ROFL, the USPTO was experiencing the/. effect before we slashdotted it - they're already way too busy taking care of stupid patents like this one.
Very true. The volume of patents submitted by MS every week would probably constitute a DOS attack from any who didn't own so many government officials.
But not words representing numbers, written in many languages.
Possibly not. If they'd submitted a patent on "I18N for numeric pattern matching" I doubt we'd be discussing this. As it is they have to try and drag in the display aspect so they can threaten anyone who makes a browser, help file viewer, or even the folks who maintain "less".
But even if they had numeric internationalized pattern matching I still wouldn't condone it.
So are you saying that the USPTO is doing a satisfactory job?
No. As stated in my sig, unless I explicity say something, don't assume or infer it.
Look at my comment again. See the question mark? That means I'm asking, not assuming. That's what you said I should do, remember? You could at least try not to get confused by your own rules.
I am making no comment whatsoever on whether the USPTO is doing anything effectively or not. All I said was that your argument holds only if it's actually true. I never conceeded it is true. Not conceding a point is not the same as asserting the opposite. It's like "NULL" in SQL.
OK, that's clear. Kind of pointless, but clear enough.
Mine and me. Don't like it? Tough.
Now that's an unwarranted assumption: Who said I didn't like it? You're as entitled to an opinion as the next man - it just doesn't count for very much unless you can back it up.
Stop arguing with me then.
mmm... you seem to be running out of things to say, and I'm sure we both have better things we could do with our time. Maybe it's about time we laid this one to rest.
Try reading ahead a sentence or two to see if a point is eleborated upon before replying. In this case, it was.
Well then perhaps if you kept these points together, people wouldn't need to make assumptions nor to draw inferences concerning what you may have meant. As it is I did what you say you prefer and requested clarification, and you still get all snippy about it. No pleasing some people, I suppose.
I never agreed that it was actually true, i.e., "All that means is... assuming your assertion were actually true."
So are you saying that the USPTO is doing a satisfactory job? I don't think even the USPTO think would go that far. They are overloaded, undermanned, lack quality technical minds and have problems retaining staff meaning that a disproportionate number of patents are being examined by inexperienced staff. No disrespect to the patent office, but I don't see how they can be doing a proper job under the circumstances.
I
never said nor implied
you did. That wild assmption is entirely your own doing.
And I never said nor implied that you said or implied any such thing. All I was doing was
offing you the courtesy of persuing the question should you have felt that I was evading the issue. For someone who gets so uptight about unwarranted inferences, you certainly seem to draw your share.
If anything, you're guilty of not being able to form sound arguments.
Irrelevant to what and how? I just challenged one of the assertions supporting your argument - if you find that irrelevant it suggests a very convenient definition of relevance.
All that means is that the patent office is doing a bad job, not that there's inherently anything wrong with software patents.
Splendid! Then you presumably conceed that small software houses are not enjoying any particular protection from the large houses under the current regime at the USPTO and that, as a minimum patent office procedures are in need of an overhaul. Perhaps you'll go so far as to grant that the benefits of software patents are not so "blindingly obvious" as you may have suggested given that the system is capable of such abuses, and that these benefits
are more theoretical than actual at this
point in time.
Now all we need is a robust method for discriminating between good and bad software patents, a means whereby the USPTO can deal with current flooding tactics and some sort of safeguard to prevent this from happening again and I might almost be ready to conceed the point. I'm sure the USPTO will find your proposals fascinating. I know I will.
That's all irrelevant.
I'm talking about "... anything else that is already patentable."
Silly me. Software patents should be allowed because everything that is already patentable is patentable and therefore so should software be patentable. That's as if I were to say the government should give me a million dollars because everyone else who has already been given a million dollars by the government has been given a million dollars by the governement and therefore, I should be given a million dollars too.
Is that what they call a circular argument do you suppose?
But I believe your case is FUD
Just to be clear: My comment shouldn't be taken
as accusing you of maliciously spreading FUD.
I'm merely suggesting that you may have been
less than critical the last time you were exposed to it. Also FUD is a woefully imprecise term to
use in a debate as picky as this one. Let's say propaganda, since that's closer to the meaning
I intended.
Of course, with those caveats, if you still think
I have a hidden agenda, I'd love to know what
it might be and who you see as benefitting.
so, as far as I'm concerned, you have yet to make any compelling argument.
Well, clearly I've failed to compel you.
On the other hand, your best argument so far in support of your case boiled down to "why not" - a tactic that not only lacks compulsion, but is somewhat juvenile to boot.
Alas, "why not" frequently seems to be the final line of defence for the pro-swpats crowd. I do hope you have something better waiting in the wings - I was rather hoping you might offer something intelligent.
Then let's say I'm a programmer and I invent a new algorithm for sorting video files (or whatever). It seems to me that my new algorithm is really no different than a machine and should be patentable.
Well, you can take the pragmatic viewpoint that patents tend to stifle innovation in software, and anti-competitive and demonstrably unnecessary for purposes of turning a profit. On the other hand I've yet to see a decent argument as to why software should be patentable.
So given that swopats have bad effects, and given that we are talking about what should be done - does it not make sense to outlaw them? I think I'd like a better argument than "it works for tractors" before I endorse the notion.
The benefit for small software houses is if they're the ones to obtain their own patents. This way, MS or Amazon can't come along and just steal their ideas.
That turns out not to be the case. Software patents are being granted so broadly and freely that the big houses can easily find a patent that is arguably infringed by the small house. That done, they can use the threat of litigation to force the small house to licence the patent for a trivial sum, or refuse to licence the patent at all and bankrupt the smaller player. The protection only applies if you can trade patents one for one with the big players.
I would think this reason would be blindingly obvious.
It's FUD put out by the big software concerns. There's a reason why the small houses campaigned against swpats in the EU and the big corps lobbied for them. The reason is that they drastically favour the big boys.
That's the wrong question.
That remains to be seen.
I see nothing special about software to exlude it from receiving the same patent protections as anything else.
Which presupposes that everything else is indeed eligible for patenting. And yet this too turns out not to be the case. There are no patents on creative works; none on ideas; none on mathematical theorems or laws of physics.
So it seems that "everything else gets patented" doesn't work, and thus the question remains to be answered: Why should software enjoy the benefits of patents? I've made the case as to why it should not - now it's your turn.
Copyrighting software is useful only if you publish the source code itself.
That's just silly. You might as well say that copyrighting movies is only useful if you publish the script.
I think you'll find copyright extends to the binary compiled form of software. This is the form under which it is generally distributed commercially. That's why downloading unlicenced software is properly called copyright infringement, rather than "piracy" as a certain Mr. Gates would have us believe.
Speaking of Mr. Gates, you are aware that he became the richest man on the planet using copyright to protect his software? I'd say that argues quite strongly that copyright is adequate to the task of protecting and exploiting software. Would you not agree?
Clear now?
I do believe we're getting there, however slowly.
What you want to protect is what the software does and for that you need a patent.
Which rather presupposes that preventing others from writing competing packages is a good thing. I'm sure MS and Amazon, (to name but two) think so, but the benefits are rather less obvious for small software houses, consumers the free market and the world at large.
But if you're in the mood for injecting a little clarity into the debate, perhaps you could explain why society should grant this monopoly over software? What is it about software that you feel qualifies it for protection under patent? What does society get in return for granting this dispensation in the realm of software? You have yet to make that case.
*sigh* your software is already protected by copyright,
Nobody in the thread said that patents or copyrights are a fundamental human right
Excellent. Then with protection already present in copyright, there is no need of protection via patent. Therefore, since they threaten at least as much "IP" as they protect, and since they are not in fact necessary, and since there is noi fundamental entitlement, the sane conclusion would seem to be that they should be outlawed.
Where is what written down? The protection of copyright as applied to software or lack of an inalienable right to double a monopoly on ideas?
The supposed fact that you can't copyright something and have a patent on it simultaneously.
I think you will find that isn't what I said. I said that a) your software is already protected by software, and b) that there is no fundamental human right to a further monopoly, or (any monopoly at all come to that) on the idea.
My assertion that you can't get copyright and patent both is purely your own, incorrect, inference, Really, if you're going to try and pre-empt the rules of engagement you might at least practice what you preach.
Of course if you have a patent on something, you don't need a copyright since a patent affords stronger protection anyway.
I expect that's dreadfully convenient for certain large software houses who happen to be currently engaged spamming the USPTO with every bloody obvious idea they can think of as well. Especially considering how most of the "intellectual property" in the software business would seem to be vested in copyright,
That by itself seems like an excellent reason for
banning software patents.
Or do you propose to tell me that patents are justified because they protect against patents?
Where is what written down? The protection of
copyright as applied to software or lack of an inalienable right to double a monopoly on ideas?
If you're going to use a sig like that you're going to need to express yourself with much greater clarity.
Anyway, a UI is not copyrightable.
I said "software interface", not "interface". The software is copyrightable. if your concern is the colours and the fonts, might I suggest that trade marks might be the measure you seek?
Whether the software implementing the UI is copyrighted is irrelevant.
Irrelevant to what, precisely?
If you reply, do so only to what I explicitly wrote. If I didn't write it, don't assume or infer it.
If you post, express yourself with clarity and let your arguments stand on their own merits. Do not presume to instruct me as to how I may and may not frame my reply.
there likely isn't going to be enough revenue generated to support the team of developers needed to add all the boring stuff like... keeping up with the myriad tax changes from year to year
Sounds like a good case for a framework and a scripting language designed to be accountant-friendly. Geeks design the framework, gui, and report innterface, and you get an accountant, probably a retired one with an interest in IT to write the initial accountancy stuff and to consult on the script language design.
Do it right and you decouple the application design from the accountancy which means that the software userbase no longer depends upon the initial development team since they can hire accountants to write tax modules each year, or use ones released under a copyleft licence.
A bit of thought into modularity would help too. You could have a basic module for the more static, universal measures, and then have add-ons for people who need to take into account specific circumstances.
There'd even potentially be money in it, if you can get software support contracts for a few medium sized concerns. It sounds an interesting project.
So if I develop a hardware stereo (i.e., totally implemented in hardware as opposed to software) with a novel set of knobs and switches, it's OK to patent that, but if I use software to display the interface on a screen instead and use a very novel scroll wheel it's suddenly not OK?
Well, because your software interface is already protected by copright and you have no inalienable right to any further monopoly on the idea. In any event you don't need the patent to make money from your knobs so there's no need for it and you already have protection to exploit your idea, so that should be sufficent.
That is unless your requirements for sufficency extend to unfairly surpressing competition. Even so, it's far from "OK", IMHO.
Further, unless your hardware interface is staggeringly novel, then there is massive prior art on that too. Come back when you get the bugs out of a telepathic interface or something. And when you do, it'll be the telepathy chip that deserves the patent. The software that draws the pretty pictures on the screen will still not be "OK" for patenting.
By the way, there have been far dumber/worse patent snafus reported here. If they didn't expose said stupidity, what makes you think this rather benign instance will?
Have standards dropped so far at the USPTO that "medium-stupid or better" is all the qualifications an idea needs to be patentable? Just because we have the crimes of Jack the Ripper on record, that doesn't mean we should cease from complaining about cases of homicide; neither do past stupidities excuse this.
This book is sold subject to the condition that it, or any part of it, shall not by way of trade or otherwise, be sold, lent, re-sold, displayed, advertised or otherwise circulated, without the publishers' prior written consent, in any form of binding, cover or title other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchaser(s).
Sounds like a EULA to me. The difference is that it has never been feasible to enforce that meaningless boilerplate on individual purcahsers, with the result that most readers treat the text, to the extent that they consider it at all, with the contempt it deserves.
Oh, and of course it isn't fifteen pages long. but then this is print media and spending fifteen or so pages on needlessly obfscated license terms would cost the licensor money.
If you are a materialist, and I think that includes most people who consider themselves educated, there is no essential difference between atmospheric pressure and religion,
Well, it would help if you were to say which usage of the word you had in mind. Although if you find atmospheric pressure to be in the same class as religion, we can probably assume definition A)
The theory that physical matter is the only reality and that everything, including thought, feeling, mind, and will, can be explained in terms of matter and physical phenomena.
If so, would you be open to the idea that you are projecting your own viewpoint onto "educated people" as a whole? Materialism as defined above is a rather narrow and extremist worldview. It's
great for winding up your classmates at school, or drunks at a party, but taken exclusively it's rather too limiting to form the basis for any useful ontology.
In this respect, it has a lot in common with its philosophical mirror image solipsism in that it willfull invalidates fifty percent of human experience.
And, just as Samuel Johnson
refuted
Bishop Berkely, so it is almost trivially easy to refute materialism as a philosophy: it takes no account of information. I can communicate the number "four" to you in a number of ways: I can voice the word (in a variety of langauages); I can send a pattern of electrons over a computer network; I can hold up a number of fingers, I can inscribe a number of different patterns on paper or other substances. All these will send the ideal of "four" from my brain to yours, but the concept of "fourfoldness" is implicit in none of these things.
Then there is mathematics whereby it is possible to construct purely abstract systems with no reference or application to physical reality.
Going a degree more mystical, there is the question of preception: it's all very well to exlain my perceptions in terms of enzymes and electrons, the fact remains that there is an "I" to whom these perceptions are presented. There is even support for this in orthodox physics in the
consciousness causes collapse interpretation of Quantum Mechanics.
I could go on, but that's enough to make the point.
Summing up, materialism can be useful but any supporting arguments tend to be circular since they have to disregard all non-materialist evidence. This would be acceptable if there was anything approaching the consensus you suggest, but that seems unlikely in the extreme: try telling your wife/GF/S.O. that "love" is nothing but an emergent effect arising from a combination of hormones and sensory stimuli. The chances are you'll get a quick grounding in why approx. 50% of the population has a problem with that proposition.
The point I'm getting to here is that Materialism is a religious belief. One of the defining characteristics of religious beliefs is that they take certain propositions on faith, both in the absence of evidence, and in the absence of any test that might falsify those propositions.
Interestingly, this makes atheism into a religion, since just as the theist professes a belief in a God in the absence of evidence, so does the atheist profess disbelief in the absence of any evidence. Agnosticism has the benefit of withholding judgement until evidence can be presented. However it is difficult to reconsice Meterialism with agnosticism since materialism tends to reject "God" as a valid concept for debate, and it's hard to define agnostic without reference to "God".
So I wonder which is most important to you: a rationalist world view, or your materialist religion? As you start to think of it in those terms, it becomes difficult to reconcile the two increasingly antagonistic viewpoints.
I think the interesting thing lies not in Tim's answers, but in the tenor of the questions asked.
Mark Lawson seems to have been desperate to elicit some response along the lines of
The web is a terrible, terrible place! It was supposed to be all kittens and fluffy bunnies and instead all they use it for is identity theft and pornography! It wasn't meant to be like this!.
I could almost start formulating consipiracy theories about laying groundwork for increased censorship, except that, the tenor of the questions is nothing unusual for a newsnight interview.
Respect is due to Sir Tim for keeping his head and not rising to the bait.
Still, the political nature of the questions can be seen as reflecting the increasingly politicised nature of the web. I wonder if he's in for more of this sort of flack in the future.
First of all: respect to you for declaring an interest up front. If more of your colleagues did so
when entering this sort of debate, there'd be fewer accusations of astroturfing.
Talk smack - that's fine - but don't lump *all* of us in to one big package and label it "evil."
It's not *all* of you that we're concerned about. It's the guys at the top. The ones who set policy. Just a few examples:
We're concerned about the people who decide it would be good strategy to bankroll SCO.
We're concerned about the people who fund AdTI to produce such horribly
skewed and falacious reports. And we worry too about the
less glaringly incompetant bodies that they have working on similar projects.
We're concerned when Steve Ballmer threatens the whole of Asia with fines for violating "intellectual property" if
they adopt Linux
We're concerned by corporate pressure on OEMs to prevent them offering Linux preloaded at a competetive price and spec.
This is the sort of behaviour that prompts the use of the word "evil". We're have no axe to grind,
or at least I personally don't, against MS developers.
However you should acknowledge that your corporate agenda is set by people who are not friends of Free Software.
Thus, if you post on slashdot defending those policies
you stand to draw the same sort of fire as would the people that instituted the policies.
Mr Hilf is a perfect example. Those questions are answered, not as I would expect a techie to answer them
but in the style of a PR flack. Where possible he spouts corporate doctrine; where not he slickly and skillfully
evades the issue. As a result he draws fire, not as a techie, but as a flack.
Where do you draw your line between "serious" and "self made hack"?
Has a red carpet/debian package distribution?
There are loads of those that you can compile to any location you choose and (and this from personal experience) they'll still work.
Of course, you need to compile them yourself, but that's a limitation of the package management system rather than the software. One that can easily be worked around, to boot.
Bingo, a truly secure OS has built-in ACLs all over the place. Back in the 1980s even the most zealous UNIX advocates would openly admit that the unix security model was flawed, as most of those users were also familiar with highly secure mainframes back then.
In fairness, I don't recall anyone identifing a specific security weakness in the UNIX model. It was just (IIRC) that the committee that wrote the standards documents were familiar with ACLs so everyone had to have ACLs. And, in fairness again, once the standards were defined, UNIX, (in many of its incarnations at the time), set about adopting them.
I can count on the fingers of one hand the number of serious applications (as opposed to self-made perl hacks one drops in $HOME/bin) that can be installed in Unix without root privileges.
Surely not. Almost any gnu based app will allow a --prefix option at configuration time. Unless your definition of "serious" means "needs suid", but that's increasingly rare for well designed apps these days.
Where do you draw your line between "serious" and "self made hack"?
and then you copy back individual files as required. You can use tar and bzip2 for bundling and compression. crontab will automate the process.
Of course, arguably the need for checkpointing (as opposed to general backups) is a weakness in the
design of tghe windows registry in the first place.
No. This is only the case in the retarded unix security model where you are either a luser or root.
Umm... no. UNIX uses a three level security model by default. About ten
years ago, people realised that this wasn't going to cut it and so we had
Access Control Lists defined as part of the POSIX standard. ACLs,
as I understand them, allow
arbitary, user defined security models. They don't come as part of Linux as standard, but there are extensions. And some *NIX implementations have had them for
a decade or so.
An application ought to be able to declare itself to the operating system without having to be (a) superuser or (b) be able to step over the toes of other applications and/or users.
No argument there. Why do you assume that *NIX cannot do this? I was
installing apps on UNIX without root access and for personal use twenty years ago.
[it's a tree structure] but so is/etc if you think about it.
Only under the most rudimentary definition of tree. When a configuration file can be a perl script or a procmail script calling this a "tree" is stretching the term.
You misundstand. The directory structure is the tree, the files are keys and the file content is purely arbitary, just as it is in the windows registry.
Do you really find a commented perl script inferior to a key whose name is a long sequence of digits and curly brackets, and whose contents is in an encrypted binary format? I'd have to ask "for what purpose".
By the way, calling/etc the equivalent of the registry is actually doing a favour to *nix.
All I'm saying is that they're structurally isomorphic. This is a quibble over
implementation details.
I would implement the registry as a DB using XML syntax with a
sophisticated set of perms (for one a serious security system remembers
both the user who created the data and the app who posted it, as old
mainframes used to do. Mod perms should be derived from that.)
The trouble here is that it raises the complexity of any *NIX app. The design philosophy is to allow small code with minimal dependancies.
What you say would require every app to link against a database library and
an XML parser. We'd need the database running, or built into the kernel,
full time. You can mitigate this a bit with large mandatory frameworks, but that too has porting problems, to say nothing of the guys who want to run *nix on legacy
hardware, keep systems on bootable USB keys, to say nothing of embedded
developers.
And it still wouldn't be clear what the format of the configuration data was,
because vendors could, as they do with windows, embed binary data in
the database.
Meanwhile we can do the same thing with existing *nix setup, plus
dotfile overrides. And everyone knows where to look, and the files are all
in plaintext. That's not "retarded"; that's a design decision.
So, to summarise: we reliquish a chunk of flexibility and transparency, raise the bar for application developers, and in return we get to impose a design philosophy on the OS. I can't personally see that the trade off is worthwhile.
The registry is a good thing [hear me out, before you mod me down]
Will do.
having a central well-organized repository of your system setup is clearly good.
Why is that then? Your central repository implies admin privs, which either means problems for user installed packages, or else local repositories which defeats the objective. Well organised is arguably good, but too rigid a format and you have problems porting apps to the platform.
So, while I'm willing to keep an open mind here, I'm having problems with "clearly" in that sentence.
As bad as the registry is, the *nix solution is even worse: the mishmash of/etc files each with its own arcane syntax and command incantations.
Swings and roundabouts, really. You can comment a *nix config file - at their best they can be self documenting. That never happens with the windows registry. And as for arcane syntax there's nothing especially cult, canny or scrutable about the windows registry format. Yes, it's a tree structure with data stored at every node in the form of a list of value=data elements, but so is/etc if you think about it. Both formats leave considerable scope for obfuscation, but only the/etc text files allow comments.
So, in the absence of support for your point, I think I have to disagree here too.
Clearly
"You keep saying that word. I don' think it means what you think it means"
Clearly a modern operating system needs something like the windows registry but implemented in a way that is not so easy to subvert or make the system unstable.
Secure and stable are admirable goals, but you've yet to show how the/etc approach fails to achieve them, or what if any aspects of the windows registry make it a better choice.
As it stands, you're comparing a hypothetical, unwritten mechanism with one that exists and works (/etc) and using it to suggest the superiority of a third system (the windows registry) that you already conceeded to be flawed.
Will glamorizing science in the movies make kids pay better attention in chemistry class?
It did for me. And physics, and maths and computer science when I finally got to a place that would teach it.
OK, by that stage, I knew I was never going to find the magic formula to turn my brother into a frog, and that my army of killer robots had a few technical hurdles to overcome before implementation.
Slashdot Mirror
How about soixante-neuf?
Hang on... let me rephrase that...
Very true. The volume of patents submitted by MS every week would probably constitute a DOS attack from any who didn't own so many government officials.
Possibly not. If they'd submitted a patent on "I18N for numeric pattern matching" I doubt we'd be discussing this. As it is they have to try and drag in the display aspect so they can threaten anyone who makes a browser, help file viewer, or even the folks who maintain "less".
But even if they had numeric internationalized pattern matching I still wouldn't condone it.
And does source code not qualify? This is emacs we're talking about. The source would have been published.
Well then perhaps if you kept these points together, people wouldn't need to make assumptions nor to draw inferences concerning what you may have meant. As it is I did what you say you prefer and requested clarification, and you still get all snippy about it. No pleasing some people, I suppose.
I never agreed that it was actually true, i.e., "All that means is ... assuming your assertion were actually true."
So are you saying that the USPTO is doing a satisfactory job? I don't think even the USPTO think would go that far. They are overloaded, undermanned, lack quality technical minds and have problems retaining staff meaning that a disproportionate number of patents are being examined by inexperienced staff. No disrespect to the patent office, but I don't see how they can be doing a proper job under the circumstances.
And I never said nor implied that you said or implied any such thing. All I was doing was offing you the courtesy of persuing the question should you have felt that I was evading the issue. For someone who gets so uptight about unwarranted inferences, you certainly seem to draw your share.If anything, you're guilty of not being able to form sound arguments.
By what criteria and according to whom?
Splendid! Then you presumably conceed that small software houses are not enjoying any particular protection from the large houses under the current regime at the USPTO and that, as a minimum patent office procedures are in need of an overhaul. Perhaps you'll go so far as to grant that the benefits of software patents are not so "blindingly obvious" as you may have suggested given that the system is capable of such abuses, and that these benefits are more theoretical than actual at this point in time.
Now all we need is a robust method for discriminating between good and bad software patents, a means whereby the USPTO can deal with current flooding tactics and some sort of safeguard to prevent this from happening again and I might almost be ready to conceed the point. I'm sure the USPTO will find your proposals fascinating. I know I will.
Silly me. Software patents should be allowed because everything that is already patentable is patentable and therefore so should software be patentable. That's as if I were to say the government should give me a million dollars because everyone else who has already been given a million dollars by the government has been given a million dollars by the governement and therefore, I should be given a million dollars too.Is that what they call a circular argument do you suppose?
Just to be clear: My comment shouldn't be taken as accusing you of maliciously spreading FUD. I'm merely suggesting that you may have been less than critical the last time you were exposed to it. Also FUD is a woefully imprecise term to use in a debate as picky as this one. Let's say propaganda, since that's closer to the meaning I intended.Of course, with those caveats, if you still think I have a hidden agenda, I'd love to know what it might be and who you see as benefitting.
Well, clearly I've failed to compel you. On the other hand, your best argument so far in support of your case boiled down to "why not" - a tactic that not only lacks compulsion, but is somewhat juvenile to boot. Alas, "why not" frequently seems to be the final line of defence for the pro-swpats crowd. I do hope you have something better waiting in the wings - I was rather hoping you might offer something intelligent.Well, you can take the pragmatic viewpoint that patents tend to stifle innovation in software, and anti-competitive and demonstrably unnecessary for purposes of turning a profit. On the other hand I've yet to see a decent argument as to why software should be patentable.
So given that swopats have bad effects, and given that we are talking about what should be done - does it not make sense to outlaw them? I think I'd like a better argument than "it works for tractors" before I endorse the notion.
That turns out not to be the case. Software patents are being granted so broadly and freely that the big houses can easily find a patent that is arguably infringed by the small house. That done, they can use the threat of litigation to force the small house to licence the patent for a trivial sum, or refuse to licence the patent at all and bankrupt the smaller player. The protection only applies if you can trade patents one for one with the big players.
I would think this reason would be blindingly obvious.
It's FUD put out by the big software concerns. There's a reason why the small houses campaigned against swpats in the EU and the big corps lobbied for them. The reason is that they drastically favour the big boys.
That's the wrong question.
That remains to be seen.
I see nothing special about software to exlude it from receiving the same patent protections as anything else.
Which presupposes that everything else is indeed eligible for patenting. And yet this too turns out not to be the case. There are no patents on creative works; none on ideas; none on mathematical theorems or laws of physics.
So it seems that "everything else gets patented" doesn't work, and thus the question remains to be answered: Why should software enjoy the benefits of patents? I've made the case as to why it should not - now it's your turn.
That's just silly. You might as well say that copyrighting movies is only useful if you publish the script.
I think you'll find copyright extends to the binary compiled form of software. This is the form under which it is generally distributed commercially. That's why downloading unlicenced software is properly called copyright infringement, rather than "piracy" as a certain Mr. Gates would have us believe.
Speaking of Mr. Gates, you are aware that he became the richest man on the planet using copyright to protect his software? I'd say that argues quite strongly that copyright is adequate to the task of protecting and exploiting software. Would you not agree?
Clear now?
I do believe we're getting there, however slowly.
What you want to protect is what the software does and for that you need a patent.
Which rather presupposes that preventing others from writing competing packages is a good thing. I'm sure MS and Amazon, (to name but two) think so, but the benefits are rather less obvious for small software houses, consumers the free market and the world at large.
But if you're in the mood for injecting a little clarity into the debate, perhaps you could explain why society should grant this monopoly over software? What is it about software that you feel qualifies it for protection under patent? What does society get in return for granting this dispensation in the realm of software? You have yet to make that case.
Clear now?
My assertion that you can't get copyright and patent both is purely your own, incorrect, inference, Really, if you're going to try and pre-empt the rules of engagement you might at least practice what you preach.
I expect that's dreadfully convenient for certain large software houses who happen to be currently engaged spamming the USPTO with every bloody obvious idea they can think of as well. Especially considering how most of the "intellectual property" in the software business would seem to be vested in copyright,That by itself seems like an excellent reason for banning software patents. Or do you propose to tell me that patents are justified because they protect against patents?
Where is what written down? The protection of copyright as applied to software or lack of an inalienable right to double a monopoly on ideas?
If you're going to use a sig like that you're going to need to express yourself with much greater clarity.
Anyway, a UI is not copyrightable.
I said "software interface", not "interface". The software is copyrightable. if your concern is the colours and the fonts, might I suggest that trade marks might be the measure you seek?
Whether the software implementing the UI is copyrighted is irrelevant.
Irrelevant to what, precisely?
If you reply, do so only to what I explicitly wrote. If I didn't write it, don't assume or infer it.
If you post, express yourself with clarity and let your arguments stand on their own merits. Do not presume to instruct me as to how I may and may not frame my reply.
Sounds like a good case for a framework and a scripting language designed to be accountant-friendly. Geeks design the framework, gui, and report innterface, and you get an accountant, probably a retired one with an interest in IT to write the initial accountancy stuff and to consult on the script language design.
Do it right and you decouple the application design from the accountancy which means that the software userbase no longer depends upon the initial development team since they can hire accountants to write tax modules each year, or use ones released under a copyleft licence.
A bit of thought into modularity would help too. You could have a basic module for the more static, universal measures, and then have add-ons for people who need to take into account specific circumstances.
There'd even potentially be money in it, if you can get software support contracts for a few medium sized concerns. It sounds an interesting project.
Well, because your software interface is already protected by copright and you have no inalienable right to any further monopoly on the idea. In any event you don't need the patent to make money from your knobs so there's no need for it and you already have protection to exploit your idea, so that should be sufficent.
That is unless your requirements for sufficency extend to unfairly surpressing competition. Even so, it's far from "OK", IMHO.
Further, unless your hardware interface is staggeringly novel, then there is massive prior art on that too. Come back when you get the bugs out of a telepathic interface or something. And when you do, it'll be the telepathy chip that deserves the patent. The software that draws the pretty pictures on the screen will still not be "OK" for patenting.
By the way, there have been far dumber/worse patent snafus reported here. If they didn't expose said stupidity, what makes you think this rather benign instance will?
Have standards dropped so far at the USPTO that "medium-stupid or better" is all the qualifications an idea needs to be patentable? Just because we have the crimes of Jack the Ripper on record, that doesn't mean we should cease from complaining about cases of homicide; neither do past stupidities excuse this.
Oh, and of course it isn't fifteen pages long. but then this is print media and spending fifteen or so pages on needlessly obfscated license terms would cost the licensor money.
And that would never do.
Well, it would help if you were to say which usage of the word you had in mind. Although if you find atmospheric pressure to be in the same class as religion, we can probably assume definition A) The theory that physical matter is the only reality and that everything, including thought, feeling, mind, and will, can be explained in terms of matter and physical phenomena.
If so, would you be open to the idea that you are projecting your own viewpoint onto "educated people" as a whole? Materialism as defined above is a rather narrow and extremist worldview. It's great for winding up your classmates at school, or drunks at a party, but taken exclusively it's rather too limiting to form the basis for any useful ontology. In this respect, it has a lot in common with its philosophical mirror image solipsism in that it willfull invalidates fifty percent of human experience.
And, just as Samuel Johnson refuted Bishop Berkely, so it is almost trivially easy to refute materialism as a philosophy: it takes no account of information. I can communicate the number "four" to you in a number of ways: I can voice the word (in a variety of langauages); I can send a pattern of electrons over a computer network; I can hold up a number of fingers, I can inscribe a number of different patterns on paper or other substances. All these will send the ideal of "four" from my brain to yours, but the concept of "fourfoldness" is implicit in none of these things.
Then there is mathematics whereby it is possible to construct purely abstract systems with no reference or application to physical reality.
Going a degree more mystical, there is the question of preception: it's all very well to exlain my perceptions in terms of enzymes and electrons, the fact remains that there is an "I" to whom these perceptions are presented. There is even support for this in orthodox physics in the consciousness causes collapse interpretation of Quantum Mechanics.
I could go on, but that's enough to make the point.
Summing up, materialism can be useful but any supporting arguments tend to be circular since they have to disregard all non-materialist evidence. This would be acceptable if there was anything approaching the consensus you suggest, but that seems unlikely in the extreme: try telling your wife/GF/S.O. that "love" is nothing but an emergent effect arising from a combination of hormones and sensory stimuli. The chances are you'll get a quick grounding in why approx. 50% of the population has a problem with that proposition.
The point I'm getting to here is that Materialism is a religious belief. One of the defining characteristics of religious beliefs is that they take certain propositions on faith, both in the absence of evidence, and in the absence of any test that might falsify those propositions.
Interestingly, this makes atheism into a religion, since just as the theist professes a belief in a God in the absence of evidence, so does the atheist profess disbelief in the absence of any evidence. Agnosticism has the benefit of withholding judgement until evidence can be presented. However it is difficult to reconsice Meterialism with agnosticism since materialism tends to reject "God" as a valid concept for debate, and it's hard to define agnostic without reference to "God".
So I wonder which is most important to you: a rationalist world view, or your materialist religion? As you start to think of it in those terms, it becomes difficult to reconcile the two increasingly antagonistic viewpoints.
Mark Lawson seems to have been desperate to elicit some response along the lines of
I could almost start formulating consipiracy theories about laying groundwork for increased censorship, except that, the tenor of the questions is nothing unusual for a newsnight interview.Respect is due to Sir Tim for keeping his head and not rising to the bait.
Still, the political nature of the questions can be seen as reflecting the increasingly politicised nature of the web. I wonder if he's in for more of this sort of flack in the future.
Talk smack - that's fine - but don't lump *all* of us in to one big package and label it "evil."
It's not *all* of you that we're concerned about. It's the guys at the top. The ones who set policy. Just a few examples:
- We're concerned about the people who decide it would be good strategy to bankroll SCO.
-
We're concerned about the people who fund AdTI to produce such horribly
skewed and falacious reports. And we worry too about the
less glaringly incompetant bodies that they have working on similar projects.
-
We're concerned when Steve Ballmer threatens the whole of Asia with fines for violating "intellectual property" if
they adopt Linux
-
We're concerned by corporate pressure on OEMs to prevent them offering Linux preloaded at a competetive price and spec.
This is the sort of behaviour that prompts the use of the word "evil". We're have no axe to grind, or at least I personally don't, against MS developers. However you should acknowledge that your corporate agenda is set by people who are not friends of Free Software. Thus, if you post on slashdot defending those policies you stand to draw the same sort of fire as would the people that instituted the policies.Mr Hilf is a perfect example. Those questions are answered, not as I would expect a techie to answer them but in the style of a PR flack. Where possible he spouts corporate doctrine; where not he slickly and skillfully evades the issue. As a result he draws fire, not as a techie, but as a flack.
Of course, you need to compile them yourself, but that's a limitation of the package management system rather than the software. One that can easily be worked around, to boot.
In fairness, I don't recall anyone identifing a specific security weakness in the UNIX model. It was just (IIRC) that the committee that wrote the standards documents were familiar with ACLs so everyone had to have ACLs. And, in fairness again, once the standards were defined, UNIX, (in many of its incarnations at the time), set about adopting them.
I can count on the fingers of one hand the number of serious applications (as opposed to self-made perl hacks one drops in $HOME/bin) that can be installed in Unix without root privileges.
Surely not. Almost any gnu based app will allow a --prefix option at configuration time. Unless your definition of "serious" means "needs suid", but that's increasingly rare for well designed apps these days.
Where do you draw your line between "serious" and "self made hack"?
mmm...
and then you copy back individual files as required. You can use tar and bzip2 for bundling and compression. crontab will automate the process.Of course, arguably the need for checkpointing (as opposed to general backups) is a weakness in the design of tghe windows registry in the first place.
No argument there. Why do you assume that *NIX cannot do this? I was installing apps on UNIX without root access and for personal use twenty years ago.
You misundstand. The directory structure is the tree, the files are keys and the file content is purely arbitary, just as it is in the windows registry.Do you really find a commented perl script inferior to a key whose name is a long sequence of digits and curly brackets, and whose contents is in an encrypted binary format? I'd have to ask "for what purpose".
All I'm saying is that they're structurally isomorphic. This is a quibble over implementation details. The trouble here is that it raises the complexity of any *NIX app. The design philosophy is to allow small code with minimal dependancies. What you say would require every app to link against a database library and an XML parser. We'd need the database running, or built into the kernel, full time. You can mitigate this a bit with large mandatory frameworks, but that too has porting problems, to say nothing of the guys who want to run *nix on legacy hardware, keep systems on bootable USB keys, to say nothing of embedded developers.And it still wouldn't be clear what the format of the configuration data was, because vendors could, as they do with windows, embed binary data in the database.
Meanwhile we can do the same thing with existing *nix setup, plus dotfile overrides. And everyone knows where to look, and the files are all in plaintext. That's not "retarded"; that's a design decision.
So, to summarise: we reliquish a chunk of flexibility and transparency, raise the bar for application developers, and in return we get to impose a design philosophy on the OS. I can't personally see that the trade off is worthwhile.
Will do.
having a central well-organized repository of your system setup is clearly good.
Why is that then? Your central repository implies admin privs, which either means problems for user installed packages, or else local repositories which defeats the objective. Well organised is arguably good, but too rigid a format and you have problems porting apps to the platform.
So, while I'm willing to keep an open mind here, I'm having problems with "clearly" in that sentence.
As bad as the registry is, the *nix solution is even worse: the mishmash of /etc files each with its own arcane syntax and command incantations.
Swings and roundabouts, really. You can comment a *nix config file - at their best they can be self documenting. That never happens with the windows registry. And as for arcane syntax there's nothing especially cult, canny or scrutable about the windows registry format. Yes, it's a tree structure with data stored at every node in the form of a list of value=data elements, but so is /etc if you think about it. Both formats leave considerable scope for obfuscation, but only the /etc text files allow comments.
So, in the absence of support for your point, I think I have to disagree here too.
Clearly
"You keep saying that word. I don' think it means what you think it means" Clearly a modern operating system needs something like the windows registry but implemented in a way that is not so easy to subvert or make the system unstable.
Secure and stable are admirable goals, but you've yet to show how the /etc approach fails to achieve them, or what if any aspects of the windows registry make it a better choice.
As it stands, you're comparing a hypothetical, unwritten mechanism with one that exists and works (/etc) and using it to suggest the superiority of a third system (the windows registry) that you already conceeded to be flawed.
It did for me. And physics, and maths and computer science when I finally got to a place that would teach it.
OK, by that stage, I knew I was never going to find the magic formula to turn my brother into a frog, and that my army of killer robots had a few technical hurdles to overcome before implementation.
It didn't matter - by that time I was hooked.