I'm also pretty good with understanding musical relationships by ear, and that's something I learned in the 8-14 age, so possibly relevant.
And there you go. See, the thing is, if you grew up without being exposed to tonal languages, you may very literally not be able to hear the differences in tones when you're an adult. That makes it a little hard to learn languages like Chinese.
So what is your point, exactly? It's a compiler bug when you turn off optimization and the code starts working properly. Optimization is not supposed to cause the code to behave differently.
And in the second case, the "branch to self" was clearly visible in the assembly code of the.lst file.
I also had a case where the compiler generated an instruction with an invalid operand, which resulted in an assembler error! Fortunately that one was fixable by adding a "volatile" keyword to a variable declaration. And all of these are within the past two years.
The biggest case of finding compiler bugs for me was in college back in the '80s. I was writing a Pascal compiler for a CS class, and kept finding all sorts of bugs in TML Pascal. The most annoying was if you had a set with more than 32 elements, "constant IN set" would be compiled as a 32-bit AND instruction, completely ignoring some of the involved data. Then Apple made MPW free (as in beer) and I didn't have to worry about those bugs anymore.
I've had optimizer problems that caused SPI transfers to break (resulting in bricked SD cards!), probably due to optimizing out a wait loop, and another that replaced a call to a delay subroutine with a branch-to-self instruction. (FWIW: It was A commeRcial cross-compiler, not gcc or LLVM)
The problem with Chinese is the tones. Depending on your genetic material, as an adult you may find it very difficult to come to grips with them.
It's not genetic. It's how your brain changes as you grow up. If you're exposed to it as a child (like maybe age 8-14 or so), you can recognize the tones as an adult. Once you're an adult, those parts of your brain have already become 'set" like jello, and it is much harder to learn them.
Or don't. It's a supposedly "international" language, but it's based on Eastern European languages, with their idiosyncrasies and strange phonemes. It might be easy to learn if you already know Polish or Czech, but otherwise it's overrated.
1: Learning to read all over again was quite interesting. I had to stop at around the 5th grade level, though, because I ran out of free time for it, and I had intentionally skipped learning the handwriting. Handwriting practice seems to be very important once you get to the 5th-6th grade level just so that you'll be familiar enough with the kanji to notice the subtle differences.
2: Forget about learning it for work purposes unless you're going to take the effort to become completely fluent in it, as in native speaker who grew up in-country level of fluency with cultural knowledge, which is a lot of work to learn as an adult. The Japanese have to take English classes in school, and it's a matter of pride to speak it (or to be the one in the company called upon to speak it) when the rare gaijin appears, even when the gaijin is speaking back in perfect Japanese. Even when the speaker has a decent knowledge of English (uncommon), the pronunciation is usually atrocious (because they don't teach that in high school classes!) And when you're not in Japan, those who contact you will be the ones who do actually know decent English.
The only reason to learn it for work is you're in your 20s and want to go over there as an exchange English teacher, and maybe make a slight dent in the bad pronunciation over there.
Learn it for fun, it's a pretty cool language, but don't learn it for work.
Grrrrr... on the internet you can always tell who is a native German speaker, because they always use* "does $FOO" for the present tense of verbs. Native English speakers normally only use that as an emphatic form. It twists something around in my brain whenever I see that.
Q: What do you call a person who only speaks 1 language? A: American
FTFY.
If I could go back in time and choose again I think I'd go for Spanish, which I would actually have found useful from time to time.
I took two semesters of Spanish in college. And I live in central Texas, where there are TV stations in Spanish. I eventually realized that about the only good thing it could get me (aside from a low-paying service-sector job) was a few more news reports about NASA.
But it was great for teaching myself Japanese. They both have similar vowel sounds (but Japanese has stuff like hyo, ryo, etc. which most English speakers can't grok, even though they have similar sounds with other consonants), and both have heavily-conjugated verbs. In fact, it overlaid what little Spanish I learned such that I want to use Spanish words with Japanese grammar.
I remember I had to turn mine off if the family wanted to watch TV channel 12. We lived far enough out in a rural area that we had to use a nice big antenna. which only made things worse.
The cassette port was often used for sound output from games, but the very act of doing the timing for sound made it not much worse to just put an AM radio next to it. You kids and your Bluetooth headsets, we had REAL wireless audio back in the day!
Oh, it's perfectly safe from fire. See, a hydrocarbon world like that is a chemical Bizarro World. It's the oxidizers that you have to keep under control.
Their very lameness and ubiquity is what makes them perfect for people living under oppressive regimes. When "everybody" uses a mail service, it becomes harder to block it without a lot of people noticing and getting pissed off. When they are so ubiquitous that even members of the regime use them, it's even better.
The information is already being sent in clear text. The only reason to encrypt it would be to avoid automated blocking.
Since all the domains are composed of two English words, they could be sent as two words, with a space in between, and possibly another codeword to indicate.com,.org, or some other TLD. That would remove the "scan the message for anything that looks like a domain name I haven't seen before and scan it for open proxies" angle.
Perhaps he could use rot13, or a substitution cipher with the key as the first line. or some similar encoding that is easily decoded both manually and with a 10-line C program. It might even be possible for every message to use a different key, and the automation necessary to do that would also let you give everybody a different random subset of the domain names. These are meant to be good for 1-4 weeks, so taking 5 minutes to decode them should be insignificant.
Getting even more tough about using low-impact steganography, the text could be converted into small 1-bit.gif encodings of the domain names and inserted as MIME data. This would be hard to reject automatically by means other than the fact that they are encoded, but a decent GUI mail reader would be able to display them. If the font is kept consistent, they can even be converted back to plain text.
It's a balance between something easy enough to decode, yet difficult enough to be annoying for automated detection. Remember, true spammers need to have their message visible as plain text because their recipients aren't motivated to decode it, but someone who really wants the data can be expected to take five minutes to decode it.
Slashdot Mirror
I'm also pretty good with understanding musical relationships by ear, and that's something I learned in the 8-14 age, so possibly relevant.
And there you go. See, the thing is, if you grew up without being exposed to tonal languages, you may very literally not be able to hear the differences in tones when you're an adult. That makes it a little hard to learn languages like Chinese.
So what is your point, exactly? It's a compiler bug when you turn off optimization and the code starts working properly. Optimization is not supposed to cause the code to behave differently.
And in the second case, the "branch to self" was clearly visible in the assembly code of the .lst file.
I also had a case where the compiler generated an instruction with an invalid operand, which resulted in an assembler error! Fortunately that one was fixable by adding a "volatile" keyword to a variable declaration. And all of these are within the past two years.
The biggest case of finding compiler bugs for me was in college back in the '80s. I was writing a Pascal compiler for a CS class, and kept finding all sorts of bugs in TML Pascal. The most annoying was if you had a set with more than 32 elements, "constant IN set" would be compiled as a 32-bit AND instruction, completely ignoring some of the involved data. Then Apple made MPW free (as in beer) and I didn't have to worry about those bugs anymore.
I've had optimizer problems that caused SPI transfers to break (resulting in bricked SD cards!), probably due to optimizing out a wait loop, and another that replaced a call to a delay subroutine with a branch-to-self instruction. (FWIW: It was A commeRcial cross-compiler, not gcc or LLVM)
Aha, just found the link I was looking for: http://www.xibalba.demon.co.uk/jbr/ranto/
crap, replied to wrong post
Aha, just found the link I was looking for: http://www.xibalba.demon.co.uk/jbr/ranto/
Because if it put in 9 legs, it wouldn't have enough legs of its own to pull the strings, duh!
The problem with Chinese is the tones. Depending on your genetic material, as an adult you may find it very difficult to come to grips with them.
It's not genetic. It's how your brain changes as you grow up. If you're exposed to it as a child (like maybe age 8-14 or so), you can recognize the tones as an adult. Once you're an adult, those parts of your brain have already become 'set" like jello, and it is much harder to learn them.
Or don't. It's a supposedly "international" language, but it's based on Eastern European languages, with their idiosyncrasies and strange phonemes. It might be easy to learn if you already know Polish or Czech, but otherwise it's overrated.
1: Learning to read all over again was quite interesting. I had to stop at around the 5th grade level, though, because I ran out of free time for it, and I had intentionally skipped learning the handwriting. Handwriting practice seems to be very important once you get to the 5th-6th grade level just so that you'll be familiar enough with the kanji to notice the subtle differences.
2: Forget about learning it for work purposes unless you're going to take the effort to become completely fluent in it, as in native speaker who grew up in-country level of fluency with cultural knowledge, which is a lot of work to learn as an adult. The Japanese have to take English classes in school, and it's a matter of pride to speak it (or to be the one in the company called upon to speak it) when the rare gaijin appears, even when the gaijin is speaking back in perfect Japanese. Even when the speaker has a decent knowledge of English (uncommon), the pronunciation is usually atrocious (because they don't teach that in high school classes!) And when you're not in Japan, those who contact you will be the ones who do actually know decent English.
The only reason to learn it for work is you're in your 20s and want to go over there as an exchange English teacher, and maybe make a slight dent in the bad pronunciation over there.
Learn it for fun, it's a pretty cool language, but don't learn it for work.
Because then you would know how to say "Desea papas fritas con eso?"
It's not really useful outside of service sector or public sector jobs. Or if you just want to know what the cleaning ladies are giggling about.
All ze Germans I know does speak the good English
Grrrrr... on the internet you can always tell who is a native German speaker, because they always use* "does $FOO" for the present tense of verbs. Native English speakers normally only use that as an emphatic form. It twists something around in my brain whenever I see that.
*sorry, I mean they "always do use"!
Q: What do you call a person who only speaks 1 language? A: American
FTFY.
If I could go back in time and choose again I think I'd go for Spanish, which I would actually have found useful from time to time.
I took two semesters of Spanish in college. And I live in central Texas, where there are TV stations in Spanish. I eventually realized that about the only good thing it could get me (aside from a low-paying service-sector job) was a few more news reports about NASA.
But it was great for teaching myself Japanese. They both have similar vowel sounds (but Japanese has stuff like hyo, ryo, etc. which most English speakers can't grok, even though they have similar sounds with other consonants), and both have heavily-conjugated verbs. In fact, it overlaid what little Spanish I learned such that I want to use Spanish words with Japanese grammar.
Right, but see, eventually you'll run out of lead, because you've turned it all into gold. So you need to get more lead!
I hear we've already got a 22km long hole dug in the ground for them here in Texas. Okay, so it's oval and not linear, but it's already been dug.
He was one of the few things she was not paranoid about. So, no.
She did get one thing right, though. Her world did indeed end in December 2012.
Run without installing Linux?
Maybe he was just a crazy paranoid asshole.
Who apparently had a crazy paranoid mother.
The fun part would be figuring out how to test such a beast.
I remember I had to turn mine off if the family wanted to watch TV channel 12. We lived far enough out in a rural area that we had to use a nice big antenna. which only made things worse.
The cassette port was often used for sound output from games, but the very act of doing the timing for sound made it not much worse to just put an AM radio next to it. You kids and your Bluetooth headsets, we had REAL wireless audio back in the day!
...named after Kalvin Coolidge, because he's so cool he can distill nitrogen out of the atmosphere.
Oh, it's perfectly safe from fire. See, a hydrocarbon world like that is a chemical Bizarro World. It's the oxidizers that you have to keep under control.
Their very lameness and ubiquity is what makes them perfect for people living under oppressive regimes. When "everybody" uses a mail service, it becomes harder to block it without a lot of people noticing and getting pissed off. When they are so ubiquitous that even members of the regime use them, it's even better.
The information is already being sent in clear text. The only reason to encrypt it would be to avoid automated blocking.
Since all the domains are composed of two English words, they could be sent as two words, with a space in between, and possibly another codeword to indicate .com, .org, or some other TLD. That would remove the "scan the message for anything that looks like a domain name I haven't seen before and scan it for open proxies" angle.
Perhaps he could use rot13, or a substitution cipher with the key as the first line. or some similar encoding that is easily decoded both manually and with a 10-line C program. It might even be possible for every message to use a different key, and the automation necessary to do that would also let you give everybody a different random subset of the domain names. These are meant to be good for 1-4 weeks, so taking 5 minutes to decode them should be insignificant.
Getting even more tough about using low-impact steganography, the text could be converted into small 1-bit .gif encodings of the domain names and inserted as MIME data. This would be hard to reject automatically by means other than the fact that they are encoded, but a decent GUI mail reader would be able to display them. If the font is kept consistent, they can even be converted back to plain text.
It's a balance between something easy enough to decode, yet difficult enough to be annoying for automated detection. Remember, true spammers need to have their message visible as plain text because their recipients aren't motivated to decode it, but someone who really wants the data can be expected to take five minutes to decode it.
I'll have you know we sell bottles of soda by the liter! Well, only the bigger ones, but that's metric!