Homogeneous grape/blackberry/mint stuff = jelly. This stuff is of uniform consistency, often overly sweetened, and usually clear as opposed to cloudy.
Smashed up strawberry/raspberry/blackberry/etc stuff = jam. This stuff has seeds, skins, etc in it.
Yes, blackberry comes both ways. Strawberry is usually jam when sold in grocery stores, but if you eat out you'll sometimes get little packets of strawberry jelly. Grape is invariably jelly.
If it's got citrus peel in it we call it marmalade instead of jam, and it's not nearly as common as the other kinds.
For readers in the US, Panadol is apparentl a foreign brand name for what's commonly sold here as Tylenol, and paracetamol is a foreign generic name for what's called acetaminophen here.
The high-definition itself does help a lot following small, fast-moving objects, so as far as seeing the action it's more of a help in hockey or tennis than basketball or football.
Independently of how "helpful" it is, though, the hi-def is nice for the same reason it's nice in movies: it just looks better, and gives you more detail. And widescreen lets you see more of the product.
I mean, it's not like you can't follow what's going on in a VHS copy of Lawrence of Arabia. But god does it look better on hi-def widescreen (or 70mm projection if you get a chance).
Similarly, seeing all the fog players blow out on a cold day, the condition of the grass, dirt stains on the uniforms after a slide, etc don't help you follow the action any better but they do make it "feel" more immediate and make it more fun to watch.
Seeing a great film cut to pan-and-scan is just silly, you miss 30% or so of the visuals. Somewhat similarly, widescreen lets you see more. You can follow positioning and motion away from the ball, and really get a feel for team strategy instead of just following the action of whoever has the ball at the moment..
I'm no fan of Bush, but even if he were a good president he'd be scared for his life. About 10% of US Presidents have been assassinated in office, and another 20% or so have had serious attempts on their lives. The ones who were killed -- Garfield, McKinley, Lincoln, Kennedy -- aren't among our worst presidents, and Lincoln is probably one of our best.
That seperates them from the average person on the street.
Apart from restricted pointers, all of these (or work-alikes) are in C++.
[SNIP next three paragraphs saying that none of them are actually in C++ but other features can do the same thing]
Obviously being a Turing-complete language C++ can do anything C can. The point is that it is not a superset of C, and you can't just use a C++ compiler to compile your C code.
Use a C++ compiler for C++. Use a C compiler for C. And always remember that C++ is not an "extended C", nor is C a subset of C++. They are seperate (albeit related) languages).
Well, all I can say is that C++ is (so much) more than just a "stricter" C.
While I agree with your core point, I have to take exception to the implication that C++ is at all a stricter C (even if it's also more). C++ and C are different languages, and C is not a subset of C++. There are valid C programs that are invalid in C++ (even not using things like variables named "new", etc), and features like implicit void casting that C++ lacks. There are programs that are valid C and valid C++ but behave differently.
And that's without getting into features of modern C (variable size arrays, language built-in complex numbers, restricted pointers, etc) that are not in C++ as far as I know.
But as far as your main point, yes, the reason to use C++ is if you want/need C++ features. My original objection was to the suggestion that you just "write C but use a C++ compiler to add namespaces and nothing else". Many of the drawbacks of C++ compared to C are pretty minor, and may be worth the tradeoff if you're going to take advantage of a lot of language features. Writing "C in C++" is just silly, though.
The most common example has to do with void pointers:
int *a; a = malloc(sizeof(int)*10);
vs. a = (int *) malloc(sizeof(int)*10);
C++ requires the latter, which does nothing but make the code more brittle in the face of change. void * is supposed to convert implicitly to other types, that is its entire reason for its existence. Code gets more fragile when you have to sprinkle references to an object's type all over the place instead of having some minimal (bare-bones, craptastic) polymorphism.
The core problem is that both the C and C++ type systems are jokes. But C++ got stricter without fixes the system itself, so you wind up having to put additional cruft in your code to no positive end.
It doesn't necessarily have to be standardized completely, but there are some language bases that any reasonable CS curriculum will cover:
* At least a semester of a lower-level EE course working with basic architecture (logic gates, VHDL or similar) * Assembly coding in at least one assembly language * At least one lower-level statically typed language (C, Fortran, C++, etc) * At least one strictly, statically typed language (ML, Haskell, etc) * At least one dynamically-typed language (scheme, smalltalk, lisp, python, etc) * At least one static class-oriented language (C++, Java, etc) * At least one more dispatch-oriented OO language (smalltalk, CLOS, python, etc)
[My personal preference would be to cover these with some RISC assembly, C, ML, lisp, python, and java; lisp especially gives you a lot of other features that are worth discussing.]
There are other language niches that are arguably important to cover, arguably stack-based languages (Forth, Postscript), non-class-oriented OO languages (self, JavaScript), etc. But at least the ones above deserve to be addressed in any respectable CS course.
Patterns aren't core CS, they're just a jargon that Java programmers use to describe common program designs. Some of them aren't often applicable to all languages or have different, older names used by non Java-heads that predate the Gang of Five's choices. They're certainly worth covering if you're going to do a lot of class-oriented programming but think of them as more of a common lexicon than anything deeper.
For the same reason that Feynman diagrams won a Nobel prize. There's often very high value in explaining something in a way that can be easily understood.
1. You'll wind up fighting the compiler figuring out how to turn off exceptions, rtti, etc 2. You don't gain much over lexical scoping. 3. C is not a subset of C++ (even forgetting trivialities like variables named "new", etc). You won't be able to re-use existing C code without strictly auditing it because of subtle difference (e.g. on 32-bit Intel, sizeof('a') is 4 in C but it's 1 in C++). Those differences are often of the "it compiles fine but behaves differently" type, which really sucks. 4. You get saddled with C++'s crappy stricter typing. When your type system is crap, making it stricter hurts rather than helps.
The standard used to be notify the vendor and wait forever. Vendors never fixed anything. Bugtraq and other security lists implemented a "full and immediate disclosure" policy, and bugs started getting fixed.
Lately, full and immediate has been pushed back to full and one month. That's a compromise so that in the unlikely event that hackers aren't already exploiting the bug it can be fixed by responsible companies before they get a chance to do so.
The drawbacks of the delay are:
1. The consumer is not informed about their risks; I might decide not to run that app at home, or not to do online banking on the same machine where I run the vulnerable app. I might be able to firewall off or otherwise protect myself using faculties outside the affected app.
2. The irresponsible vendors now just don't start working on the bug until a month later, so fixes get delayed. Moreover, when immediate disclosure was the norm there was no yelling at the person who disclosed the bug as though they had done something wrong; now, companies are trying to make it seem as though the people who discover the holes are somehow the ones in the wrong instead of the vendors/authors who wrote the insecure code in the first place.
Really, the one month delay ought to be the norm only with companies/individual who show that they actually use the month to work on a fix and often manage to get one out. For those who ignore things until they're disclosed, we should go back to immediate disclosure.
And absolutely as a consumer I want to know if there's a problem so that I can take steps to protect myself. The month is an okay compromise but if you're leaving known holes open and unpatched, and not letting me know so I can decide how to deal with that risk, it's hurting me. And it's hurting other consumers.
And ultimately the goal of 3rd-party security researchers should be to protect the consumers and users, not to cover the company's ass.
GTA is a fun game, but isn't particularly deep; it's not too difficult to make a semi-effective knockoff and bank on the fact that GTA is wildly popular
Of course, the reason that GTA still dominates is because it (especially in the form of both GTA3 and San Andreas) _is_ particularly deep in the genre, and the knockoffs are at best semi-effective. The clones tend to be urban shooters with some street racing aspects, and miss the other 90% of what GTA has. GTA is really more comparable to a single-player Sims than to True Crime.
[Rugby League, Rugby Union, Aussie Rules, and soccer]...You could also count Gailic football and there's the two slightly different American and Canadian types of Gridiron. Apart from that there's also indoor soccer.
And indoor football (gridiron), which (in the form of the Arena Football League) is at least as big (viewer-wise and money-wise) as Gaelic football or Canadian gridiron.
That's a controversial theory you're putting forward as fact. Lillienfeld certainly patented the MESFET and then followed on with the MOSFET and other transistors later on, and Heil followed on with additional refinements.
It's possible that they just came up with methods of building working transistors, never tried to build them, and then came up with more practical ways to build them, but it seems pretty unlikely. It's much more likely that the engineering issues of trying to build reliable solid-state devices in the 1930s made it impossible to build anything more than unreliable prototypes.
Ultimately, whether they built anything working is an open question and stating that they didn't as fact is misleading at best.
Whichever it was, it's not like they just had some vague ideas. Indeed, Bell had to abandon their attempts to patent FETs because of Lilienfeld and Heil's prior art.
Further, the Nobel the Bell crew won was clearly erroneous; it credited the Bell researchers with discovering the transistor effect, and whether you believe he built a transistor or not, Lillienfeld (and Heil) clearly discovered and described (and published papers on!) the transistor effect.
Sorry about the "founded years earlier" portion of my rant, you had that covered correctly (must've been scrolled off the top of my screen or something).
AT&T was actually founded years earlier (in the mid-1870s) and was renamed to AT&T in the 1880s. When Vail took over he lobbied heavily for the "One Policy, One System" mantra and worked aggressively to get government rights of way and even for government usage of eminent domain power to sieze private rights of way to grant to his company. When the government began investigating the company under his watch for monopolistic practice, he compromised (around 1912 or so) on a "we can buy one phone company if we sell another" policy and promptly started acquiring geographic monopolies--if he held a company in LA and another in New York, and a large competitor did likewise, they'd swap so that one owned New York and the other owned LA and there was no messy competition to keep prices down.
The only reason they got the European system working was because of the AT&T/RCA/GE/Westinghouse compromise a year earlier that colluded to give each company a monopoly in a certain sector (and incidentally resulted in the formation of broadcasting powerhouse NBC)
Earlier, Vail lobbied heavily to get the federal government to "nationalize" phone service by giving AT&T a monopoly, and by 1918 that came to pass and they had a government-mandated control of the market which lasted until the early 1980s.
Ones they had that, they used it aggressively to exterminate small inventors. By the 1980s, they had banners hanging in corporate offices saying something like "There are 2 major forces in the US; one brought you phones, radios, radar, sonar, transistors, etc and the other brought you the Civil War, WWI, WWII, Vietnam, etc--guess which one is trying to tell the other one what to do?"
That despite the fact that AT&T had been denied patents on the transistor because it was previously invented and refined by Lillienfield and Heil (although if you look up AT&T scietists Brattain, Bardeen, and Shockley in wikipedia, it lists them as the inventors with a straight face), they'd acted to bankrupt Tesla and leave Marconi as unknown as possible as the inventors of the radio, and they'd similar coopted the radar from Hulsmeyer and Bay.
Any argument that they (or Vail) were in favor of free-market private enterprise and against government subsidies and grants has a lot of evidence to the contrary that it needs to address.
AT&T is private. Much of their infrastructure was originally government funded (they are one of several companies that replaced Bell when Bell was broken up in the early 1980s). Partially because of that, and partially because of historical monopoly concerns, and partially because telcom is considered basic infrastructure, they are more heavily regulated than many private companies. Consequently, they tend to have an interest in making nice with the feds.
For example, when you speak, what do you do to separate words form one another? The surprising answer is, nothing. Take a tape of ordinary conversation. Run it through an oscilloscope. Look for the breaks. You won't find them. We "blur" words together in sentences. (I suspect this is why anyone speaking a different tongue always sounds like he/she is speaking very quickly... your brain hasn't learned to put the "spaces" back in by context.)
One of my Spanish classes way back in junior high was explaining how words tend to get run together and malformed in the spoken language, which drew some anger from the class about how they had to not only learn the words themselves but how to understand "slurred" speech.
The teacher went over to the chalkboard and wrote "that's true, but watcha gonna do about it?"
In fact, the European Union is even the largest economy in the world.
No it isn't. The NAFTA zone is significantly larger economy.
Of course, you could say that the EU has a larger economy than the United States. But that makes as much sense as saying the NAFTA states have a larger economy than Germany, or that the Western Hemisphere has a larger economy than China.
At least around here (Washington, DC area) the primary purpose of parking meter's isn't revenue. It's to keep people from parking all day, allowing for turnover and giving access to people who aren't there all day. Employees park blocks away or in garages or whatever; people doing grocery shopping, eating lunch, getting their hair cut, etc use the metered spaces.
If it was about maximizing meter revenues, it would be legal to feed the meter and stay more than the alloted 2 hours. But they've determined turning over the meters gets more revenue coming into the city's businesses (I have no idea if that's a correct conclusion, but it is their rationale).
Slashdot Mirror
From wikipedia's entry on gelatin: "Contrary to popular belief, horns and hooves are not commonly used."
Being from the US, our jelly is always smooth.
Homogeneous grape/blackberry/mint stuff = jelly. This stuff is of uniform consistency, often overly sweetened, and usually clear as opposed to cloudy.
Smashed up strawberry/raspberry/blackberry/etc stuff = jam. This stuff has seeds, skins, etc in it.
Yes, blackberry comes both ways. Strawberry is usually jam when sold in grocery stores, but if you eat out you'll sometimes get little packets of strawberry jelly. Grape is invariably jelly.
If it's got citrus peel in it we call it marmalade instead of jam, and it's not nearly as common as the other kinds.
For readers in the US, Panadol is apparentl a foreign brand name for what's commonly sold here as Tylenol, and paracetamol is a foreign generic name for what's called acetaminophen here.
Here, we call that jeevesing.
The high-definition itself does help a lot following small, fast-moving objects, so as far as seeing the action it's more of a help in hockey or tennis than basketball or football.
Independently of how "helpful" it is, though, the hi-def is nice for the same reason it's nice in movies: it just looks better, and gives you more detail. And widescreen lets you see more of the product.
I mean, it's not like you can't follow what's going on in a VHS copy of Lawrence of Arabia. But god does it look better on hi-def widescreen (or 70mm projection if you get a chance).
Similarly, seeing all the fog players blow out on a cold day, the condition of the grass, dirt stains on the uniforms after a slide, etc don't help you follow the action any better but they do make it "feel" more immediate and make it more fun to watch.
Seeing a great film cut to pan-and-scan is just silly, you miss 30% or so of the visuals. Somewhat similarly, widescreen lets you see more. You can follow positioning and motion away from the ball, and really get a feel for team strategy instead of just following the action of whoever has the ball at the moment..
I'm no fan of Bush, but even if he were a good president he'd be scared for his life. About 10% of US Presidents have been assassinated in office, and another 20% or so have had serious attempts on their lives. The ones who were killed -- Garfield, McKinley, Lincoln, Kennedy -- aren't among our worst presidents, and Lincoln is probably one of our best.
That seperates them from the average person on the street.
Apart from restricted pointers, all of these (or work-alikes) are in C++.
[SNIP next three paragraphs saying that none of them are actually in C++ but other features can do the same thing]
Obviously being a Turing-complete language C++ can do anything C can. The point is that it is not a superset of C, and you can't just use a C++ compiler to compile your C code.
Use a C++ compiler for C++. Use a C compiler for C. And always remember that C++ is not an "extended C", nor is C a subset of C++. They are seperate (albeit related) languages).
Well, all I can say is that C++ is (so much) more than just a "stricter" C.
While I agree with your core point, I have to take exception to the implication that C++ is at all a stricter C (even if it's also more). C++ and C are different languages, and C is not a subset of C++. There are valid C programs that are invalid in C++ (even not using things like variables named "new", etc), and features like implicit void casting that C++ lacks. There are programs that are valid C and valid C++ but behave differently.
And that's without getting into features of modern C (variable size arrays, language built-in complex numbers, restricted pointers, etc) that are not in C++ as far as I know.
But as far as your main point, yes, the reason to use C++ is if you want/need C++ features. My original objection was to the suggestion that you just "write C but use a C++ compiler to add namespaces and nothing else". Many of the drawbacks of C++ compared to C are pretty minor, and may be worth the tradeoff if you're going to take advantage of a lot of language features. Writing "C in C++" is just silly, though.
The most common example has to do with void pointers:
int *a;
a = malloc(sizeof(int)*10);
vs.
a = (int *) malloc(sizeof(int)*10);
C++ requires the latter, which does nothing but make the code more brittle in the face of change. void * is supposed to convert implicitly to other types, that is its entire reason for its existence. Code gets more fragile when you have to sprinkle references to an object's type all over the place instead of having some minimal (bare-bones, craptastic) polymorphism.
The core problem is that both the C and C++ type systems are jokes. But C++ got stricter without fixes the system itself, so you wind up having to put additional cruft in your code to no positive end.
It doesn't necessarily have to be standardized completely, but there are some language bases that any reasonable CS curriculum will cover:
* At least a semester of a lower-level EE course working with basic architecture (logic gates, VHDL or similar)
* Assembly coding in at least one assembly language
* At least one lower-level statically typed language (C, Fortran, C++, etc)
* At least one strictly, statically typed language (ML, Haskell, etc)
* At least one dynamically-typed language (scheme, smalltalk, lisp, python, etc)
* At least one static class-oriented language (C++, Java, etc)
* At least one more dispatch-oriented OO language (smalltalk, CLOS, python, etc)
[My personal preference would be to cover these with some RISC assembly, C, ML, lisp, python, and java; lisp especially gives you a lot of other features that are worth discussing.]
There are other language niches that are arguably important to cover, arguably stack-based languages (Forth, Postscript), non-class-oriented OO languages (self, JavaScript), etc. But at least the ones above deserve to be addressed in any respectable CS course.
Patterns aren't core CS, they're just a jargon that Java programmers use to describe common program designs. Some of them aren't often applicable to all languages or have different, older names used by non Java-heads that predate the Gang of Five's choices. They're certainly worth covering if you're going to do a lot of class-oriented programming but think of them as more of a common lexicon than anything deeper.
For the same reason that Feynman diagrams won a Nobel prize. There's often very high value in explaining something in a way that can be easily understood.
1. You'll wind up fighting the compiler figuring out how to turn off exceptions, rtti, etc
2. You don't gain much over lexical scoping.
3. C is not a subset of C++ (even forgetting trivialities like variables named "new", etc). You won't be able to re-use existing C code without strictly auditing it because of subtle difference (e.g. on 32-bit Intel, sizeof('a') is 4 in C but it's 1 in C++). Those differences are often of the "it compiles fine but behaves differently" type, which really sucks.
4. You get saddled with C++'s crappy stricter typing. When your type system is crap, making it stricter hurts rather than helps.
And stackless Python has string objects _and_ continuations.
No. 3 months is _way_ too long.
The standard used to be notify the vendor and wait forever. Vendors never fixed anything. Bugtraq and other security lists implemented a "full and immediate disclosure" policy, and bugs started getting fixed.
Lately, full and immediate has been pushed back to full and one month. That's a compromise so that in the unlikely event that hackers aren't already exploiting the bug it can be fixed by responsible companies before they get a chance to do so.
The drawbacks of the delay are:
1. The consumer is not informed about their risks; I might decide not to run that app at home, or not to do online banking on the same machine where I run the vulnerable app. I might be able to firewall off or otherwise protect myself using faculties outside the affected app.
2. The irresponsible vendors now just don't start working on the bug until a month later, so fixes get delayed. Moreover, when immediate disclosure was the norm there was no yelling at the person who disclosed the bug as though they had done something wrong; now, companies are trying to make it seem as though the people who discover the holes are somehow the ones in the wrong instead of the vendors/authors who wrote the insecure code in the first place.
Really, the one month delay ought to be the norm only with companies/individual who show that they actually use the month to work on a fix and often manage to get one out. For those who ignore things until they're disclosed, we should go back to immediate disclosure.
And absolutely as a consumer I want to know if there's a problem so that I can take steps to protect myself. The month is an okay compromise but if you're leaving known holes open and unpatched, and not letting me know so I can decide how to deal with that risk, it's hurting me. And it's hurting other consumers.
And ultimately the goal of 3rd-party security researchers should be to protect the consumers and users, not to cover the company's ass.
GTA is a fun game, but isn't particularly deep; it's not too difficult to make a semi-effective knockoff and bank on the fact that GTA is wildly popular
Of course, the reason that GTA still dominates is because it (especially in the form of both GTA3 and San Andreas) _is_ particularly deep in the genre, and the knockoffs are at best semi-effective. The clones tend to be urban shooters with some street racing aspects, and miss the other 90% of what GTA has. GTA is really more comparable to a single-player Sims than to True Crime.
[Rugby League, Rugby Union, Aussie Rules, and soccer]...You could also count Gailic football and there's the two slightly different American and Canadian types of Gridiron. Apart from that there's also indoor soccer.
And indoor football (gridiron), which (in the form of the Arena Football League) is at least as big (viewer-wise and money-wise) as Gaelic football or Canadian gridiron.
That's a controversial theory you're putting forward as fact. Lillienfeld certainly patented the MESFET and then followed on with the MOSFET and other transistors later on, and Heil followed on with additional refinements.
It's possible that they just came up with methods of building working transistors, never tried to build them, and then came up with more practical ways to build them, but it seems pretty unlikely. It's much more likely that the engineering issues of trying to build reliable solid-state devices in the 1930s made it impossible to build anything more than unreliable prototypes.
Ultimately, whether they built anything working is an open question and stating that they didn't as fact is misleading at best.
Whichever it was, it's not like they just had some vague ideas. Indeed, Bell had to abandon their attempts to patent FETs because of Lilienfeld and Heil's prior art.
Further, the Nobel the Bell crew won was clearly erroneous; it credited the Bell researchers with discovering the transistor effect, and whether you believe he built a transistor or not, Lillienfeld (and Heil) clearly discovered and described (and published papers on!) the transistor effect.
Sorry about the "founded years earlier" portion of my rant, you had that covered correctly (must've been scrolled off the top of my screen or something).
AT&T was actually founded years earlier (in the mid-1870s) and was renamed to AT&T in the 1880s. When Vail took over he lobbied heavily for the "One Policy, One System" mantra and worked aggressively to get government rights of way and even for government usage of eminent domain power to sieze private rights of way to grant to his company. When the government began investigating the company under his watch for monopolistic practice, he compromised (around 1912 or so) on a "we can buy one phone company if we sell another" policy and promptly started acquiring geographic monopolies--if he held a company in LA and another in New York, and a large competitor did likewise, they'd swap so that one owned New York and the other owned LA and there was no messy competition to keep prices down.
The only reason they got the European system working was because of the AT&T/RCA/GE/Westinghouse compromise a year earlier that colluded to give each company a monopoly in a certain sector (and incidentally resulted in the formation of broadcasting powerhouse NBC)
Earlier, Vail lobbied heavily to get the federal government to "nationalize" phone service by giving AT&T a monopoly, and by 1918 that came to pass and they had a government-mandated control of the market which lasted until the early 1980s.
Ones they had that, they used it aggressively to exterminate small inventors. By the 1980s, they had banners hanging in corporate offices saying something like "There are 2 major forces in the US; one brought you phones, radios, radar, sonar, transistors, etc and the other brought you the Civil War, WWI, WWII, Vietnam, etc--guess which one is trying to tell the other one what to do?"
That despite the fact that AT&T had been denied patents on the transistor because it was previously invented and refined by Lillienfield and Heil (although if you look up AT&T scietists Brattain, Bardeen, and Shockley in wikipedia, it lists them as the inventors with a straight face), they'd acted to bankrupt Tesla and leave Marconi as unknown as possible as the inventors of the radio, and they'd similar coopted the radar from Hulsmeyer and Bay.
Any argument that they (or Vail) were in favor of free-market private enterprise and against government subsidies and grants has a lot of evidence to the contrary that it needs to address.
AT&T is private. Much of their infrastructure was originally government funded (they are one of several companies that replaced Bell when Bell was broken up in the early 1980s). Partially because of that, and partially because of historical monopoly concerns, and partially because telcom is considered basic infrastructure, they are more heavily regulated than many private companies. Consequently, they tend to have an interest in making nice with the feds.
For example, when you speak, what do you do to separate words form one another? The surprising answer is, nothing. Take a tape of ordinary conversation. Run it through an oscilloscope. Look for the breaks. You won't find them. We "blur" words together in sentences. (I suspect this is why anyone speaking a different tongue always sounds like he/she is speaking very quickly... your brain hasn't learned to put the "spaces" back in by context.)
One of my Spanish classes way back in junior high was explaining how words tend to get run together and malformed in the spoken language, which drew some anger from the class about how they had to not only learn the words themselves but how to understand "slurred" speech.
The teacher went over to the chalkboard and wrote "that's true, but watcha gonna do about it?"
In fact, the European Union is even the largest economy in the world.
No it isn't. The NAFTA zone is significantly larger economy.
Of course, you could say that the EU has a larger economy than the United States. But that makes as much sense as saying the NAFTA states have a larger economy than Germany, or that the Western Hemisphere has a larger economy than China.
No, it's at best the second largest. The NAFTA zone is larger.
It's just duplicitous to compare an entire trading block with one member of another.
At least around here (Washington, DC area) the primary purpose of parking meter's isn't revenue. It's to keep people from parking all day, allowing for turnover and giving access to people who aren't there all day. Employees park blocks away or in garages or whatever; people doing grocery shopping, eating lunch, getting their hair cut, etc use the metered spaces.
If it was about maximizing meter revenues, it would be legal to feed the meter and stay more than the alloted 2 hours. But they've determined turning over the meters gets more revenue coming into the city's businesses (I have no idea if that's a correct conclusion, but it is their rationale).
FWIW, I've never had trouble paying for plane tickets with cash. If you do it the day of the flight, expect to be flagged for additional screening.