Until they over-use them and get into a jam. Then they call on us techies to help them make their contorted mess act like a database.
This seems to be something that resolver one is aimed directly at, since a big selling point of its representing everything in Python is enabling IT staff to easily extract business-user-created spreadsheet logic and transfer it to other systems.
That is true, but the user cannot add columns quickly. I'm thinking of having a reserved row for the column name. If you don't put a name there, then it is automatically named the spreadsheet column letter, or the like (and column letters can perhaps still be aliases even if you do have an explicit name).
That's not really a feature of the DBMS, but of the UI, though. Given the generality of the programmability that resolver one appears to offer, it seems like it should be nearly trivial to implement a spreadsheet in it that would use a appropriate SQL backend this way.
Well, but spreadsheets have a grid feel instead of a table feel and follow grid rules instead of table rules.
That's generally true, though good spreadsheets make it possible to implement worksheets, or ranges of worksheets, that obey table-like rules to a greater or lesser extent, and the particular features highlighted in resolver one suggest that it does so far more than most out of the box (in ways that allow columns and rows to act like table columns and rows, and even allow worksheets to work like views backed by some set of other worksheets.) At least, that's what I infer from the description of worksheet formulae on the resolver one home page:
Worksheet formulae
Resolver features formulae for worksheets as well as for cells; this allows a single formula to fill a worksheet with values calculated from one or more other worksheets. Result worksheets calculated in this way will be updated correctly as the input data grows or shrinks; this method also diminishes the incidence of single cells with incorrect formulae, since there is no need to mechanically specify the formula for every cell in the worksheet.
Sorry if I'm ignorant, but I don't think you can put functions in cells and use these to manipulate data in other cells.
Lat I checked, you expressly could not in Excel with VBA custom worksheet functions. Even if you could, I'd certainly rather do it in Python than VBA, though.
A good portion of spreadsheets actually should be database tables of some kind.
A good portion of spreadsheets are actually backed by database tables of some kind.
People end up manually grouping and other stuff that report-writers can do automatically. What is needed is a kind of "dynamic" RDBMS tool that has open-ended columns and column widths. A "spreadbase"? The Oracle clones are all too rigid.
While I think you are selling Oracle and its object-relational kin short if you think they can't handle what you seem to be describing, a more simply "flexible" approach is that taken by, e.g., SQLite where types are advisory rather than rigid. But in either case the DBMS is just the back-end storage engine, you still need a front-end piece that the business user can interact with in a friendly, visual way or program if necessary, that's where something spreadsheet-like comes in. Resolver one seems, at the outline level, to be a good way of approaching that.
SQL databases have become much lighter and more efficient these days. Why should I use this store data over a lightweight SQL database?
One of the selling points is using it to interact with SQL databases that are used to store data. From the resolver one homepage:
Entire database tables, or the results of arbitrary SQL queries, can be imported directly into Resolver documents as new worksheets, which can immediately be used in calculations. If necessary, Resolver can update the worksheet in real time as the underlying dataset changes. Because there is no need for the user to write code themselves, this allows for sophisticated data analysis without assistance from IT.
because the automation system controlling the infrastructure is not connected to a public network, like say, the internet - right ?
Just because its not connected to a "public" network doesn't mean that an attacker couldn't get access to the network at a vulnerable location. Sure, physical security on the infrastructure of a private network is an important part of the solution, but if there is no reason to provide a remote self-destruct function, then its a good idea to make sure that even if the network is compromised, it can't be used to self-destruct the generator. This is true whether or not it is accessible via a public network.
The one time pad, where the key length = message length is still safe as long as you never reuse the key.
The one time pad is always going to be secure, but it is of limited utility, since it requires a method of securely communicating at least as much information as you are trying to protect. There are certainly things that one-time pads are very good for, but there are lots of applications of encryption for which a one-time pad is never going to be a practical solution,
Otherwise you have to do that same math with every possible key, which means that every bit that is added to the length of the key doubles the key space, and drastically increases the number of computations a computer would need to try to brute force the key.
Well, assuming that you need to do a brute force attack, that's true even for quantum computers (though the "drastic" increase is far less than for traditional computers, where brute force is O(N) in size of the key space, since for quantum computers Grover's Algorithm, assuming Wikipedia is correct, is O(N^1/2).)
OTOH, the speedups quantum computing offers against at leastmany common forms of public key encryption are even more devastating, and I don't know of any form of public key encryption that seems to be safe (I'm not an expert, I'd welcome hearing of one that isn't clearly at extreme risk!). Even if quantum computer just makes public-key encryption worthless, that's a pretty serious blow.
They'll just increase the key size to the point where it won't be easy for even a quantum computer to decrypt...Since there is no theoretical limit to the size of the key, and the only practical limit is processing power, this is almost trivial.
If encryption doesn't scale better than decryption, then there is a problem, since then (at best) someone with K times your processing power (for some value of K that is independent of key size) will be able to decrypt your transmission as easily as you encrypt it, no matter how many bits you use for the key.
As pointed out elsewhere, the people who would participate are too self-selecting. They would just be too small a segment of the U.S. population.
And that would be different from the present US system how? You think substantive participation in the US political system isn't something done by a very small, largely self-selected (though wealth plays a big role in the capacity to "self-select") cadre of the population now?
People who do have alot of time on their hands and are interested to spend that time on reading and voting on bills would be able to put more weight to their views, which aren't necessary mine. So I'd need a representative, who is able to spend alot of time representing my views. Hence a party. Too bad this isn't always the way it works in current governments.
Well, one reason it may not work that way in real governments is that, even with a representative, you have the same effect going on: the people that have the time and interest in influencing legislation spend a lot of time, money, and effort presenting arguments and "facts" (sometimes actual facts, sometimes not) supporting their position to legislators, the people who lack time or interest or both don't, and so the former group is overrepresented in exactly the same way they would be in a direct democracy.
(I can't really understand how the US ended up with only two parties; overhere we have at least five, and in that handfull of parties I don't find a complete match with my views.)
In the US we have pretty much always had two major national political parties (though which two has changed over time), and each state one or two major local parties (in rare cases, not the same two major parties as the national parties -- e.g., before the Minnesota Farmer-Labor Party merged with the Democratic Party after the New Deal, it had displaced the locally-weak Democratic Party to become the second major party in Minnesota, and actually had then displaced the Republican Party as the dominant party in that state) because the structure of our electoral systems at almost every level encourage two parties; winner-take-all elections, particularly plurality elections, do that, and all of our elections for federal office are single-winner, and most are plurality (the Presidential election is a bit weird, but with, IIRC, two exceptions each state election for Presidential electors is a winner-take-all plurality election for multiple electors.)
A system where more elections are conducted in a more proportional manner will tend have more parties (and, generally, better popular satisfaction with government).
This is not a wiki for creating laws. It is a wiki for public input on laws that are being drafted by legislators.
It would just invite vandalism, and instead of leading to formation of new laws, it would waste money and manpower involved in maintanence and moderation.
There's a lot of manpower and money that goes through separating wheat from chaff in public comments submitted in input on legislation no matter what the form, that's unavoidable. The only way to avoid that is to have a legislature that ignores the public.
Very cool, a law wiki. Now what law do I want to create?... I know, how about making it illegal to walk forwards on a Thursday. Anyone breaking this law, will be locked in stocks and flogged to death, with custard coated kippers.
From the wiki's homepage:
On this site, wiki technology is paired with ongoing work to develop a Policing Act in 2008.
An official Bill is currently being written-up by parliamentary drafters, but in parallel there's an opportunity for others to suggest how a new Policing Act might look by contributing to a wiki Act. It'll be kept open until 1 November 2007, when the results can be fed back into the official law-making process.
Creating this online environment is a continuation of the open process used throughout the Police Act Review. It's all about encouraging a national conversation on policing.
The new Act will need to cover a wide range of topics, from high-level governance to day-to-day administration. To help get people started, we've included some headings and a few example ideas.
But don't feel constrained. For instance, if you'd prefer to work offline and upload a complete Act for others to comment on, by all means add it beneath the one we've started (there's a space provided under the "Alternative versions" heading).
Already, there's been heaps of interest in the wiki, with lots of people adding things, editing parts they feel can be improved, or offering alternatives. To help ensure positive ideas aren't lost, and provide a space for people to explain their suggestions, we've added discussion pages linking off the main pages.
Good luck!
So, no, I don't think you've found a flaw with the concept of this law wiki. I think you've found a flaw with a concept that you invented yourself when you saw the Slashdot summary, which isn't at all the same thing. This isn't a law carved in stone by wiki editors; the wiki is just adopted as a way to solicit public input that allows a number of different forms of input. There may well be flaws with that concept, but not the one you pointed to.
The design for both was continually downsized until the only purpose of both the shuttle and the station is to exist for each other, like some crazy love story.
Um, the shuttle was operational long before the ISS was planned. In fact, it was operational before the far more ambitious NASA "Space Station Freedom" was planned. So, it simply doesn't make sense to say the design of both the shuttle and station were downsized until their only purposes were to exist for eachother: there weren't any concrete plans for a station to downsize until the shuttle was not only completely designed, but actually flying.
If the money supply is stable, you can only increase gdp if you have deflation (at least equally bad) or if people start using the same money more. Since thats unrealistic, you need to increase the money supply.
Actually, increases in the velocity of money (particularly the velocity of money within the domestic economy) are not unrealistic. They aren't things you can achieve through monetary policy, of course, though other government policies can affect them; for instance, if you allow checks to be processed as EBTs, you reduce delays and uncertainty in processing transactions, and thereby increase the velocity of money. If you can alter tax policy to selectively reduce the share of taxes paid by a class of consumers that has, on average, a higher propensity to spend in the domestic economy, you increase the (domestic) velocity of money.
What you try to do is increase it so that it exactly equals the change in GDP, but thats pretty impossible to do.
As I understand it, central bankers don't, in practice, aim to exactly match GDP growth, instead they usually try to slightly overshoot for two reasons: first, if its not excessive, slightly easy money itself is generally seen as being stimulative and accelerating GDP growth; second, inadequate money supply is seen as substantially worse than oversupply of similar magnitude even outside of the range in which the stimulative effect of easy money is outweighed by the inflationary effect, so its better to miss high.
That is the most comical statement of all. Sure, all they do is "manage" the economy.
No, they manage the money supply. Now, that is important to the economy, and particular important in regards to inflation, but its far from the only lever that is used to manage the economy, and it is a fairly blunt instrument compared to, say, fiscal policy when it comes to managing the economy.
Inflation is, quite simply, the printing of additional money
No, its not. Inflation is an increase in general price levels. It may be (or may not be, depending on other factors in the economy) an effect of creating additional money (either by printing additional notes or easing credit terms), though if the money supply is expanding at exactly the pace at which the output is expanding, you wouldn't expect it to have that effect. It may also be a result of causes other than expansion of the money supply. Shortage of a key, broadly important economic input (energy/fuel inputs are the best examples) resulting in declining production can produce inflation, as can anything else that reduces production, given a constant money supply, as can other persistent supply reductions that may accompany war or domestic instability (and by "war" I mean one where attacks are occurring on the country being studied, not a "foreign war", although I suppose a foreign war that a country participates in could limit production by reducing the supply of labor and thereby produce inflation as the same money is chasing fewer goods, though it would be rather unusual for a modern foreign war to involve enough of the population to have a significant effect through that mechanism.)
My knowledge of copyright judgments in the US is lacking, but doesn't the plaintiff usually need to register the work with the copyright office in order to get punitive damages?
The requirement is actually broader than just punitive damages; in general, the only action for violation of copyright that can occur without registration is an action by the original author of a visual work (whether or not the copyright holder) to address false claims or imputations of authorship or distortion of the work that would harm the reputation of the author or destruction of a work of "recognized stature". See 17 U.S.C. Section 411(a), 17 U.S.C. Section 106A(a).
But note that registration is prerequisite for other legal actions, it does not have to occur before the alleged violation on which those legal actions are based, just before a suit is filed.
Er, no. That would be 1/2, not 1/3. (And you mean "denominator".)
When adding 1/3 + 1/3 + 1/3, the computer needs to do 4 integer adds.
It needs to do two integer adds, and two compares. If the denominators weren't equal, it would also need to go through several more operations to find a common multiple of the denominators and adjust the numerators correspondingly; OTOH, this doesn't really invalidate your point, since its not a lot of operations, and they are all integer operations.
Do not compare floating point numbers for equality. Doing so is almost always a bug.
Why do spreadsheets still use floating point, anyway? Accuracy seems more important than speed, in general, in spreadsheets (sure, you can do complex realtime simulations in a spreadsheet if you want to, but that's not the main point of them), so defaulting to using something a unlimited precision decimal or rational format as the default number storage format would seem to make a lot more sense than floating point.
My point was that all these projects can be counted as GPLv3 projects, or is it that important that I formally fork such a project to be counted in the numbers?
Well, yes, if it is not currently licensed with the restrictions in the GPLv3 (but merely allows other people to relicense their own redistribution that way), it is inaccurate to describe it as a GPLv3 project. I mean, by your argument, every project under a GPLv3-compatible license (or, presumably, in the public domain like SQLite) should be counted as a GPLv3 project because someone could conceivably redistribute a derivative of it under the GPLv3 at some point in the future.
Slashdot Mirror
Jan 1 -> July 1 is 6 months.
Jan 1 -> June 30 is, truncated to an integer number of months, 5 months.
This seems to be something that resolver one is aimed directly at, since a big selling point of its representing everything in Python is enabling IT staff to easily extract business-user-created spreadsheet logic and transfer it to other systems.
That's not really a feature of the DBMS, but of the UI, though. Given the generality of the programmability that resolver one appears to offer, it seems like it should be nearly trivial to implement a spreadsheet in it that would use a appropriate SQL backend this way.
That's generally true, though good spreadsheets make it possible to implement worksheets, or ranges of worksheets, that obey table-like rules to a greater or lesser extent, and the particular features highlighted in resolver one suggest that it does so far more than most out of the box (in ways that allow columns and rows to act like table columns and rows, and even allow worksheets to work like views backed by some set of other worksheets.) At least, that's what I infer from the description of worksheet formulae on the resolver one home page:
Lat I checked, you expressly could not in Excel with VBA custom worksheet functions. Even if you could, I'd certainly rather do it in Python than VBA, though.
A good portion of spreadsheets are actually backed by database tables of some kind.
While I think you are selling Oracle and its object-relational kin short if you think they can't handle what you seem to be describing, a more simply "flexible" approach is that taken by, e.g., SQLite where types are advisory rather than rigid. But in either case the DBMS is just the back-end storage engine, you still need a front-end piece that the business user can interact with in a friendly, visual way or program if necessary, that's where something spreadsheet-like comes in. Resolver one seems, at the outline level, to be a good way of approaching that.
One of the selling points is using it to interact with SQL databases that are used to store data. From the resolver one homepage:
Entire database tables, or the results of arbitrary SQL queries, can be imported directly into Resolver documents as new worksheets, which can immediately be used in calculations. If necessary, Resolver can update the worksheet in real time as the underlying dataset changes. Because there is no need for the user to write code themselves, this allows for sophisticated data analysis without assistance from IT.
Good point; the distinction is important.
Just because its not connected to a "public" network doesn't mean that an attacker couldn't get access to the network at a vulnerable location. Sure, physical security on the infrastructure of a private network is an important part of the solution, but if there is no reason to provide a remote self-destruct function, then its a good idea to make sure that even if the network is compromised, it can't be used to self-destruct the generator. This is true whether or not it is accessible via a public network.
The one time pad is always going to be secure, but it is of limited utility, since it requires a method of securely communicating at least as much information as you are trying to protect. There are certainly things that one-time pads are very good for, but there are lots of applications of encryption for which a one-time pad is never going to be a practical solution,
Well, assuming that you need to do a brute force attack, that's true even for quantum computers (though the "drastic" increase is far less than for traditional computers, where brute force is O(N) in size of the key space, since for quantum computers Grover's Algorithm, assuming Wikipedia is correct, is O(N^1/2).)
OTOH, the speedups quantum computing offers against at leastmany common forms of public key encryption are even more devastating, and I don't know of any form of public key encryption that seems to be safe (I'm not an expert, I'd welcome hearing of one that isn't clearly at extreme risk!). Even if quantum computer just makes public-key encryption worthless, that's a pretty serious blow.
If encryption doesn't scale better than decryption, then there is a problem, since then (at best) someone with K times your processing power (for some value of K that is independent of key size) will be able to decrypt your transmission as easily as you encrypt it, no matter how many bits you use for the key.
And that would be different from the present US system how? You think substantive participation in the US political system isn't something done by a very small, largely self-selected (though wealth plays a big role in the capacity to "self-select") cadre of the population now?
Well, one reason it may not work that way in real governments is that, even with a representative, you have the same effect going on: the people that have the time and interest in influencing legislation spend a lot of time, money, and effort presenting arguments and "facts" (sometimes actual facts, sometimes not) supporting their position to legislators, the people who lack time or interest or both don't, and so the former group is overrepresented in exactly the same way they would be in a direct democracy.
In the US we have pretty much always had two major national political parties (though which two has changed over time), and each state one or two major local parties (in rare cases, not the same two major parties as the national parties -- e.g., before the Minnesota Farmer-Labor Party merged with the Democratic Party after the New Deal, it had displaced the locally-weak Democratic Party to become the second major party in Minnesota, and actually had then displaced the Republican Party as the dominant party in that state) because the structure of our electoral systems at almost every level encourage two parties; winner-take-all elections, particularly plurality elections, do that, and all of our elections for federal office are single-winner, and most are plurality (the Presidential election is a bit weird, but with, IIRC, two exceptions each state election for Presidential electors is a winner-take-all plurality election for multiple electors.)
A system where more elections are conducted in a more proportional manner will tend have more parties (and, generally, better popular satisfaction with government).
This is not a wiki for creating laws. It is a wiki for public input on laws that are being drafted by legislators.
There's a lot of manpower and money that goes through separating wheat from chaff in public comments submitted in input on legislation no matter what the form, that's unavoidable. The only way to avoid that is to have a legislature that ignores the public.
From the wiki's homepage:
So, no, I don't think you've found a flaw with the concept of this law wiki. I think you've found a flaw with a concept that you invented yourself when you saw the Slashdot summary, which isn't at all the same thing. This isn't a law carved in stone by wiki editors; the wiki is just adopted as a way to solicit public input that allows a number of different forms of input. There may well be flaws with that concept, but not the one you pointed to.
Um, the shuttle was operational long before the ISS was planned. In fact, it was operational before the far more ambitious NASA "Space Station Freedom" was planned. So, it simply doesn't make sense to say the design of both the shuttle and station were downsized until their only purposes were to exist for eachother: there weren't any concrete plans for a station to downsize until the shuttle was not only completely designed, but actually flying.
Couldn't you just compile one of the existing open-source computer algebra systems to an existing Linux-based phone or handheld platform?
Actually, increases in the velocity of money (particularly the velocity of money within the domestic economy) are not unrealistic. They aren't things you can achieve through monetary policy, of course, though other government policies can affect them; for instance, if you allow checks to be processed as EBTs, you reduce delays and uncertainty in processing transactions, and thereby increase the velocity of money. If you can alter tax policy to selectively reduce the share of taxes paid by a class of consumers that has, on average, a higher propensity to spend in the domestic economy, you increase the (domestic) velocity of money.
As I understand it, central bankers don't, in practice, aim to exactly match GDP growth, instead they usually try to slightly overshoot for two reasons: first, if its not excessive, slightly easy money itself is generally seen as being stimulative and accelerating GDP growth; second, inadequate money supply is seen as substantially worse than oversupply of similar magnitude even outside of the range in which the stimulative effect of easy money is outweighed by the inflationary effect, so its better to miss high.
No, they manage the money supply. Now, that is important to the economy, and particular important in regards to inflation, but its far from the only lever that is used to manage the economy, and it is a fairly blunt instrument compared to, say, fiscal policy when it comes to managing the economy.
No, its not. Inflation is an increase in general price levels. It may be (or may not be, depending on other factors in the economy) an effect of creating additional money (either by printing additional notes or easing credit terms), though if the money supply is expanding at exactly the pace at which the output is expanding, you wouldn't expect it to have that effect. It may also be a result of causes other than expansion of the money supply. Shortage of a key, broadly important economic input (energy/fuel inputs are the best examples) resulting in declining production can produce inflation, as can anything else that reduces production, given a constant money supply, as can other persistent supply reductions that may accompany war or domestic instability (and by "war" I mean one where attacks are occurring on the country being studied, not a "foreign war", although I suppose a foreign war that a country participates in could limit production by reducing the supply of labor and thereby produce inflation as the same money is chasing fewer goods, though it would be rather unusual for a modern foreign war to involve enough of the population to have a significant effect through that mechanism.)
The requirement is actually broader than just punitive damages; in general, the only action for violation of copyright that can occur without registration is an action by the original author of a visual work (whether or not the copyright holder) to address false claims or imputations of authorship or distortion of the work that would harm the reputation of the author or destruction of a work of "recognized stature". See 17 U.S.C. Section 411(a), 17 U.S.C. Section 106A(a).
But note that registration is prerequisite for other legal actions, it does not have to occur before the alleged violation on which those legal actions are based, just before a suit is filed.
Er, no. That would be 1/2, not 1/3. (And you mean "denominator".)
It needs to do two integer adds, and two compares. If the denominators weren't equal, it would also need to go through several more operations to find a common multiple of the denominators and adjust the numerators correspondingly; OTOH, this doesn't really invalidate your point, since its not a lot of operations, and they are all integer operations.
SmartDraw is a very nice package, IMO; its not Free or even free, though, and the request was specifically for OSS.
Why do spreadsheets still use floating point, anyway? Accuracy seems more important than speed, in general, in spreadsheets (sure, you can do complex realtime simulations in a spreadsheet if you want to, but that's not the main point of them), so defaulting to using something a unlimited precision decimal or rational format as the default number storage format would seem to make a lot more sense than floating point.
Well, yes, if it is not currently licensed with the restrictions in the GPLv3 (but merely allows other people to relicense their own redistribution that way), it is inaccurate to describe it as a GPLv3 project. I mean, by your argument, every project under a GPLv3-compatible license (or, presumably, in the public domain like SQLite) should be counted as a GPLv3 project because someone could conceivably redistribute a derivative of it under the GPLv3 at some point in the future.