That force profile will be jarring to the fingers. Good ergonomic keyboards start off with high initial resistance, decrease the resistance, and then *increase* it again before getting to the stop at the bottom. This is actually significant in avoiding RSI.
You know, this might actually work if you extremely limit the unauthenticated e-mail streams rather than dumping them entirely. This would allow legitimate e-mails from new acquaintances while strangling of spam. What I envision is that each server has a maximum e-mail reception rate from various other servers. This is structured (1) to encourage efficient flow, i.e. max_in_rate = max_out_rate, and (2) so that others are given apportionments based on the trust level. So for example servers within the same organization may have unlimited interchange rates (or for a complex network, limited to reflect the true capacity and encourage load balancing), edges between different ISPs may have rates established by contract, and unknown hosts are limited to a rate which allows legitimate e-mail but makes spamming excruciating. In addition to simple rates, other rules could be established about message flow to favor legitimate e-mail. The key benefit of this scheme is that e-mail delivery can be kept reliable: messages are either enqueued or rejected depending on capacity availability; at the same time, no information is given out about whether an e-mail address is valid or not when turning away spam. A potential downside is the huge number of mailhosts which would have to be tracked. However, I think if we set the threshold rate for untrusted servers correctly, we can greatly limit the number of mail servers which we need to track with larger rate allocations. Furthermore, traditional blacklisting can be used to demote servers rather than entirely cut them off, and if a particular mail server sees demand for two-way communication with a blacklisted mail server, it can ease the restrictions. Another possibility is to use a reliable non-real-time dual to QoS at the IP level for routes along which e-mail is sent, if it's possible to use the existing and emerging IPv6 standards and infrastructure for this purpose.
Actually, I wonder if you could engineer the springs and such in the Model M keys to only make ultrasonic noise? Is that at all possible or feasible? How would it work? Anyhow, personally I find the clicking brings cadence to my work. You really know you're typing away when you almost think you hear a machine gun rattling off a round!
"C++'s templates are basically textual substitution, you can't type-constrain template arguments, you just have to see if it works"
Hmmm...no, your remark refers to the pre-processor.
No they don't, but I can see how the wording can make you think otherwise. My point is that when you define a template, you cannot place constraints on template arguments other than by writing code which uses them in a certain way. This is because judgment whether a particular type can be substituted into a template is on a basis of "would the code compile if we simply rewrote it with that type in place of the template parameter?" Clearly, without this capability, it's much more difficult to make a family of different templates based on properties of the type with which we are instantiating.
WHEREAS, I was just kidding around,
WHEREAS, the totally lacking grammar in my above post is uncharacteristic of my previous posts,
WHEREAS, BitchKapoor is actually a good friend of mine,
RESOLVED, YHBT,
RESOLVED, but I'll fuck off anyway,
RESOLVED, now IHBT.
truely a tesamint to the sorry shape/.'s in when someone named "BitchKapoor" get's modded upto +5 Informative... just glad my 2-year-old can't read yet... or maybe he's just hiding it from me? anyone else wish this sight had an optinal obscenity filter for this kindof nonsense.
"Mail Call shows it all the time" -- you mean you're just claiming this stuff from TV? I've personally seen these FX(R) marking rounds the other posters have mentioned (new, in use and expired littering the ground), and have had to avoid sitting on the many paint spots splattered all over the mock-up training villages when conducting field experiments.
From the website you cited: "Contrary to local belief, the unique idea of a Duck Tour did not originate in Boston. Duck operations have been in existence in the Midwest for decades, and in fact, continue to thrive."
By Midwest, I believe they mean Wisconsin, or more specifically, the Wisconsin Dells a stupid tourist trap river/lake kind of area.
Unless, of course, the password hash is stored on the disk itself. After all, why add a little nonvolatile memory chip to a huge nonvolatile storage device?
Your car analogy is spot-on, IBM Thinkpads and Dell Latitudes are just worlds apart. But in my experience, both work fine under Linux, as long as you don't choose some weird video controller. In any event, make sure that you get a sufficient warranty, and don't break the LCD, since those generally aren't covered under warranty (but you should check -- better warrant coverage on that kind of thing could be the deciding factor). I had a T22, but it fell out of my bag on the stairs and cracked the LCD; now I'm stuck with a C610, and someone uses the IBM as a desktop.
I know what you mean, man, I can assemble an Athlon 64 system with the same specs with quality equipment from reputable dealers (mostly newegg) for under $900, with a 3-year warranty on all parts. For a while last year, vendors like Dell had good systems cheaper than what you or I could buy the parts for on the open market, but now it's again at least $200 cheaper to build it yourself.
But my lecturer says the main distinction is that languages fall into three categories: imperative (assembly, C, BASIC, etc.), functional (Haskell and all that) and logic (prolog). How does that fit with your idea of programming languages?
That is the standard categorization used, but you are right that there are other distinctions, such as line-based vs. structured. Similarly along these lines we have lexical vs. dynamic scoping, strong vs. weak typing, explicit vs. implicit typing, and sequential vs. implicit vs. explicit concurrency. There's also a general sense of "how much you have to type" to write or modify a program, i.e. how compact the notation is. Note that Visual Basic (and Quick Basic before it) are actually structured languages, unlike the line-number-oriented BASICs before them.
I think one of the most useful combinations of language features is something compact, between imperative and functional with structured, lexically scoped, implicit static typing with built-in support for (including fine-grain) explicit concurrency (explicit concurrency is required for on-line I/O, which conflicts with a pure functional language's simple input-compute-output-stop program model). Unfortunately, there aren't any popular languages like this (especially for good concurrency & I/O support).
Do we know what definition of "coding" Mr. Johnson was using? Maybe he doesn't realize what programming is, and was equating it with data entry, which is also called coding (you look at what people filled out on the survey and then enter it into the computer, e.g. reducing free-form answers to multiple choice).
That's a good point.
Slashdot Mirror
Uh, no, that would allow two different possible positions for you assuming distances can be estimated accurately, which is not really even true.
That force profile will be jarring to the fingers. Good ergonomic keyboards start off with high initial resistance, decrease the resistance, and then *increase* it again before getting to the stop at the bottom. This is actually significant in avoiding RSI.
You know, this might actually work if you extremely limit the unauthenticated e-mail streams rather than dumping them entirely. This would allow legitimate e-mails from new acquaintances while strangling of spam. What I envision is that each server has a maximum e-mail reception rate from various other servers. This is structured (1) to encourage efficient flow, i.e. max_in_rate = max_out_rate, and (2) so that others are given apportionments based on the trust level. So for example servers within the same organization may have unlimited interchange rates (or for a complex network, limited to reflect the true capacity and encourage load balancing), edges between different ISPs may have rates established by contract, and unknown hosts are limited to a rate which allows legitimate e-mail but makes spamming excruciating. In addition to simple rates, other rules could be established about message flow to favor legitimate e-mail. The key benefit of this scheme is that e-mail delivery can be kept reliable: messages are either enqueued or rejected depending on capacity availability; at the same time, no information is given out about whether an e-mail address is valid or not when turning away spam. A potential downside is the huge number of mailhosts which would have to be tracked. However, I think if we set the threshold rate for untrusted servers correctly, we can greatly limit the number of mail servers which we need to track with larger rate allocations. Furthermore, traditional blacklisting can be used to demote servers rather than entirely cut them off, and if a particular mail server sees demand for two-way communication with a blacklisted mail server, it can ease the restrictions. Another possibility is to use a reliable non-real-time dual to QoS at the IP level for routes along which e-mail is sent, if it's possible to use the existing and emerging IPv6 standards and infrastructure for this purpose.
Actually, I wonder if you could engineer the springs and such in the Model M keys to only make ultrasonic noise? Is that at all possible or feasible? How would it work? Anyhow, personally I find the clicking brings cadence to my work. You really know you're typing away when you almost think you hear a machine gun rattling off a round!
That is truly bizarre.
Hmmm...no, your remark refers to the pre-processor.
No they don't, but I can see how the wording can make you think otherwise. My point is that when you define a template, you cannot place constraints on template arguments other than by writing code which uses them in a certain way. This is because judgment whether a particular type can be substituted into a template is on a basis of "would the code compile if we simply rewrote it with that type in place of the template parameter?" Clearly, without this capability, it's much more difficult to make a family of different templates based on properties of the type with which we are instantiating.
LET IT FURTHER BE RESOLVED that I left off a line break. That sucks.
WHEREAS, I was just kidding around,
WHEREAS, the totally lacking grammar in my above post is uncharacteristic of my previous posts, WHEREAS, BitchKapoor is actually a good friend of mine,
RESOLVED, YHBT,
RESOLVED, but I'll fuck off anyway,
RESOLVED, now IHBT.
truely a tesamint to the sorry shape /.'s in when someone named "BitchKapoor" get's modded upto +5 Informative... just glad my 2-year-old can't read yet... or maybe he's just hiding it from me? anyone else wish this sight had an optinal obscenity filter for this kindof nonsense.
"Mail Call shows it all the time" -- you mean you're just claiming this stuff from TV? I've personally seen these FX(R) marking rounds the other posters have mentioned (new, in use and expired littering the ground), and have had to avoid sitting on the many paint spots splattered all over the mock-up training villages when conducting field experiments.
From the website you cited: "Contrary to local belief, the unique idea of a Duck Tour did not originate in Boston. Duck operations have been in existence in the Midwest for decades, and in fact, continue to thrive."
By Midwest, I believe they mean Wisconsin, or more specifically, the Wisconsin Dells a stupid tourist trap river/lake kind of area.
What do you call a wheelchair?
Actually, they use special paintball bullets for their M-16s.
Curiously, both of their first strips came out during the same week of March, 1951...
No, what you're talking about is an umlaut. But since Slashdot is Anglo-centric, we'll have to make do with "ueber."
Kingston makes memory boards, but not the chips.
No, no, you're supposed to call it '52 pickup, so the kid thinks it's about an old truck.
Unless, of course, the password hash is stored on the disk itself. After all, why add a little nonvolatile memory chip to a huge nonvolatile storage device?
Your car analogy is spot-on, IBM Thinkpads and Dell Latitudes are just worlds apart. But in my experience, both work fine under Linux, as long as you don't choose some weird video controller. In any event, make sure that you get a sufficient warranty, and don't break the LCD, since those generally aren't covered under warranty (but you should check -- better warrant coverage on that kind of thing could be the deciding factor). I had a T22, but it fell out of my bag on the stairs and cracked the LCD; now I'm stuck with a C610, and someone uses the IBM as a desktop.
I know what you mean, man, I can assemble an Athlon 64 system with the same specs with quality equipment from reputable dealers (mostly newegg) for under $900, with a 3-year warranty on all parts. For a while last year, vendors like Dell had good systems cheaper than what you or I could buy the parts for on the open market, but now it's again at least $200 cheaper to build it yourself.
That 128MB is just on the graphics card. The main memory is 512MB.
But my lecturer says the main distinction is that languages fall into three categories: imperative (assembly, C, BASIC, etc.), functional (Haskell and all that) and logic (prolog). How does that fit with your idea of programming languages?
That is the standard categorization used, but you are right that there are other distinctions, such as line-based vs. structured. Similarly along these lines we have lexical vs. dynamic scoping, strong vs. weak typing, explicit vs. implicit typing, and sequential vs. implicit vs. explicit concurrency. There's also a general sense of "how much you have to type" to write or modify a program, i.e. how compact the notation is. Note that Visual Basic (and Quick Basic before it) are actually structured languages, unlike the line-number-oriented BASICs before them.
I think one of the most useful combinations of language features is something compact, between imperative and functional with structured, lexically scoped, implicit static typing with built-in support for (including fine-grain) explicit concurrency (explicit concurrency is required for on-line I/O, which conflicts with a pure functional language's simple input-compute-output-stop program model). Unfortunately, there aren't any popular languages like this (especially for good concurrency & I/O support).
Do we know what definition of "coding" Mr. Johnson was using? Maybe he doesn't realize what programming is, and was equating it with data entry, which is also called coding (you look at what people filled out on the survey and then enter it into the computer, e.g. reducing free-form answers to multiple choice).
That's a good point.