I'd say the obvious solution is not to use a server to assign serial numbers, but instead to use, say, a hash of the operations/data involved or the submission timestamp coupled with the global static thread number or something.
The problem then is that transactions may arrive out of their globally defined order... you either have to back out subsequent ones and reapply in the correct order if this happens, or wait for some indeterminate period of time before applying a transaction to make sure one doesn't arrive after it that should be executed before it.
I'm sure there is a solution to this that doesn't present a scalability issue, but I don't really see it right now. My current project is likely to be hitting this kind of issue in ~12-18 months, so I keep thinking about it, but right now the serial number server seems to be the only practical way.
The editors have a loose definition of the work prove. I read the article and they provide some compelling arguments. However, I saw no proof in a mathematics or scientific way.
Perhaps you miss where they say they have a forthcoming paper?
NoSQL's two big features are scalability and the arbitrary schemas.
It amazes me how often the third crucial feature is missed: efficiency of implementation. Sending data to and from a database server by serializing to text and back to binary again is a seriously inefficient way of processing it, yet this is how it is typically done with SQL databases. A binary in-process API for manipulating the data without serializing it can have a profound impact on speed, which is a different thing entirely from scalability.
In essence, TFA claims that if the traditional ACID guarantee "if three transactions (let's call them A, B and C) are active... the resulting database state will be the same as if it had run them one-by-one. No promises are made, however, about which particular order execution it will be equivalent to: A-B-C, B-A-C, A-C-B" is not abandoned (as in NoSQL systems), but is even strengthened to a guarantee that the result will always be as if they arrived in A-B-C order, then it solves all kinds of possible replication problems, requires less networking between the many servers involved, and allows for high scaling while also keeping all the integrity constraints.
Which, to anyone who has seriously thought about how to implement atomic transactions in a nosql environment, should not exactly come as a shock. It's the obvious solution to the problem, and I'm sure if you dig into it you'll find hundreds of implementations that work just like that.
The interesting problem then becomes coming up with an efficient way of deciding what that complete ordering on transactions is going to be. You can only get so far with a single server that assigns serial numbers to the transactions on arrival (at least if your data is distributed and not replicated; in the latter situation you should be able to do so easily).
Angry at the revelation, Matthew Prichard, Christie's grandson, who describes the decision of Wikipedia as 'unfortunate,' says he will raise the matter with the play's producer, Sir Stephen Waley-Cohen.
I'm pissed off that the train I wanted to catch this morning was late. I shall be raising this matter with my housemate when he gets home from work later.
I've never seen the play nor plan to. So, I just read the Wikipedia article. The Wikipedia has a brief Synopsis of Act I & II and then a section titled 'Identity of the murderer'. My guess was correct after reading the breif synopsis. Really after the twist endings of The Usual Suspects, Palahnuik, and dare I say M. Night Shamwow, this is not a big shocker, and really it's very predicable.
Is there? I've spent a frustration 30 minutes bouncing from link to link looking for some actual details without much luck.
Yeah, the details are rather scant. But the decision was apparently made by a magistrate, which means he filed either in a local court or in the federal magistrates court. Either of these can be appealed (*almost* automatically) to a higher court, if he chooses to do so.
Last time I looked, there were several serious shortcomings with calc, in that it couldn't cope with (IIRC) more than 64K rows in a sheet, which can seriously hinder using it for statistical analysis.
I am surprised, though, at the article claiming that alcohol abstination is more common in lower socioeconomic classes. I was always under the impression that lower standard of living leads to higher prevalence of alcoholism and that higher income usually correlates with better higher health consciousness.
As I understand it, most abstainers do so because they have a close relative or friend who suffers from alcoholism, and this has put them off the idea. This is, of course, much more likely in deprived social groups, rather than affluent ones.
Was good for you. Then bad for you. Now has good cholestorol. It's the prime example of why "studies" are nothing but trash.
When exactly were eggs supposed to be bad for you? Checking studies back as far as 1982, I haven't found a single one that supports such a view, and plenty that suggest they are (at least) not harmful in terms of blood cholesterol levels.
I think the problem is amateur nutritionists who look at what they contain and panic. Professionals have known for about as long as such things can be measured that cholesterol consumption does not necessarily lead to blood cholesterol increases.
But the truth is it's been known for over a century that drinkers have cleaner arteries. Thinner blood, and/or some chemistry with the alcohol seems to help keep the plaques from forming.
Bingo. You know all those results for how aspirin prevents heart disease? Red wine, and probably some other alcoholic drinks as well, has exactly the same effect.
(Also worth noting: aspirin-allergics also often have negative reactions to red wine... not heard an explanation why, but it does seem to be true)
This gets better and better... They only timed the compute time. Cudamalloc was not part of the timing or cudamemcpy.
Sorry, I only count 'time to solution'. That is all i'm interested in.
Total time including initialisation and reporting result is only useful if what you're looking at is a complete process -- but when was the last time you just needed to sort something and that sorting wasn't part of a larger algorithm? It's at least not totally out of the question that the data will already be in GPU memory, and that the next step of processing can occur there too, at which point these (O(n)) overheads become irrelevant.
Except that to be able to use quantum crypto at all, you need to provide a physical way to pass the quantum state. And with that requirement, why won't you just pass the key the good old fashioned way? Strictly more secure, and much cheaper.
More secure? Hardly. All you have to do is eavesdrop on the key exchange and you have the key. In a real world scenario, typically this means bribing a few security guards, breaking into one of the communicators' homes or offices and retrieving the key from their computer, or intercepting a message sent over a physical line, probably encrypted via a non-100%-reliable cryptographic system, with the (at least) theoretical possibility that the encryption on the key exchange can be broken.
In a properly implemented quantum crypto system, this is theoretically impossible: the key passes directly from one endpoint to the other, and any interference between the two is easily detectable. It isn't stored for longer than the message takes to be sent, so breaking in to retrieve it is impractical. Done properly, the quantum crypto system is as secure as it is possible to be. As it happens, the system here was not done properly; it failed to detect interference on the line (and as ability to detect interference is, essentially, the point of quantum crypto, this is bad news).
Like many here you're not getting one thing - developers / geeks do not account for 90% of possible iPhone customers. There is something that is a problem for YOU and a problem for many OPEN-SOURCE type people - but not really something that is seen as a problem by the majority of people out there.
No, actually, this particular issue is one standing in the way of corporate adoption, not geek adoption. Corporate types don't want to have to hire Objective C developers - who are rare and expensive - to develop their iPhone apps. Apple, however, won't allow any other language to be implemented (other than Safari's javascript interpreter).
Geeks don't want rapid development tools, which is what the app in question is (a brief read of the web site makes it sound like a modernised implementation of the same sort of idea Lotus Notes started out with -- a quick way of defining your data structures and a very easy way to prototype UIs to manipulate them). We're generally happier getting down to the nuts and bolts and playing with the hardware at a lower level.
And it's not even limited to the iPhone - most people still use MS Office, despite how many competitors again? Despite the free OpenOffice?
MS Office *is* technically superior to OpenOffice. It has a large number of features that are not duplicated in OO. Its programmability is substantially superior to OO's. Also, it is the only piece of software that is able to read 100% of uncorrupted MS Office documents, and as that is the de-facto standard document exchange format in most fields of business, it's an important consideration.
The iPhone is not technically superior to Android, or even Symbian, both of which have important features that are missing on the iPhone. Flash is merely the most obvious one, but it is well worth noting that it is not only geeks who care about it. Many geeks, I suspect, are happier without it.
The closed Appstore may be something you hate - on the other hand, as far as non-geeks are concerned, I'd rather have the AppStore than seeing a proliferation of new phone threats (like - wouldn't you hate being spammed by a mobile botnet?).
I'm not sure in what way a mobile botnet is any worse than a static one. Sure, I see your point, but when it comes down to it you can make the same argument for your desktop computer... why run the risk of installing malware by mistake? Let's get a PC which will only install software that's been approved by Microsoft. And, admittedly, there are a lot of people that would be an adequate solution for, maybe even better than the current situation. But there are also a lot of people it isn't adequate for, and you can bet that if it was a situation that existed we'd be making a lot of noise about it. Why? So that people don't get one *by mistake* without realising how limited they'll be.
The same with the iPad - the iPad came out to much ridicule from the tech-savvy crowd - but see how many projects there are out to 'innovate' a tablet computer now that the iPad is out?
You seem to be missing something, which is the large number of people who were already working on tablet computers before Apple even announced the iPad. Several manufacturers released devices in a similar factor before the iPad was released (e.g. Asus's T91). Microsoft have been working with numerous manufacturers on similar devices since the late 90s. It's hardly a new idea.
For instance, Dune is another series that's a little ambiguous in this respect. By the time I get to "God-emperor of Dune", I see it morphing from SF to Fantasy (and boring Fantasy at that), but I'll still give it the benefit of the doubt and classify it SF (despite the fact that it IS a wonderful example of space opera, while still being entertaining and insightful as all hell). Hell, look at the whole Qwisatz Haderach nonsense and tell me how that's any less woo-woo than the Force crap. Both fairy-tale like if you ask me (but spanking good stories what?).
Are you seriously telling me you've watched Star Wars and read Dune and haven't realized that huge elements of Star Wars were ripped off from it?
Anyway, I'm going to disagree with your argument. There is now widely held to be a genre of "science fantasy" which is outside of both science fiction and fantasy; Star Wars and Dune both fit into this genre rather than either of the others.
It's only the readout from the quantum computer which is (or appears) random. The actual calculation is completely deterministic and doesn't exhibit randomness.
Currently implemented and proposed algorithms, yes. However, there is no reason a nondeterministic algorithm couldn't be implemented.
Good suggestion, but keep in mind that CC licenses are not designed to be used with software. As they say on their FAQ:
Can I use a Creative Commons license for software?
We do not recommend it. Creative Commons licenses should not be used for software. We strongly encourage you to use one of the very good software licenses which are already available. We recommend considering licenses made available by the Free Software Foundation or listed at the Open Source Initiative. Unlike our licenses, which do not make mention of source or object code, these existing licenses were designed specifically for use with software.
CC is a great set of licenses, but as they say, if you're dealing with software you're probably better off using one of the licenses designed with it in mind.
The reason their FAQ says this is that some of the clauses in other CC licenses break software distribution. For instance, the no-derivitives license would prevent distribution of binaries. The attribution requirements would likely have some strange effects with regards to how binaries can be distributed. CC share-alike (their equivalent of copyleft) requires only relicensing under the same terms, but doesn't require distribution of source code, so would be effectively useless. But CC0 has none of these problems. The only potential issue is that it explicitly doesn't include trademark or patent grants, so if you have either of these rights over your program you'll need to deal with those separately.
You want lawyers to not be by the book?! Read the license terms, it was legal.
Except that they (and consequently we) didn't have the rights to distribute it under the terms of that license. All distribution of glibc until now (and since the NFS-related portions were added) has been illegal. Sun/Oracle could have sued anyone distributing it if they chose to.
However, quantum effects would not help either, because a quantum computer can solve exactly the same set of problems a classical computer can.
Only under the assumption that the apparent randomness exhibited in quantum effects is *actually* random. If you instead hypothesize that it follows some pattern that we cannot determine or guess, and that the pattern may be influenced in ways that we do not (and cannot) understand by the physical universe, then quantum computers become able to perform computations that a classical system cannot emulate because it does not have access to this source of information.
Make no mistake about it : people who talk about unspeakable quantum phenomenon to explain thoughts are just people who are uncomfortable about the idea that we don't need any soul-thingie to explain sentience and consciousness.
Or are Roger Penrose, who seems to be more concerned about the implication that human intelligence can never be perfect (because if it can be simulated, it's [isomorphic to] a formal reasoning system, and is therefore subject to the incompleteness theorem).
Phillip Dick wrote Second Variety ten years before that third-rate knock-off. If anyone deserves credit for being ripped off, it's him.
Except that Berserker is a much closer match to the article's idea. Second Variety describes robots we created waging war on humanity (i.e. it prefigures Terminator). Berserker at least comes close to the theme of TFA: alien AIs that we make first contact with. And then they start trying to kill us. Much more relevant.
Of course, Berserker itself had earlier antecedants, and perhaps A for Andromeda is an even closer match to what the article is talking about, particularly as it discusses the result of a SETI-like program. I believe it may be in the sequel, Andromeda Breakthrough, that it is revealed the intelligence that originated the messages is an AI.
So, well done Shostak: you're only coming at this idea 50 years behind the SF writers.;)
Yes, manually programmable. I don't know about controllers that automatically program themselves.
I don't see anything in the patent that requires this to be automatic. The "mapping means" could easily be implemented as, for example, an array of switches to program device ids, and it appears to me this would satisfy the claims.
That's odd... I don't see the X10 protocol describing dynamic mapping of devices. Certainly not in the 1970s version.
Such mapping would be implemented within a programmable controller, not by the protocol itself which only describes the controller -> controlled device connection. Programmable controllers have been widely available for a long time.
Slashdot Mirror
If you can't tell the difference, please go back to drinking Budweiser.
And preferably not the stuff that comes from Budweis.
(What I find annoying: geographic indications are protected in wines, but nobody seems to bother when it comes to beers.)
I'd say the obvious solution is not to use a server to assign serial numbers, but instead to use, say, a hash of the operations/data involved or the submission timestamp coupled with the global static thread number or something.
The problem then is that transactions may arrive out of their globally defined order... you either have to back out subsequent ones and reapply in the correct order if this happens, or wait for some indeterminate period of time before applying a transaction to make sure one doesn't arrive after it that should be executed before it.
I'm sure there is a solution to this that doesn't present a scalability issue, but I don't really see it right now. My current project is likely to be hitting this kind of issue in ~12-18 months, so I keep thinking about it, but right now the serial number server seems to be the only practical way.
The editors have a loose definition of the work prove. I read the article and they provide some compelling arguments. However, I saw no proof in a mathematics or scientific way.
Perhaps you miss where they say they have a forthcoming paper?
NoSQL's two big features are scalability and the arbitrary schemas.
It amazes me how often the third crucial feature is missed: efficiency of implementation. Sending data to and from a database server by serializing to text and back to binary again is a seriously inefficient way of processing it, yet this is how it is typically done with SQL databases. A binary in-process API for manipulating the data without serializing it can have a profound impact on speed, which is a different thing entirely from scalability.
In essence, TFA claims that if the traditional ACID guarantee "if three transactions (let's call them A, B and C) are active ... the resulting database state will be the same as if it had run them one-by-one. No promises are made, however, about which particular order execution it will be equivalent to: A-B-C, B-A-C, A-C-B" is not abandoned (as in NoSQL systems), but is even strengthened to a guarantee that the result will always be as if they arrived in A-B-C order, then it solves all kinds of possible replication problems, requires less networking between the many servers involved, and allows for high scaling while also keeping all the integrity constraints.
Which, to anyone who has seriously thought about how to implement atomic transactions in a nosql environment, should not exactly come as a shock. It's the obvious solution to the problem, and I'm sure if you dig into it you'll find hundreds of implementations that work just like that.
The interesting problem then becomes coming up with an efficient way of deciding what that complete ordering on transactions is going to be. You can only get so far with a single server that assigns serial numbers to the transactions on arrival (at least if your data is distributed and not replicated; in the latter situation you should be able to do so easily).
I'm pissed off that the train I wanted to catch this morning was late. I shall be raising this matter with my housemate when he gets home from work later.
I've never seen the play nor plan to. So, I just read the Wikipedia article. The Wikipedia has a brief Synopsis of Act I & II and then a section titled 'Identity of the murderer'. My guess was correct after reading the breif synopsis. Really after the twist endings of The Usual Suspects, Palahnuik, and dare I say M. Night Shamwow, this is not a big shocker, and really it's very predicable.
Actually, I was kind-of hoping the butler did it.
Why don't they just edit it with "spoiler alert"
The relevant fact is just below a bold, large type heading with the text "Identity of the killer".
I don't really think much more than that is required, is it?
Is there? I've spent a frustration 30 minutes bouncing from link to link looking for some actual details without much luck.
Yeah, the details are rather scant. But the decision was apparently made by a magistrate, which means he filed either in a local court or in the federal magistrates court. Either of these can be appealed (*almost* automatically) to a higher court, if he chooses to do so.
Is MS Office really superior in anything that 90% of the users actually need from their word processor?
As of the latest version, you still can't have a word count in your status bar. This is pretty-much a killer failure for anyone who writes to word targets (i.e., most professional writers).
Last time I looked, there were several serious shortcomings with calc, in that it couldn't cope with (IIRC) more than 64K rows in a sheet, which can seriously hinder using it for statistical analysis.
I am surprised, though, at the article claiming that alcohol abstination is more common in lower socioeconomic classes. I was always under the impression that lower standard of living leads to higher prevalence of alcoholism and that higher income usually correlates with better higher health consciousness.
As I understand it, most abstainers do so because they have a close relative or friend who suffers from alcoholism, and this has put them off the idea. This is, of course, much more likely in deprived social groups, rather than affluent ones.
Eggs.
Was good for you. Then bad for you. Now has good cholestorol. It's the prime example of why "studies" are nothing but trash.
When exactly were eggs supposed to be bad for you? Checking studies back as far as 1982, I haven't found a single one that supports such a view, and plenty that suggest they are (at least) not harmful in terms of blood cholesterol levels.
I think the problem is amateur nutritionists who look at what they contain and panic. Professionals have known for about as long as such things can be measured that cholesterol consumption does not necessarily lead to blood cholesterol increases.
But the truth is it's been known for over a century that drinkers have cleaner arteries. Thinner blood, and/or some chemistry with the alcohol seems to help keep the plaques from forming.
Bingo. You know all those results for how aspirin prevents heart disease? Red wine, and probably some other alcoholic drinks as well, has exactly the same effect.
(Also worth noting: aspirin-allergics also often have negative reactions to red wine... not heard an explanation why, but it does seem to be true)
This gets better and better...
They only timed the compute time. Cudamalloc was not part of the timing or cudamemcpy.
Sorry, I only count 'time to solution'. That is all i'm interested in.
Total time including initialisation and reporting result is only useful if what you're looking at is a complete process -- but when was the last time you just needed to sort something and that sorting wasn't part of a larger algorithm? It's at least not totally out of the question that the data will already be in GPU memory, and that the next step of processing can occur there too, at which point these (O(n)) overheads become irrelevant.
Except that to be able to use quantum crypto at all, you need to provide a physical way to pass the quantum state. And with that requirement, why won't you just pass the key the good old fashioned way? Strictly more secure, and much cheaper.
More secure? Hardly. All you have to do is eavesdrop on the key exchange and you have the key. In a real world scenario, typically this means bribing a few security guards, breaking into one of the communicators' homes or offices and retrieving the key from their computer, or intercepting a message sent over a physical line, probably encrypted via a non-100%-reliable cryptographic system, with the (at least) theoretical possibility that the encryption on the key exchange can be broken.
In a properly implemented quantum crypto system, this is theoretically impossible: the key passes directly from one endpoint to the other, and any interference between the two is easily detectable. It isn't stored for longer than the message takes to be sent, so breaking in to retrieve it is impractical. Done properly, the quantum crypto system is as secure as it is possible to be. As it happens, the system here was not done properly; it failed to detect interference on the line (and as ability to detect interference is, essentially, the point of quantum crypto, this is bad news).
Like many here you're not getting one thing - developers / geeks do not account for 90% of possible iPhone customers. There is something that is a problem for YOU and a problem for many OPEN-SOURCE type people - but not really something that is seen as a problem by the majority of people out there.
No, actually, this particular issue is one standing in the way of corporate adoption, not geek adoption. Corporate types don't want to have to hire Objective C developers - who are rare and expensive - to develop their iPhone apps. Apple, however, won't allow any other language to be implemented (other than Safari's javascript interpreter).
Geeks don't want rapid development tools, which is what the app in question is (a brief read of the web site makes it sound like a modernised implementation of the same sort of idea Lotus Notes started out with -- a quick way of defining your data structures and a very easy way to prototype UIs to manipulate them). We're generally happier getting down to the nuts and bolts and playing with the hardware at a lower level.
And it's not even limited to the iPhone - most people still use MS Office, despite how many competitors again? Despite the free OpenOffice?
MS Office *is* technically superior to OpenOffice. It has a large number of features that are not duplicated in OO. Its programmability is substantially superior to OO's. Also, it is the only piece of software that is able to read 100% of uncorrupted MS Office documents, and as that is the de-facto standard document exchange format in most fields of business, it's an important consideration.
The iPhone is not technically superior to Android, or even Symbian, both of which have important features that are missing on the iPhone. Flash is merely the most obvious one, but it is well worth noting that it is not only geeks who care about it. Many geeks, I suspect, are happier without it.
The closed Appstore may be something you hate - on the other hand, as far as non-geeks are concerned, I'd rather have the AppStore than seeing a proliferation of new phone threats (like - wouldn't you hate being spammed by a mobile botnet?).
I'm not sure in what way a mobile botnet is any worse than a static one. Sure, I see your point, but when it comes down to it you can make the same argument for your desktop computer... why run the risk of installing malware by mistake? Let's get a PC which will only install software that's been approved by Microsoft. And, admittedly, there are a lot of people that would be an adequate solution for, maybe even better than the current situation. But there are also a lot of people it isn't adequate for, and you can bet that if it was a situation that existed we'd be making a lot of noise about it. Why? So that people don't get one *by mistake* without realising how limited they'll be.
The same with the iPad - the iPad came out to much ridicule from the tech-savvy crowd - but see how many projects there are out to 'innovate' a tablet computer now that the iPad is out?
You seem to be missing something, which is the large number of people who were already working on tablet computers before Apple even announced the iPad. Several manufacturers released devices in a similar factor before the iPad was released (e.g. Asus's T91). Microsoft have been working with numerous manufacturers on similar devices since the late 90s. It's hardly a new idea.
For instance, Dune is another series that's a little ambiguous in this respect. By the time I get to "God-emperor of Dune", I see it morphing from SF to Fantasy (and boring Fantasy at that), but I'll still give it the benefit of the doubt and classify it SF (despite the fact that it IS a wonderful example of space opera, while still being entertaining and insightful as all hell). Hell, look at the whole Qwisatz Haderach nonsense and tell me how that's any less woo-woo than the Force crap. Both fairy-tale like if you ask me (but spanking good stories what?).
Are you seriously telling me you've watched Star Wars and read Dune and haven't realized that huge elements of Star Wars were ripped off from it?
Anyway, I'm going to disagree with your argument. There is now widely held to be a genre of "science fantasy" which is outside of both science fiction and fantasy; Star Wars and Dune both fit into this genre rather than either of the others.
It's only the readout from the quantum computer which is (or appears) random. The actual calculation is completely deterministic and doesn't exhibit randomness.
Currently implemented and proposed algorithms, yes. However, there is no reason a nondeterministic algorithm couldn't be implemented.
The reason their FAQ says this is that some of the clauses in other CC licenses break software distribution. For instance, the no-derivitives license would prevent distribution of binaries. The attribution requirements would likely have some strange effects with regards to how binaries can be distributed. CC share-alike (their equivalent of copyleft) requires only relicensing under the same terms, but doesn't require distribution of source code, so would be effectively useless. But CC0 has none of these problems. The only potential issue is that it explicitly doesn't include trademark or patent grants, so if you have either of these rights over your program you'll need to deal with those separately.
You want lawyers to not be by the book?! Read the license terms, it was legal.
Except that they (and consequently we) didn't have the rights to distribute it under the terms of that license. All distribution of glibc until now (and since the NFS-related portions were added) has been illegal. Sun/Oracle could have sued anyone distributing it if they chose to.
However, quantum effects would not help either, because a quantum computer can solve exactly the same set of problems a classical computer can.
Only under the assumption that the apparent randomness exhibited in quantum effects is *actually* random. If you instead hypothesize that it follows some pattern that we cannot determine or guess, and that the pattern may be influenced in ways that we do not (and cannot) understand by the physical universe, then quantum computers become able to perform computations that a classical system cannot emulate because it does not have access to this source of information.
Make no mistake about it : people who talk about unspeakable quantum phenomenon to explain thoughts are just people who are uncomfortable about the idea that we don't need any soul-thingie to explain sentience and consciousness.
Or are Roger Penrose, who seems to be more concerned about the implication that human intelligence can never be perfect (because if it can be simulated, it's [isomorphic to] a formal reasoning system, and is therefore subject to the incompleteness theorem).
Phillip Dick wrote Second Variety ten years before that third-rate knock-off. If anyone deserves credit for being ripped off, it's him.
Except that Berserker is a much closer match to the article's idea. Second Variety describes robots we created waging war on humanity (i.e. it prefigures Terminator). Berserker at least comes close to the theme of TFA: alien AIs that we make first contact with. And then they start trying to kill us. Much more relevant.
Of course, Berserker itself had earlier antecedants, and perhaps A for Andromeda is an even closer match to what the article is talking about, particularly as it discusses the result of a SETI-like program. I believe it may be in the sequel, Andromeda Breakthrough, that it is revealed the intelligence that originated the messages is an AI.
So, well done Shostak: you're only coming at this idea 50 years behind the SF writers. ;)
Yes, manually programmable. I don't know about controllers that automatically program themselves.
I don't see anything in the patent that requires this to be automatic. The "mapping means" could easily be implemented as, for example, an array of switches to program device ids, and it appears to me this would satisfy the claims.
That's odd... I don't see the X10 protocol describing dynamic mapping of devices. Certainly not in the 1970s version.
Such mapping would be implemented within a programmable controller, not by the protocol itself which only describes the controller -> controlled device connection. Programmable controllers have been widely available for a long time.