
A single point of space in that universe would be bigger than the whole of this universe
It would exist for only a brief instant of our time
A single instant of time in that universe would last longer than the lifetime of our universe
Of course, it's also possible {but extremely unlikely} that there is no Grand Unifying Theory, just a supreme being with a sick sense of humour who keeps changing the rules slightly every time we get close to discovering what they are.....
* Canonical example of difference between quantum and classical phenomena: Why can't a chair just spontaneously shift position? My own take is that quantum wave functions do exist in large systems, but "quantum" phenomena are not generally observed because the waves are not coherent {just as you don't see interference fringes where the light from two candles falls on the same surface}.
I'm not sure that will ever work. Plastic coins, RFID or not, would be too prone to forgery. The devices are easy enough to spoof, and the whole point with coins is that they are used in machines in unsupervised locations. It doesn't matter that you need a rucksack full of gear to simulate a few pound coins, if nobody is going to see you doing it.
Forging actual metal coins is difficult, not impossible but difficult and expensive, and usually not worth it for the returns involved. Better quality requires more investment, which can only be recouped by doing a bigger run; but larger quantities increase the risk of discovery. Forging RFID is trivial and inexpensive by comparison -- and you just know somebody is going to want to have a go at it.
If there is an upside to knowing the history of every banknote, it's this: shops could erect signs saying MONEY WON IN LAWSUITS IS NOT WELCOME IN THIS STORE. And mean it.
Use coins. I already do anyway. The authorities must think I have a massive gambling habit, but really I'm just going into amusement arcades to change serial-numbered notes for unnumbered coins. Coins, being made of metal, cannot have RFID devices embedded in them. Radio waves will not travel through anything that conducts electricity {this is a fundamental limitation of the universe and cannot be overcome by invention}. If you are really paranoid, you can test each coin for conductivity in several places using a simple home-built device {a store-bought AVO may have been rigged}.
No, you misunderstand. It's not the fact that it was given away that changes anything: it's the fact that the source code was included that changes everything. If you give away -- or even sell -- sweets which contain poison and the wrappers clearly say that they contain poison, you are not liable. The person who failed to read the wrapper is at fault. They had access to enough information, they were in a position to make an informed choice, so we have to assume that the choice they made {however detrimental an effect it may have had} was an informed one.
The source code is a guarantee of functionality: it is an irrefutable statement of exactly what the program will and will not do if it is compiled and run on a properly-working computer. Of course it is written in a language that not everybody understands; but, to be blunt about it, that's their problem. For instance, if I write
then any sufficiently-competent Perl programmer should know that this program will not print "Hello World!" on a Saturday; and anybody else who is unsure but wants to find out anyway should engage the services of a sufficiently-competent programmer. {They might well be charged money for this service; but just because we are giving our give our source code away certainly does not mean that we shouldn't be allowed to charge money for explaining what other people's source code does to people who don't understand it.}
But they wouldn't be liable, though! If it was Open Source, they would be giving away the source code. Then it's up to the recipient to make an informed decision {possibly engaging the services of an independent third party} as to whether or not the goods are fit for purpose. Without the source code, they would not be able to make a fully informed decision, and so have to take it on faith that the vendor is being truthful. That's the crucial difference. The source code is a guarantee in its own right.
If the new kernel you installed on your server came with source code, it would be your responsibility to read that source code and check that it did what it was supposed to do.
If you buy a TV stand in kit form, you have the opportunity to decide, at the moment that you build it, whether or not it will suit your intended purpose and if not, what modifications you can make to ensure that it will. For instance, it might come with 3x35 screws: upgrading them to 4x50 and using plenty of hot melt adhesive might allow it to support an extra 10kg., which might make the difference between supporting your nice new widescreen set with NICAM and fastext, and not.
The point is, it's up to you to decide whether goods furnished in kit form will be suitable for the intended application. All the manufacturer is responsible for is making sure that all the parts can be fitted together in such a way as it looks like the picture on the box. If you want to be able to sue someone if your TV stand falls apart and your TV falls on the floor, then get an odd-job person to build it for you, and check their liability insurance to make sure you are covered.
As long as software supplied as source code was treated analogously to goods supplied in kit form, I can see absolutely no problem with obliging software suppliers to offer a guarantee of performance. If you give someone some source code files, you can confidently say "I can guarantee you that if you compile this source code and run it on a computer which is working properly, it will do exactly what the source code says it will do".
I remember, back in the 1970s, there was an appliance known as the "radio-cassette recorder". This was transportable {i.e. it had a handle} and allowed you to enjoy listening to the wireless {MW, LW and FM -- FM stereo on the bigger, more expensive models with two speakers} or your favourite audio cassettes, anywhere there was a power point. {If you put batteries in the thing, you didn't even need the power point, and you could listen to both sides of a C-90 on a single set. Supposedly. Made the machine even heavier to carry, though.} But these devices did not only allow you just to listen to music: you could also record from the radio onto a blank cassette.
Nobody ever had a problem with the existence of radio cassette recorders. Everyone used to record their favourite songs from the radio. Sometimes we'd even go out and buy a "45" if we liked the song enough. Ah, the 1970s and 80s were wonderful times if you were a kid.....
I am allowed to invite my friends into my home to watch my TV, even if they have not subscribed to the channel I am watching -- in fact, even if they do not have a TV licence. And I'm fairly sure I am allowed to lend my home recordings to my friends -- I've certainly never heard of anyone being busted for that, and the practice is almost universal.
The needs of the many outweight the needs of the few, but the needs of the few outweigh the whims and caprices of the many. And even if your assertion were correct {which I severely doubt, given my better than 3:1 fan to freak ratio} it still falls into the "whims and caprices" category rather than the "needs" category.
MySQL does what it does well, and doesn't pretend to do what it doesn't do.
As a flexible but quirky set of persistent array extensions for most known programming languages, including the Three P's, it works really well. Unfortunately, the fact that it uses a bastardised dialect of SQL has led people to mistake it for a relational database management system.
MySQL makes it easy to write web pages with changeable content {such as message boards, diaries, online auctions and personal rant sites}; and with the addition of phpmyadmin, it even becomes a kind of MS-access replacement. You will have to do a lot more work in the application layer if you want to emulate stored procedures and triggers, or if you don't want to run afoul of graceful degradation. But in the 90% of cases where you don't need SPs and triggers, and where you'd rather let a few characters go missing than seize up with a fatal error, it performs just fine. And because it uses SQL {albeit with a broad regional accent}, it can provide a n00b with a sort of gentle introduction to real databases.
The bottom line is, you can't expect to tow a four-berth trailer round the twisty mountain roads with a one-litre Ford Fiesta. But for someone who just wants to drop the kids off at school and then nip into town to buy another pair of shoes, a 4x4 with a three litre engine and six gears is overkill. And you don't see many high performance sports cars with L-plates either.
The fundamental difference between RPM and DEB is that RPM packages depend on specific files while DEB packages depend on specific packages. Debian packages can also implicitly provide other packages. So, there is a Debian "virtual package" called "www-browser". It doesn't contain any files, but it has dependencies which will be satisfied by any of links, lynx, firefox, konqueror, mozilla, &c. In turn, the browser packages claim to provide "www-browser" when any one is installed. If you have some application that requires something that implements "sendmail" functionality {exim, postfix and so forth all call their binary "sendmail" and use the same command line options} then you need only depend on the "mail-transport-agent" package.
Portage is different again. I haven't used it too much. By the time Gentoo arrived on the scene, I was already fluent in Debian, so my experience is limited to an experimental "just for the hell of it" installation. I was impressed with what I saw, though: with a GUI front end, I think it would have the potential to make the whole business of downloading and compiling from source as painless as InstallShield but without the spyware. I've resolved to do some serious GTK-perl hacking at some stage in future, so I'll see if anything comes of that.
Whether APT {which Conectiva bodged to work with RPM packages: it will be interesting to see whether they stick with Mandrake's urpmi tool, which is an APT work-alike} and Portage are actually superior to RPM is not certain, because both Debian and Gentoo feature enormous package repositories including pretty much the whole of Open Source; it's like comparing a few apples to several kilos of oranges.
By the way, circular dependencies can be solved very easily, but it took me ages to figure it out. What you have to do is specify both package names in the same command.
I can name one feature I'd like to see in Mandriva.
Either Debdrake or Portagedrake.
RPMs are bollocks - either that, or every RPM distro that I've used or seen had a package repository that was bollocks, which amounts to the same thing either way: bollocks. Anyway, back in the days when I was using Mandrake 8.2, one of the first things I learned was how to spell "make install".
How exactly is someone downloading a programme that they presumably had the right to watch {TV licence and cable/satellite subscription fees up to date; I am guessing that the cable/satellite TV companies insist to see your TV licence before they'll let you sign up, and can cut you off in a heartbeat if necessary. Sky and NTL certainly do} any different than recording it on a VCR / DVD+RW / Sky Plus / any other kind of TV-recorder, at the time of broadcast?
The needs of the many outweigh the needs of the few.
Users outnumber developers.
Therefore, the needs of users outweigh the needs of developers.
You may have done something socially useful by writing a piece of software, but that does not mean you are automatically entitled to any kind of reward. Does your water company pay you for the shit you flush into their sewers? No, but you'll get fined if you try doing your business out in the street. Does your council pay you for all the bottles, cans and papers you recycle? No, but {at least if you live in Hertfordshire} you'll get fined if you put them in the rubbish.
The choice is almost never between paying for something or "stealing" it. There is almost invariably a third option, which is doing without. However, the marketing people are so convinced that their product is indispensable, that they consider it inconceivable that anyone would choose to forgo their product altogether. {Of course, doing without a DVD hurts the movie studios every bit as much as watching it without paying. But it doesn't get their backs up so much, because their heads are so far up their own arses that they don't notice this kind of lost sale}.
Let's consider a group of mates. {Mostly} honest, working-class people. They are closely-knit, often doing things together -- dinner parties, barbeques, minding one another's kids, that sort of thing. Not necessarily always the whole group together. Their means are modest, such that the purchase of a DVD is something that requires a good deal of evaluation and may well ultimately be eschewed in favour of something else.
If any one of them buys a DVD of a film that they all like, they will all end up watching it {either by borrowing it in turn, or by having a viewing party}. Each method has its own problems: in the first case there is a risk that the DVD might be returned late, in unsatisfactory condition or not at all; and in the second case, there is an obligation to supply food, drink and drugs and a risk of damage to the home. And both methods carry a slight risk that the purchaser might not like the film after all. Although these people are generally good friends and trust one another, these are real practical considerations given the costs involved and will undoubtedly figure in the pre-purchase evaluation.
So a group of people who can ill afford a DVD {costing somewhere between 80 and 100 fags} that they may or may not like, have to engage in much soul-searching -- and in the end might well decide to wait for it to come around on TV. If one of them has Sky {or a relative with Sky and a VCR/DVD+RW} then this is a more promising path. Waiting costs nothing but time, which is plentiful when you don't have much money, and they have already paid for the subscription anyway.
This presents a window of opportunity for an entrepreneur, distributing independently-sourced copies of feature films and passing on the cost savings {from things like, not having paid any royalties to the studio} to customers. These "Pirates" can offer their product more cheaply than the legitimate outlets despite costing significantly more to manufacture. Contrast this situation with the printed word. Many newsagents' shops have photocopiers; but the total cost {including non-monetary factors} of copying even one article from a newspaper or magazine usually exceeds the purchase price of the paper. Likewise, scanning and OCRing a Harry Potter book and making it available for download would be a significant effort. Whoever downloads it has either to print it out, or make do with reading it on-screen; and it's a PITA when it costs so little to buy the "real thing". {Of course, if someone has a point to prove, they'll be prepared to go through with it. But they wouldn't have bought the original anyway.}
The price of DVDs is a psychological barrier to purchase. If a DVD cost £4.99, people would be less likely to think too hard about the implications and more likely just to buy it, on the grounds that if it goes wrong it's "only a fiver". {A team of experts probably have already worked out the exact threshhold for impulsive purchases.} Out of our hypothetical group {but be honest, you do know people who are just like that}, it's far more likely that at least one person will buy the movie rather than everyone waiting for it to show up cheaper elsewhere.
So my plan to stop piracy is this: Make DVDs cheaper and not only will you sell more of them, you will end up making more money than before because people will be buying them who ordinarily would not have.
Copy prevention is mathematically impossible. Not just supremely difficult, like cracking RSA encryption; actually impossible. Like perpetual motion machines or faster-than-light travel. And limited-read media, by virtue of the fact they are as susceptible to copying as any other media now known or ever to be invented, do nothing to prevent illegal copying.
There was a bar that I used to drink in, back in my student days, which had a juke box. An NSM Prestige 160 if you care about these things; a lot like a Seeburg inside. It cost 10 pence a record {remember records?} and it was always playing. Once a fortnight, the amusement machine company came out to change the records. Well, one time, not only did they put in a whole load of new records, they also cranked up the price from 10p to 20p. And from that day on, the bar was like a Wetherspoons.
I guess the point I'm trying to make is that if DVDs cost £3 each instead of £20, then more people would be more prepared to buy them; and they'd actually sell enough copies to make more of a profit. Instead of waiting to see if one of my friends bought a movie I would like to watch {in a kind of "chicken" game, where the loser is the one who actually buys the disc and then has either to lend it out to everyone else, thereby risking the disc becoming trashed; or invite them over for a viewing, thereby risking an enormous cost in drink, drugs, broken furniture and freaked-out neighbours} we could all just buy our own copy of the disc, and not have to worry about the intricate politics of the situation. Likewise, there would be next to no market in "piracy", since the margins involved would be ridiculously small. Back in 1998-99, a "pirated" music CD cost £3 {handwritten track listing, labelled with indelible marker} or £4 {inkjet printed cover artwork and label}. Writers were rare, not much faster than 4* or 8* and hardly anybody had ADSL. As a cottage industry, it was fine for awhile but it soon became unsustainable.
When you think about it, we all know that the typical Linux installation is far less resource intensive than Windows, so why don't they sell it on their lesser hardware?
I'm not sure that's true anymore. Sure, Linux will run on a '486SX with 16 megs of RAM. And handle all the usual stuff like apache and sendmail. But Windows 98SE will run on that same machine and give you a GUI, and even run Microsoft Office 2000. Slowly, slowly, I'll grant you; but there's a perception about that shiny flashy graphics are somehow always easier to deal with than simple text-mode commands. And while I believe that is false, I also know that offering lower-spec hardware with Linux is not going to do anything to challenge that perception; in fact, it will only serve to reinforce it. Most people are clueless and just want a machine with big numbers, under the impression that it must be better {car analogy: they only care about engine cc's, not how far it will go on a litre of fuel}. You're essentially making out that a lower spec machine is only good enough for running Linux, not good enough for Windows.
What eats resources is the X11 windowing system. {Though object-oriented, interpreted languages -- such as the JavaScript embedded into web browsers -- probably don't help much either.} It used to be that KDE was horribly bloated, but GNOME is no longer a lightweight alternative. Of course there are less resource-intensive desktops {my favourite, which I will be using in my own distro, is WindowMaker} but most people are expecting a Windows XP clone. Hence, KDE or a heavily-customised GNOME.
I'm sure that you could create a display server optimised for applications running locally on a desktop machine with a single monitor {most people's configuration} and it probably would be less resource-intensive. But would it really be worth it? Who is the intended market? The people that are running older hardware generally know what they are doing. There are still a few '486 and first-generation Pentium boxes in every co-lo; and they churn out web pages and e-mails that are viewed on machines with ten times the RAM and twenty times the processor speed.
If it is a transformer power supply, then a small amount of energy is used in remagnetising the steel core once each half-cycle while the switch of the wall socket is on. If it is a switched mode power supply, then a small amount of energy is wasted due to leakage in the capacitor {which of course will only increase over time; until the metal film resistor in series with the diode bridge finds new - but brief - gainful employment as a fuse}. Again, only as long as the switch is on at the wall.
Although plug-in PSUs do get warm {which obviously indicates energy wastage} it's still nothing even close to the wastage of manufacturing, delivering and disposing of batteries {mains is already here and doesn't need removing when done with}. And, since I buy my own meter tokens, I know how much electricity I am getting through.
I have had electricity coming out of the power points on my walls for a lot longer than that!
Seriously, I make a point never to use disposable batteries if I can avoid it. Fossil fuels suck balls, but mains power is still more efficient than any battery technology will ever be.
There are two reasons why this product will not take off. Firstly, it will be illegal in Europe, where manufacturers have a responsibility to make sure their products are recycled when finished with; and secondly, it won't work anyway.
I invented a one-time-play audio cassette in 1979; but even before I had gathered together the bits to build the first prototype, I thought of a crack. And the crack I thought of was not one that could be "invented around". The limitation was a fundamental one with the universe itself, and so I never proceeded with the idea. Also, I had no idea that it would ever be useful for anything more than a practical joke.
There are two methods to ensure one-time-play, and two corresponding methods to defeat it. The easy one is destructive read-out; that is, the process of reading the data from the media also erases it. Or, more likely, the data is erased soon after being read. In either case, the corresponding crack is just to prevent the erasure from happening. My cassette idea {it was long enough ago for any patent I might have taken out to have expired by now, so I'm quite safe telling you this} involved a small permanent magnet downstream of the playback head; and the crack was to ensure that the tape never got as far as that magnet, by pulling out a loop of tape between the capstan/pinch roller assembly and the take-up spool.
The hard one is recorded read-out; that is, the fact that the disc has been read is remembered somewhere. This time, the crack is to cause the fact that the disc has been read, not to be remembered. If the record is kept in the player, then any other player should be able to read the disc {unless the disc carries the serial number of the player in which it is intended to be used; in which case it would be necessary to change the serial number stored in the player to match the one stored on the disc}. If the record is kept on the disc, then we effectively have a trivial emulation of destructive read-out {where the data itself remains intact and only a "permission to play" flag is altered}. The crack would be either to prevent the disc from being marked as played, or to prevent the "played" marking from being read.
Actually, there's a third reason why this scheme is doomed..... Almost nobody will want to buy a disc that you can only watch once, especially not if they will need to buy a new player just to watch it. After all, DVD players -- region-free with tape-recordable outputs -- are already available for £30.....
Slashdot Mirror
The issues are nothing to do with the schema itself, but rather to do with openness. The OpenOffice.org data format was conceived so that anybody who cares can write applications that speak it, as a right. By contrast, the Microsoft format is closed. If you want to write an application that speaks it, you have to ask Microsoft; they can charge you money for telling you, withhold bits if they see fit, and withdraw the privilege anytime. And if you do anything that Microsoft told you not to do, they can punish you. You can expect the problem with Microsoft's format that only Microsoft -- and a chosen few appointed by Microsoft -- are allowed to write programs that can retrieve your data once it has been saved in Microsoft's proprietary format. OpenOffice.org's format is better because any competent programmer can help you to retrieve that data, without being beholden to anyone. Any data that belongs to you rather than to Microsoft. That is not the problem. The problem is if, five or ten years down the line, you decide for some reason to move away from Microsoft. There are any number of reasons why you might want to do that: for argument's sake, let's say MS have kept cranking up the cost of Office to the point where you now have to decide whether to try to save money on software licences or lay off staff. Now someone else's document converter may well not be able to handle Microsoft's proprietary format correctly. Your data might become inaccessible! There is also a very real possibility that Microsoft may not exist 10 years from now, and they may take their proprietary formats to the grave with them.
In five, fifty or a hundred years, any competent programmer will still be able to obtain the schema which will enable them to make sense of an OpenOffice.org document, because no one person or organisation controls that schema. No such guarantee can be made in respect of Microsoft's schema.
Or, let me put it this way. Imagine you buy a new car. The bonnet is fastened shut with a tamperproof seal, so only authorised dealers can make repairs -- and they have to use the manufacturer's original specified parts and procedures. You have to buy petrol from the manufacturer's specified filling stations {who will check from time to time that you haven't been tampering with things that do not concern you}. When the car reaches the end of its life {which may come sooner than you think, since the manufacturer can order their service centres not to repair it on a whim} you have to replace it with another one from that same manufacturer; otherwise everything and everybody you ever carried in that car will be left in limbo somewhere, and not fit properly in your new car.
- pay a software vendor for a worthless piece of paper {"This may or may not do what you wanted, and we couldn't give a flying toss one way or the other; but if you try to poke about inside it to find out, we will rip you limb from limb and feed you to the hounds"}; or
- pay a competent programmer, independent of the original authors and having nothing to lose whatever they say, to determine the suitability of a product for a particular application by inspecting the source code?
In effect, the above is already what most EULAs say anyway. It's just a matter of time before someone gets burned by closed software, tries to sue a major software vendor and calls the whole setup into question. I'm surprised it has not already happened, but maybe it has and they have been hushed up {bought off and given a stern warning not to talk about it ever againMy guess is that source code would stand up much better as a defence in court than an EULA, because source code does not actively seek to diminish your statutory rights. And if a bug is obvious to anyone in the courtroom from looking at the source, then it should have been obvious to the plaintiff or their agent in the first place.
I block all kinds of advertisements.
I block TV advertisements -- when I'm not watching the Beeb, of course -- by getting up and taking a leak, or brewing a cuppa, or skinning a d00b, or doing anything else I can fit into five minutes. And muting the sound for good measure. {If I'm recording the show I'm watching, I'll stick a chapter marker after the break, so that whoever watches it later can just skip straight past the crap.}
I block magazine advertisements by skipping the pages with the adverts on. Though if the truth be told, I am highly selective with buying magazines. I generally prefer just to rotate myself through several newsagents, browsing without purchase though never for long enough to elicit unwanted attention -- it's a skill I taught myself many years ago. The publishers can't be too worried about it anyway; they are obviously losing less money through people like me than the couple of pence per copy it would take just to seal their magazines inside an envelope.
I block internet advertisements using Squid. It's probably overkill if the truth be told, but it's not hard to set up.
As for why I do it, that's easy. I am the sole judge of what I view. I have already decided that if you feel the need to shove your product in my face, then I will do my damnedest to avoid buying it. I do not owe advertisers anything.
I think this -- randomly-present allergens rather than deliberately-placed poisons -- is really more like the software situation. If we're being kind to the closed-source developers, the bugs are there by accident, not design. But even so, the closed source vendors are still effectively saying "This chocolate bar categorically does not contain any kind of nuts whatsoever and you'd better believe us, we will punish you if you try to check with a matter analyser". Because someone managed to persuade a judge that two wrongs make a right. Still, it opened up the way for the families of all those people who had been killed by gu
Any idiot can go to B&Q and buy some copper pipe, fittings, solder, flux and a blowtorch. If their self-installed pipes leak, or they manage to set their bathroom on fire, or they drip hot solder on themself
Open Source software is not necessarily free as in gratis, because you may well have to pay to get the code checked over. But that's still not a service you are ever likely to get at any price from the vendors of closed-source software; they have their own reasons for keeping their secrets which exclude all possibility of independent audit.
Now, we still have a few gaps, including that small things appear not to behave the same way as big things. No doubt, if we can quantify the differences* -- or explain why that would be impossible -- we can take a stab at a single Grand Unifying Theory which would underpin all of Physics.
It's also possible that there could be another possible set of laws of physics which would be mutually consistent, even consistent with the G.U.T., just contrary to all our observations. If there existed a parallel universe which obeyed this set of laws, one of four things could happen:
- It would collapse to a single point in our space
- A single point of space in that universe would be bigger than the whole of this universe
- It would exist for only a brief instant of our time
- A single instant of time in that universe would last longer than the lifetime of our universe
Of course, it's also possible {but extremely unlikely} that there is no Grand Unifying Theory, just a supreme being with a sick sense of humour who keeps changing the rules slightly every time we get close to discovering what they are* Canonical example of difference between quantum and classical phenomena: Why can't a chair just spontaneously shift position? My own take is that quantum wave functions do exist in large systems, but "quantum" phenomena are not generally observed because the waves are not coherent {just as you don't see interference fringes where the light from two candles falls on the same surface}.
I'm not sure that will ever work. Plastic coins, RFID or not, would be too prone to forgery. The devices are easy enough to spoof, and the whole point with coins is that they are used in machines in unsupervised locations. It doesn't matter that you need a rucksack full of gear to simulate a few pound coins, if nobody is going to see you doing it.
Forging actual metal coins is difficult, not impossible but difficult and expensive, and usually not worth it for the returns involved. Better quality requires more investment, which can only be recouped by doing a bigger run; but larger quantities increase the risk of discovery. Forging RFID is trivial and inexpensive by comparison -- and you just know somebody is going to want to have a go at it.
If there is an upside to knowing the history of every banknote, it's this: shops could erect signs saying MONEY WON IN LAWSUITS IS NOT WELCOME IN THIS STORE. And mean it.
Use coins. I already do anyway. The authorities must think I have a massive gambling habit, but really I'm just going into amusement arcades to change serial-numbered notes for unnumbered coins. Coins, being made of metal, cannot have RFID devices embedded in them. Radio waves will not travel through anything that conducts electricity {this is a fundamental limitation of the universe and cannot be overcome by invention}. If you are really paranoid, you can test each coin for conductivity in several places using a simple home-built device {a store-bought AVO may have been rigged}.
The source code is a guarantee of functionality: it is an irrefutable statement of exactly what the program will and will not do if it is compiled and run on a properly-working computer. Of course it is written in a language that not everybody understands; but, to be blunt about it, that's their problem. For instance, if I writethen any sufficiently-competent Perl programmer should know that this program will not print "Hello World!" on a Saturday; and anybody else who is unsure but wants to find out anyway should engage the services of a sufficiently-competent programmer. {They might well be charged money for this service; but just because we are giving our give our source code away certainly does not mean that we shouldn't be allowed to charge money for explaining what other people's source code does to people who don't understand it.}
But they wouldn't be liable, though! If it was Open Source, they would be giving away the source code. Then it's up to the recipient to make an informed decision {possibly engaging the services of an independent third party} as to whether or not the goods are fit for purpose. Without the source code, they would not be able to make a fully informed decision, and so have to take it on faith that the vendor is being truthful. That's the crucial difference. The source code is a guarantee in its own right.
If the new kernel you installed on your server came with source code, it would be your responsibility to read that source code and check that it did what it was supposed to do.
If you buy a TV stand in kit form, you have the opportunity to decide, at the moment that you build it, whether or not it will suit your intended purpose and if not, what modifications you can make to ensure that it will. For instance, it might come with 3x35 screws: upgrading them to 4x50 and using plenty of hot melt adhesive might allow it to support an extra 10kg., which might make the difference between supporting your nice new widescreen set with NICAM and fastext, and not.
The point is, it's up to you to decide whether goods furnished in kit form will be suitable for the intended application. All the manufacturer is responsible for is making sure that all the parts can be fitted together in such a way as it looks like the picture on the box. If you want to be able to sue someone if your TV stand falls apart and your TV falls on the floor, then get an odd-job person to build it for you, and check their liability insurance to make sure you are covered.
As long as software supplied as source code was treated analogously to goods supplied in kit form, I can see absolutely no problem with obliging software suppliers to offer a guarantee of performance. If you give someone some source code files, you can confidently say "I can guarantee you that if you compile this source code and run it on a computer which is working properly, it will do exactly what the source code says it will do".
I remember, back in the 1970s, there was an appliance known as the "radio-cassette recorder". This was transportable {i.e. it had a handle} and allowed you to enjoy listening to the wireless {MW, LW and FM -- FM stereo on the bigger, more expensive models with two speakers} or your favourite audio cassettes, anywhere there was a power point. {If you put batteries in the thing, you didn't even need the power point, and you could listen to both sides of a C-90 on a single set. Supposedly. Made the machine even heavier to carry, though.} But these devices did not only allow you just to listen to music: you could also record from the radio onto a blank cassette.
.....
Nobody ever had a problem with the existence of radio cassette recorders. Everyone used to record their favourite songs from the radio. Sometimes we'd even go out and buy a "45" if we liked the song enough. Ah, the 1970s and 80s were wonderful times if you were a kid
I am allowed to invite my friends into my home to watch my TV, even if they have not subscribed to the channel I am watching -- in fact, even if they do not have a TV licence. And I'm fairly sure I am allowed to lend my home recordings to my friends -- I've certainly never heard of anyone being busted for that, and the practice is almost universal.
The needs of the many outweight the needs of the few, but the needs of the few outweigh the whims and caprices of the many. And even if your assertion were correct {which I severely doubt, given my better than 3:1 fan to freak ratio} it still falls into the "whims and caprices" category rather than the "needs" category.
MySQL does what it does well, and doesn't pretend to do what it doesn't do.
As a flexible but quirky set of persistent array extensions for most known programming languages, including the Three P's, it works really well. Unfortunately, the fact that it uses a bastardised dialect of SQL has led people to mistake it for a relational database management system.
MySQL makes it easy to write web pages with changeable content {such as message boards, diaries, online auctions and personal rant sites}; and with the addition of phpmyadmin, it even becomes a kind of MS-access replacement. You will have to do a lot more work in the application layer if you want to emulate stored procedures and triggers, or if you don't want to run afoul of graceful degradation. But in the 90% of cases where you don't need SPs and triggers, and where you'd rather let a few characters go missing than seize up with a fatal error, it performs just fine. And because it uses SQL {albeit with a broad regional accent}, it can provide a n00b with a sort of gentle introduction to real databases.
The bottom line is, you can't expect to tow a four-berth trailer round the twisty mountain roads with a one-litre Ford Fiesta. But for someone who just wants to drop the kids off at school and then nip into town to buy another pair of shoes, a 4x4 with a three litre engine and six gears is overkill. And you don't see many high performance sports cars with L-plates either.
The fundamental difference between RPM and DEB is that RPM packages depend on specific files while DEB packages depend on specific packages. Debian packages can also implicitly provide other packages. So, there is a Debian "virtual package" called "www-browser". It doesn't contain any files, but it has dependencies which will be satisfied by any of links, lynx, firefox, konqueror, mozilla, &c. In turn, the browser packages claim to provide "www-browser" when any one is installed. If you have some application that requires something that implements "sendmail" functionality {exim, postfix and so forth all call their binary "sendmail" and use the same command line options} then you need only depend on the "mail-transport-agent" package.
Portage is different again. I haven't used it too much. By the time Gentoo arrived on the scene, I was already fluent in Debian, so my experience is limited to an experimental "just for the hell of it" installation. I was impressed with what I saw, though: with a GUI front end, I think it would have the potential to make the whole business of downloading and compiling from source as painless as InstallShield but without the spyware. I've resolved to do some serious GTK-perl hacking at some stage in future, so I'll see if anything comes of that.
Whether APT {which Conectiva bodged to work with RPM packages: it will be interesting to see whether they stick with Mandrake's urpmi tool, which is an APT work-alike} and Portage are actually superior to RPM is not certain, because both Debian and Gentoo feature enormous package repositories including pretty much the whole of Open Source; it's like comparing a few apples to several kilos of oranges.
By the way, circular dependencies can be solved very easily, but it took me ages to figure it out. What you have to do is specify both package names in the same command.
I can name one feature I'd like to see in Mandriva.
Either Debdrake or Portagedrake.
RPMs are bollocks - either that, or every RPM distro that I've used or seen had a package repository that was bollocks, which amounts to the same thing either way: bollocks. Anyway, back in the days when I was using Mandrake 8.2, one of the first things I learned was how to spell "make install".
How exactly is someone downloading a programme that they presumably had the right to watch {TV licence and cable/satellite subscription fees up to date; I am guessing that the cable/satellite TV companies insist to see your TV licence before they'll let you sign up, and can cut you off in a heartbeat if necessary. Sky and NTL certainly do} any different than recording it on a VCR / DVD+RW / Sky Plus / any other kind of TV-recorder, at the time of broadcast?
The needs of the many outweigh the needs of the few.
Users outnumber developers.
Therefore, the needs of users outweigh the needs of developers.
You may have done something socially useful by writing a piece of software, but that does not mean you are automatically entitled to any kind of reward. Does your water company pay you for the shit you flush into their sewers? No, but you'll get fined if you try doing your business out in the street. Does your council pay you for all the bottles, cans and papers you recycle? No, but {at least if you live in Hertfordshire} you'll get fined if you put them in the rubbish.
The choice is almost never between paying for something or "stealing" it. There is almost invariably a third option, which is doing without. However, the marketing people are so convinced that their product is indispensable, that they consider it inconceivable that anyone would choose to forgo their product altogether. {Of course, doing without a DVD hurts the movie studios every bit as much as watching it without paying. But it doesn't get their backs up so much, because their heads are so far up their own arses that they don't notice this kind of lost sale}.
Let's consider a group of mates. {Mostly} honest, working-class people. They are closely-knit, often doing things together -- dinner parties, barbeques, minding one another's kids, that sort of thing. Not necessarily always the whole group together. Their means are modest, such that the purchase of a DVD is something that requires a good deal of evaluation and may well ultimately be eschewed in favour of something else.
If any one of them buys a DVD of a film that they all like, they will all end up watching it {either by borrowing it in turn, or by having a viewing party}. Each method has its own problems: in the first case there is a risk that the DVD might be returned late, in unsatisfactory condition or not at all; and in the second case, there is an obligation to supply food, drink and drugs and a risk of damage to the home. And both methods carry a slight risk that the purchaser might not like the film after all. Although these people are generally good friends and trust one another, these are real practical considerations given the costs involved and will undoubtedly figure in the pre-purchase evaluation.
So a group of people who can ill afford a DVD {costing somewhere between 80 and 100 fags} that they may or may not like, have to engage in much soul-searching -- and in the end might well decide to wait for it to come around on TV. If one of them has Sky {or a relative with Sky and a VCR/DVD+RW} then this is a more promising path. Waiting costs nothing but time, which is plentiful when you don't have much money, and they have already paid for the subscription anyway.
This presents a window of opportunity for an entrepreneur, distributing independently-sourced copies of feature films and passing on the cost savings {from things like, not having paid any royalties to the studio} to customers. These "Pirates" can offer their product more cheaply than the legitimate outlets despite costing significantly more to manufacture. Contrast this situation with the printed word. Many newsagents' shops have photocopiers; but the total cost {including non-monetary factors} of copying even one article from a newspaper or magazine usually exceeds the purchase price of the paper. Likewise, scanning and OCRing a Harry Potter book and making it available for download would be a significant effort. Whoever downloads it has either to print it out, or make do with reading it on-screen; and it's a PITA when it costs so little to buy the "real thing". {Of course, if someone has a point to prove, they'll be prepared to go through with it. But they wouldn't have bought the original anyway.}
The price of DVDs is a psychological barrier to purchase. If a DVD cost £4.99, people would be less likely to think too hard about the implications and more likely just to buy it, on the grounds that if it goes wrong it's "only a fiver". {A team of experts probably have already worked out the exact threshhold for impulsive purchases.} Out of our hypothetical group {but be honest, you do know people who are just like that}, it's far more likely that at least one person will buy the movie rather than everyone waiting for it to show up cheaper elsewhere.
So my plan to stop piracy is this: Make DVDs cheaper and not only will you sell more of them, you will end up making more money than before because people will be buying them who ordinarily would not have.
Copy prevention is mathematically impossible. Not just supremely difficult, like cracking RSA encryption; actually impossible. Like perpetual motion machines or faster-than-light travel. And limited-read media, by virtue of the fact they are as susceptible to copying as any other media now known or ever to be invented, do nothing to prevent illegal copying.
There was a bar that I used to drink in, back in my student days, which had a juke box. An NSM Prestige 160 if you care about these things; a lot like a Seeburg inside. It cost 10 pence a record {remember records?} and it was always playing. Once a fortnight, the amusement machine company came out to change the records. Well, one time, not only did they put in a whole load of new records, they also cranked up the price from 10p to 20p. And from that day on, the bar was like a Wetherspoons.
I guess the point I'm trying to make is that if DVDs cost £3 each instead of £20, then more people would be more prepared to buy them; and they'd actually sell enough copies to make more of a profit. Instead of waiting to see if one of my friends bought a movie I would like to watch {in a kind of "chicken" game, where the loser is the one who actually buys the disc and then has either to lend it out to everyone else, thereby risking the disc becoming trashed; or invite them over for a viewing, thereby risking an enormous cost in drink, drugs, broken furniture and freaked-out neighbours} we could all just buy our own copy of the disc, and not have to worry about the intricate politics of the situation. Likewise, there would be next to no market in "piracy", since the margins involved would be ridiculously small. Back in 1998-99, a "pirated" music CD cost £3 {handwritten track listing, labelled with indelible marker} or £4 {inkjet printed cover artwork and label}. Writers were rare, not much faster than 4* or 8* and hardly anybody had ADSL. As a cottage industry, it was fine for awhile but it soon became unsustainable.
What eats resources is the X11 windowing system. {Though object-oriented, interpreted languages -- such as the JavaScript embedded into web browsers -- probably don't help much either.} It used to be that KDE was horribly bloated, but GNOME is no longer a lightweight alternative. Of course there are less resource-intensive desktops {my favourite, which I will be using in my own distro, is WindowMaker} but most people are expecting a Windows XP clone. Hence, KDE or a heavily-customised GNOME.
I'm sure that you could create a display server optimised for applications running locally on a desktop machine with a single monitor {most people's configuration} and it probably would be less resource-intensive. But would it really be worth it? Who is the intended market? The people that are running older hardware generally know what they are doing. There are still a few '486 and first-generation Pentium boxes in every co-lo; and they churn out web pages and e-mails that are viewed on machines with ten times the RAM and twenty times the processor speed.
If it is a transformer power supply, then a small amount of energy is used in remagnetising the steel core once each half-cycle while the switch of the wall socket is on. If it is a switched mode power supply, then a small amount of energy is wasted due to leakage in the capacitor {which of course will only increase over time; until the metal film resistor in series with the diode bridge finds new - but brief - gainful employment as a fuse}. Again, only as long as the switch is on at the wall.
Although plug-in PSUs do get warm {which obviously indicates energy wastage} it's still nothing even close to the wastage of manufacturing, delivering and disposing of batteries {mains is already here and doesn't need removing when done with}. And, since I buy my own meter tokens, I know how much electricity I am getting through.
I have had electricity coming out of the power points on my walls for a lot longer than that!
Seriously, I make a point never to use disposable batteries if I can avoid it. Fossil fuels suck balls, but mains power is still more efficient than any battery technology will ever be.
There are two reasons why this product will not take off. Firstly, it will be illegal in Europe, where manufacturers have a responsibility to make sure their products are recycled when finished with; and secondly, it won't work anyway.
..... Almost nobody will want to buy a disc that you can only watch once, especially not if they will need to buy a new player just to watch it. After all, DVD players -- region-free with tape-recordable outputs -- are already available for £30 .....
I invented a one-time-play audio cassette in 1979; but even before I had gathered together the bits to build the first prototype, I thought of a crack. And the crack I thought of was not one that could be "invented around". The limitation was a fundamental one with the universe itself, and so I never proceeded with the idea. Also, I had no idea that it would ever be useful for anything more than a practical joke.
There are two methods to ensure one-time-play, and two corresponding methods to defeat it. The easy one is destructive read-out; that is, the process of reading the data from the media also erases it. Or, more likely, the data is erased soon after being read. In either case, the corresponding crack is just to prevent the erasure from happening. My cassette idea {it was long enough ago for any patent I might have taken out to have expired by now, so I'm quite safe telling you this} involved a small permanent magnet downstream of the playback head; and the crack was to ensure that the tape never got as far as that magnet, by pulling out a loop of tape between the capstan/pinch roller assembly and the take-up spool.
The hard one is recorded read-out; that is, the fact that the disc has been read is remembered somewhere. This time, the crack is to cause the fact that the disc has been read, not to be remembered. If the record is kept in the player, then any other player should be able to read the disc {unless the disc carries the serial number of the player in which it is intended to be used; in which case it would be necessary to change the serial number stored in the player to match the one stored on the disc}. If the record is kept on the disc, then we effectively have a trivial emulation of destructive read-out {where the data itself remains intact and only a "permission to play" flag is altered}. The crack would be either to prevent the disc from being marked as played, or to prevent the "played" marking from being read.
Actually, there's a third reason why this scheme is doomed