Google of course was not the first web search thingie. Having been around for a while, I am of the impression, though, that it was the first web search to actually provide relevant results systematically. And *that* was innovating.
In the same vein, maps have been around for ever, but yet it is quite undeniable that their maps have a certain shinyness of its own (possibly coming from the coolness factor associated with underlying technology, at least if you understand it) that sets it quite apart from other mapping services. Innovation may also mean come up with a really good way to do something that already exists.
On the other hand, one thing that impressed me quite muchly was the fact that they actually were able to innovate with their email service: gmail proposed a different way to dealing with one's email, and that had not happened since... well, probably since email existed.
I am really using the word verifiability as a sinomym for resistance against attempts to falsability; this abuse of language (which it is) is standard (no one says that 'the discovery of Neptune did not falsify Newtonian mechanics'...) and needs no defense. I probably was not clear about this, I guess. Effectively verifying universal propositions is, if the idea makes any sense at all, not a requirement for science. I cling not to any rotten corpses.
Regarding your two steps...
The 'observation' you refer to is really a petition of principle. For one thing, it really depends on some fixing of what the rules of natural selection are, but of course these are dynamic, as is any other hypothesis in the realm of science; natural selection as a hypothesis has changed in time, and has changed in a way which allows now to explain things which could not be explained with older versions of the hypothesis--- this is nothing to be surprised at, as science works exactly this way. This does not mean that an hypothesis cannot be justifiably rejected, and there are abundant candidates which would prove the hypothesis of natural selection false. 'Excesive' diversity does not seem one of this candidates: diversity would have to be made quantifiable objectively (as being excesive for my understanding is certainly not a problem...), a rather precise argument should be made in order to establish what the 'reasonable' amount of diversity that would ensue from natural selection is, and justification for the assertion that the actual diversity present in nature exceeds the reasonable amount of diversity consequent to natural selection.
As for your second step: The assertion "an intelligent designer is the best available explanation" is an assertion about available explanations, not about the universe or the world or anything else. Its content is dependent of making precise what intelligence is (which is a very difficult problem even when one tries to distinguish non-intelligent 'trivial' actions from intelligent ones, let alone when one is trying to describe 'creations of the world' as intelligent actions; does a non-intelligent designer make sense?), what designing means in this context and what is it that is being designed (which is not really obvious: did the proposed designer designe the world as we see it piece by piece, or did he/she/it design rules which when put to work turned an initial state, picked by the designer, into the state of the world we see now---the two alternatives are not equivalent: for an amusing toy example, consider the creation of the game of life by Conway), what the set of available explanations is, and to fix a sense in which an explanation is better than another. For example, why is the absence of a rule a worse explanation for anything than the existence of an intelligent background actor? Can the two hypothesis be distinguished in any way? Before all this is done, your step 2 is nothign mre than a petition of principle.
Any intelligent discussion regarding the subject at hand requires a minimal understanding of the subject. Now, the subject is not really biology but epistemology and the philosophy of science: the objections to ID are not based on the content of the assertions it proposes, but on their standing with regards to the methods of science.
That a statement such as "The complex combinations of chemicals needed for blood coagulation are too complex to have evolved on their own" (and any other statement for that matter) needs to be verifiable to acquire the lustre of scientificity is tautological. This is independent of its being a statement about biology: it is a condition of scientificity. I for one have absolutely no idea of what the biology of coagulation is, yet it seems obvious to me that such an assertion requires to be verifiable.
Now, I have to confess that I do not see how a verification of that particular statement might look like. I'd say that a definition of what complexity is and a way to measure it, or at least, to make comparaisons of complexity, would be required; this can be very hard, as it shpould be an objective definition, not based on what we can understand or not, and would probably need a rather delicate mathematical formulation to be of any use. Also, I'd expect that a very precise description of what one means by "on its own" would be needed, and, accompanying that, a description of what "not on its own" would mean, too. After those preliminaries, one would have simply gotten to a point in which the statement to be verified has a precise meaning. As it currently stands, I simply cannot make much sense of it, except as a declaration of what we do not understand. Next, one would need to devise experiments (even thought experiments) which would allow one to verify it or to reject it---it would not be needed that the experiments could be effectively carried out (this is what happens with the experiments needed to validate string theories: we cannot at the current time perform them, but physicists are all the time thinking about them and coming up with new ones and so on): one might device processes to determine an empirical upper bounds on the complexity of what can happen "on their own", for example, and to compare that to the complexity of coagulation (which one would have to measure somehow), or something. Once you do that, the statement would be have adquired the quality of being falseable, and it would become scientific.
Without doing all this 'homework', it is impossible to take the statement seriously.
When one speaks of 'testing' or 'predicting' with respect to a scientific theory one is not referring to actually predicting a result, before it happens. Of course, that is not excluded, and it is a most spectacular event when that happens (the prediction of the existence of Neptune is a remarkable example of this---you can read about it here---but there are of course many others, among which one should also name the fact that the theory of selection and heredity suggested the plausibility of mechanisms like those that we now know DNA is involved in in the transmition of hereditary characteristics)
That a theory 'predicts' something in this context means that the something happens can be derived from other statements in the theory but that the theory itself, in its formulation, does not contain the statement that the something happens. For example, Newtonian mechanics predict the time needed by a stone to fall from my table to the floor. It is irrelevant with respect to the usual criteria for scientificity of theories that I may or may not already know how long that will take, the only important thing is that I can somehow derive from more 'basic' statements in the theory the time.
This may seem like a bad criterion, but it is not. It is essentially the best thing we've come up with in millenia of thinking about this.
Also, it is interesting that, for example, no theory that we know of predicts gravitation (except maybe some forms of string theory, I think...) despite the fact that we are all well aware of it. We have theories that include gravitation in their formulation: for example, the Newtonian theory of gravitation essentially axiomatically says that there is a force between massive bodies of such and such magnitude. This is a rather economical way of introducing gravitation, but it is nonetheless a deux-ex-machina introduction.
You can, though, agree that none of those systems of assertions (ID, Allah's desire for jihad, any religous statement) are of scientific nature by the very fact that they are not testable. This is really tautological, and not an arguable point, since this is indeed just using the 'standard' definition of science.
On the other hand, an argument can be made by which anything which is not testable, that is, anything which has no observable consequence, can safely be said to not exist: any true statement that can be made regarding an universe which contains some thing which is not observable is equivalent to the conjunction of a true statement regarding an alternate universe in which this non-observable thing does not exist and (possibly) the statement that this non-observable thing exists.
Therefore a consistent theory of the universe-with-this-non-observable-thing provides an essentially equivalent consistent theory of the universe-in-which-everything-which-exists-is-obser vable.
Reciprocally, one can adjoint to any consistent theory of the universe the existence of an arbitrary number of non-observable things as axioms without disturbing consistency, and without enlarging the class of true statements regarding the universe except by those statements stating that the adjoined things exist.
In this precise sense, anything which is not observable can safely be assumed no to exist.
From your second paragraph, I guess you would prefer a situation in which christianity would not have been divided, and you attribute (in part, I guess...) the current division to the fact that Luther proposed that anyone interpret the bible on his own. Does the (pre-Luther for some, current for the Roman Catholic) situation in which the interpretation of the bible is mandated by an authority and in which, presumably in consequence, the Christian Church is united, seem preferable to you?
Actually, all theories of gravitation before the general theory of relativity were discarded because they made bad predictions, or because they could not withstand contrast with observation, which is essentially the same thing. For example,
the introduction of epicycles and similar devices was motivated by the failure of previous theories to predict astronomical events in a way concordant with the measurements done with the precision available at the time. Similarly, Newtonian celestial mechanics failed to predict the observed peculiarities in the movement of Mercury, and, more spectacularly, could not accomodate the observed behaviour of light rays in the presence of large masses.
(Of course, I am using the word 'discarded' here with the usual qualifications involving domains of validity and so on, which underlie the justification for using the Newtonian theory, which in principle is passé, when computing the trajectory of a stone falling from my table)
It is true that if you are allowed to use arbitrarily many epicycles you can aproximate arbitrarily closely any periodic motion (and obtain it exactly if you are allowed to use infinitely many): this is a restatement of the fact that trigonometric polinomials can arbitrarily aproximate (in mostly any reasonable sense chosen for this word) a continuous periodic function. But this introduces parameters (infinitely many if you are going for perfect matches of trajectories) that need to be determined empirically, since such a theory of planetary motion will be not much more than a best fit theory in the sense of least squares. Now, these parameters are not independent (this is a consequence of the fact---let's keep the discussion Newtonian, for simplicity---that we can also provide a much more economic explanation for the movement, depending only on a considerably smaller number of parameters: the masses of the bodies in the system, their relative distances at any particular moment, the constant of gravitation), and the Ptolemaic theory does not describe the dependencies, so in a very precise sense the Ptolemaic description is not complete.
Only when you complete it with the whole set of relations among the parameters involved in the description of movements (this is possible: it amounts to a translation in terms of Fourier coefficients of Newton's equations of motion) you obtain a theory which you can describe as equivalent to Newton's, and only then are you in a position to decide that you prefer the latter because of its economy of description. Before this 'completion', Ptolemy's theory is purely descriptive.
The world is not locally flat. It only appears so if you do not look hard enough at the evidence.
Indeed, it is essentially the content of Gauss's Theorema Egregium (this name, bestowed on the theorem by Gauss himself, reflects the fact that this is a quite remarkable result) that you can tell a flat surface from a rounded one by performing local experiments only. For example, measuring the sum of angles in geodesic triangles.
Well, you can always not compile your apps under Cygwin.
See, they are as free to do what they want with the fruits of their labor as you are free to do what you want with the fruits of your labor. Now: they apparently have made their minds to let you use the fruits of their labor, with the only provision that they in turn get to use the fruits of yours. You can, similarily, make up your mind and not want to allow them to use the fruits of your labor. There is absolutely no problem with either of you making up your minds thusly. Only, you'll have to keep in mind that you will not get to use the fruits of their labor: because your decition implies the non-acceptance of their conditions.
There is an essential difference between software and food: the marginal cost of production of a unit. This consideration is rather basic to the philosophy underlying the GPL, and you just cannot overlook it.
Now, the situation with music and movies is both more interesting and more difficult, since for one thing the `usual' considerations about marginal cost are insufficient or, at least, require that subtleties specific to music and movies be taken into account. Invoking a famous precedent, I'll leave it to you to analyse it. One reasonable point of departure is the consideration of what exactly it is that is the product that the music and movie industry distributes (which, note, has changed over time).
You are forced to comply when you accept: true; but that is true of every licence, not only the GPL. Try accepting and not complying to any one of Microsoft's EULA (and telling them you do).
You are forced to comply when you accept any kind of legally binding agreement: that is what `legally binding' means.
Also, the way you `get there' is important. Most legislations consider void any kind of legally binding agreement which is forced upon one of the `agreeing' parties, so under those legislations, you simply cannot legally force a party to comply to an `agreement' unto which that party was forcefully subjected. (Note that this is essentially taulological).
I think it is fair to give enough credit to the people that wrote the GPL as to believe they know this elementaly facts, don't you think?
Not really: what happens is that the time the developers of those apps put into them is very valuable, much more than mine or yours.
The source of value for things like software is essentially the labor put into making them, specially in the case of those particular apps in which the role of capital in the production process was (at least initially) close to nil (except in the case of Mozilla, OpenOffice, and a few others)
Maybe what shocked him/her is the fact that using revision control on a text stored in the essentially opaque Word format is very very similar to simply storing backups for every single version in a way that essentially prevents useful comparaisons (without a tool which understands the format, that is). Doing the mind experiment of visualizing what you could do with good ol' plain diff on good ol' plain text/TeX/etc (which is not by far the best way you can think of to deal with changesets on textual data) might give you a hint of why he/she thinks it's the stupidest thing ever.
it's just x^x (which is weird if you try to work it using x^n or a^x formulae, and you don't know logarithmic differentiation)
It's not weird---it's just wrong.
"Reverse engineering" the derivative of a function might have been fun, but it is of rather doubtful utility.
In the particular case of x^x, the difficulty most often resides in the fact that most people do not really even know what function that particular notation denotes.
Slashdot Mirror
The US is not oversensitive to `their' space?
Google of course was not the first web search thingie. Having been around for a while, I am of the impression, though, that it was the first web search to actually provide relevant results systematically. And *that* was innovating. In the same vein, maps have been around for ever, but yet it is quite undeniable that their maps have a certain shinyness of its own (possibly coming from the coolness factor associated with underlying technology, at least if you understand it) that sets it quite apart from other mapping services. Innovation may also mean come up with a really good way to do something that already exists.
On the other hand, one thing that impressed me quite muchly was the fact that they actually were able to innovate with their email service: gmail proposed a different way to dealing with one's email, and that had not happened since... well, probably since email existed.
Cute.
Does any liability survive the EULA?
I am really using the word verifiability as a sinomym for resistance against attempts to falsability; this abuse of language (which it is) is standard (no one says that 'the discovery of Neptune did not falsify Newtonian mechanics'...) and needs no defense. I probably was not clear about this, I guess. Effectively verifying universal propositions is, if the idea makes any sense at all, not a requirement for science. I cling not to any rotten corpses.
Regarding your two steps...
The 'observation' you refer to is really a petition of principle. For one thing, it really depends on some fixing of what the rules of natural selection are, but of course these are dynamic, as is any other hypothesis in the realm of science; natural selection as a hypothesis has changed in time, and has changed in a way which allows now to explain things which could not be explained with older versions of the hypothesis--- this is nothing to be surprised at, as science works exactly this way. This does not mean that an hypothesis cannot be justifiably rejected, and there are abundant candidates which would prove the hypothesis of natural selection false. 'Excesive' diversity does not seem one of this candidates: diversity would have to be made quantifiable objectively (as being excesive for my understanding is certainly not a problem...), a rather precise argument should be made in order to establish what the 'reasonable' amount of diversity that would ensue from natural selection is, and justification for the assertion that the actual diversity present in nature exceeds the reasonable amount of diversity consequent to natural selection.
As for your second step: The assertion "an intelligent designer is the best available explanation" is an assertion about available explanations, not about the universe or the world or anything else. Its content is dependent of making precise what intelligence is (which is a very difficult problem even when one tries to distinguish non-intelligent 'trivial' actions from intelligent ones, let alone when one is trying to describe 'creations of the world' as intelligent actions; does a non-intelligent designer make sense?), what designing means in this context and what is it that is being designed (which is not really obvious: did the proposed designer designe the world as we see it piece by piece, or did he/she/it design rules which when put to work turned an initial state, picked by the designer, into the state of the world we see now---the two alternatives are not equivalent: for an amusing toy example, consider the creation of the game of life by Conway), what the set of available explanations is, and to fix a sense in which an explanation is better than another. For example, why is the absence of a rule a worse explanation for anything than the existence of an intelligent background actor? Can the two hypothesis be distinguished in any way? Before all this is done, your step 2 is nothign mre than a petition of principle.
Can you list make a list of 3 of those?
Any intelligent discussion regarding the subject at hand requires a minimal understanding of the subject. Now, the subject is not really biology but epistemology and the philosophy of science: the objections to ID are not based on the content of the assertions it proposes, but on their standing with regards to the methods of science.
That a statement such as "The complex combinations of chemicals needed for blood coagulation are too complex to have evolved on their own" (and any other statement for that matter) needs to be verifiable to acquire the lustre of scientificity is tautological. This is independent of its being a statement about biology: it is a condition of scientificity. I for one have absolutely no idea of what the biology of coagulation is, yet it seems obvious to me that such an assertion requires to be verifiable.
Now, I have to confess that I do not see how a verification of that particular statement might look like. I'd say that a definition of what complexity is and a way to measure it, or at least, to make comparaisons of complexity, would be required; this can be very hard, as it shpould be an objective definition, not based on what we can understand or not, and would probably need a rather delicate mathematical formulation to be of any use. Also, I'd expect that a very precise description of what one means by "on its own" would be needed, and, accompanying that, a description of what "not on its own" would mean, too. After those preliminaries, one would have simply gotten to a point in which the statement to be verified has a precise meaning. As it currently stands, I simply cannot make much sense of it, except as a declaration of what we do not understand. Next, one would need to devise experiments (even thought experiments) which would allow one to verify it or to reject it---it would not be needed that the experiments could be effectively carried out (this is what happens with the experiments needed to validate string theories: we cannot at the current time perform them, but physicists are all the time thinking about them and coming up with new ones and so on): one might device processes to determine an empirical upper bounds on the complexity of what can happen "on their own", for example, and to compare that to the complexity of coagulation (which one would have to measure somehow), or something. Once you do that, the statement would be have adquired the quality of being falseable, and it would become scientific.
Without doing all this 'homework', it is impossible to take the statement seriously.
When one speaks of 'testing' or 'predicting' with respect to a scientific theory one is not referring to actually predicting a result, before it happens. Of course, that is not excluded, and it is a most spectacular event when that happens (the prediction of the existence of Neptune is a remarkable example of this---you can read about it here---but there are of course many others, among which one should also name the fact that the theory of selection and heredity suggested the plausibility of mechanisms like those that we now know DNA is involved in in the transmition of hereditary characteristics)
That a theory 'predicts' something in this context means that the something happens can be derived from other statements in the theory but that the theory itself, in its formulation, does not contain the statement that the something happens. For example, Newtonian mechanics predict the time needed by a stone to fall from my table to the floor. It is irrelevant with respect to the usual criteria for scientificity of theories that I may or may not already know how long that will take, the only important thing is that I can somehow derive from more 'basic' statements in the theory the time.
This may seem like a bad criterion, but it is not. It is essentially the best thing we've come up with in millenia of thinking about this.
Also, it is interesting that, for example, no theory that we know of predicts gravitation (except maybe some forms of string theory, I think...) despite the fact that we are all well aware of it. We have theories that include gravitation in their formulation: for example, the Newtonian theory of gravitation essentially axiomatically says that there is a force between massive bodies of such and such magnitude. This is a rather economical way of introducing gravitation, but it is nonetheless a deux-ex-machina introduction.
You can, though, agree that none of those systems of assertions (ID, Allah's desire for jihad, any religous statement) are of scientific nature by the very fact that they are not testable. This is really tautological, and not an arguable point, since this is indeed just using the 'standard' definition of science.
On the other hand, an argument can be made by which anything which is not testable, that is, anything which has no observable consequence, can safely be said to not exist: any true statement that can be made regarding an universe which contains some thing which is not observable is equivalent to the conjunction of a true statement regarding an alternate universe in which this non-observable thing does not exist and (possibly) the statement that this non-observable thing exists.
Therefore a consistent theory of the universe-with-this-non-observable-thing provides an essentially equivalent consistent theory of the universe-in-which-everything-which-exists-is-obser vable.
Reciprocally, one can adjoint to any consistent theory of the universe the existence of an arbitrary number of non-observable things as axioms without disturbing consistency, and without enlarging the class of true statements regarding the universe except by those statements stating that the adjoined things exist.
In this precise sense, anything which is not observable can safely be assumed no to exist.
A honest question:
From your second paragraph, I guess you would prefer a situation in which christianity would not have been divided, and you attribute (in part, I guess...) the current division to the fact that Luther proposed that anyone interpret the bible on his own. Does the (pre-Luther for some, current for the Roman Catholic) situation in which the interpretation of the bible is mandated by an authority and in which, presumably in consequence, the Christian Church is united, seem preferable to you?
Actually, all theories of gravitation before the general theory of relativity were discarded because they made bad predictions, or because they could not withstand contrast with observation, which is essentially the same thing. For example, the introduction of epicycles and similar devices was motivated by the failure of previous theories to predict astronomical events in a way concordant with the measurements done with the precision available at the time. Similarly, Newtonian celestial mechanics failed to predict the observed peculiarities in the movement of Mercury, and, more spectacularly, could not accomodate the observed behaviour of light rays in the presence of large masses.
(Of course, I am using the word 'discarded' here with the usual qualifications involving domains of validity and so on, which underlie the justification for using the Newtonian theory, which in principle is passé, when computing the trajectory of a stone falling from my table)
It is true that if you are allowed to use arbitrarily many epicycles you can aproximate arbitrarily closely any periodic motion (and obtain it exactly if you are allowed to use infinitely many): this is a restatement of the fact that trigonometric polinomials can arbitrarily aproximate (in mostly any reasonable sense chosen for this word) a continuous periodic function. But this introduces parameters (infinitely many if you are going for perfect matches of trajectories) that need to be determined empirically, since such a theory of planetary motion will be not much more than a best fit theory in the sense of least squares. Now, these parameters are not independent (this is a consequence of the fact---let's keep the discussion Newtonian, for simplicity---that we can also provide a much more economic explanation for the movement, depending only on a considerably smaller number of parameters: the masses of the bodies in the system, their relative distances at any particular moment, the constant of gravitation), and the Ptolemaic theory does not describe the dependencies, so in a very precise sense the Ptolemaic description is not complete.
Only when you complete it with the whole set of relations among the parameters involved in the description of movements (this is possible: it amounts to a translation in terms of Fourier coefficients of Newton's equations of motion) you obtain a theory which you can describe as equivalent to Newton's, and only then are you in a position to decide that you prefer the latter because of its economy of description. Before this 'completion', Ptolemy's theory is purely descriptive.
One should note that these episodes seem to occur with a geographical distribution that would not pass any standard statistical test for uniformity...
The world is not locally flat. It only appears so if you do not look hard enough at the evidence.
Indeed, it is essentially the content of Gauss's Theorema Egregium (this name, bestowed on the theorem by Gauss himself, reflects the fact that this is a quite remarkable result) that you can tell a flat surface from a rounded one by performing local experiments only. For example, measuring the sum of angles in geodesic triangles.
Well, you can always not compile your apps under Cygwin.
See, they are as free to do what they want with the fruits of their labor as you are free to do what you want with the fruits of your labor. Now: they apparently have made their minds to let you use the fruits of their labor, with the only provision that they in turn get to use the fruits of yours. You can, similarily, make up your mind and not want to allow them to use the fruits of your labor. There is absolutely no problem with either of you making up your minds thusly. Only, you'll have to keep in mind that you will not get to use the fruits of their labor: because your decition implies the non-acceptance of their conditions.
It's not that hard.
Hm, there are `FOSS analogues' for all kind of content, not just software, and there are licences designed for exactly that. Google around a bit...
I'd love to know, though, whether there exist licences in common use for such content containing the `viral' provisions.
There is an essential difference between software and food: the marginal cost of production of a unit. This consideration is rather basic to the philosophy underlying the GPL, and you just cannot overlook it.
Now, the situation with music and movies is both more interesting and more difficult, since for one thing the `usual' considerations about marginal cost are insufficient or, at least, require that subtleties specific to music and movies be taken into account. Invoking a famous precedent, I'll leave it to you to analyse it. One reasonable point of departure is the consideration of what exactly it is that is the product that the music and movie industry distributes (which, note, has changed over time).
You are forced to comply when you accept: true; but that is true of every licence, not only the GPL. Try accepting and not complying to any one of Microsoft's EULA (and telling them you do).
You are forced to comply when you accept any kind of legally binding agreement: that is what `legally binding' means.
Also, the way you `get there' is important. Most legislations consider void any kind of legally binding agreement which is forced upon one of the `agreeing' parties, so under those legislations, you simply cannot legally force a party to comply to an `agreement' unto which that party was forcefully subjected. (Note that this is essentially taulological).
I think it is fair to give enough credit to the people that wrote the GPL as to believe they know this elementaly facts, don't you think?
Not really: what happens is that the time the developers of those apps put into them is very valuable, much more than mine or yours.
The source of value for things like software is essentially the labor put into making them, specially in the case of those particular apps in which the role of capital in the production process was (at least initially) close to nil (except in the case of Mozilla, OpenOffice, and a few others)
It forces no one to do anything. No one is forced to accept a licence, be it the GPL of Microsoft's EULA du jour.
Heh. That made me laugh. "Self-avowed communist"! That sounds like taken from a bad tv special on the McCarthy era.
What else would it be mostly driven by?
Maybe what shocked him/her is the fact that using revision control on a text stored in the essentially opaque Word format is very very similar to simply storing backups for every single version in a way that essentially prevents useful comparaisons (without a tool which understands the format, that is). Doing the mind experiment of visualizing what you could do with good ol' plain diff on good ol' plain text/TeX/etc (which is not by far the best way you can think of to deal with changesets on textual data) might give you a hint of why he/she thinks it's the stupidest thing ever.
Note that the whole year 2000 was dedicated to Math...
Google for geomview.
It's not weird---it's just wrong.
"Reverse engineering" the derivative of a function might have been fun, but it is of rather doubtful utility.
In the particular case of x^x, the difficulty most often resides in the fact that most people do not really even know what function that particular notation denotes.