>>16 There's a cover story going around that claims it was named for no operation, but it was actually named for monopolizing your processor. I heard this from the underground illuminati molepeople so it must be true.
Name:
Anonymous2021-03-21 9:59
Floats were a mistake. They are the source of the highest percentage of bugs among beginners (``why is 0.3333333 * 3 not == 1 the computer is retarded''), intermediate (``lol .1+.2!=.3, upvote on hackernews plox''), and experienced (``There is a corner case with negative infinity that we have to handle for compliance with a standard drafted by a deranged lisper'') programmers.
Name:
Anonymous2021-04-06 6:41
sockets
Name:
Anonymous2021-04-06 9:44
>>20 They aren't even close to the most common bug. Just use rationals.
Name:
Anonymous2021-04-06 13:33
>>1 C++ iostreams templates: Have anyone tried to use this for anything more complex than toy programs?
Name:
Anonymous2021-04-06 14:19
>>23 I've seen iostream used in the original Majesty game, when I decompiled it. They are supposed to be much faster than fread, since templates get inlined.
Name:
Anonymous2021-04-06 16:54
>>22 Rationals are not in the stdlib of most common languages.
Name:
Anonymous2021-04-06 17:13
>>22 It is impossible to implement rationals efficiently.
Name:
Anonymous2021-04-06 18:38
>>26 Most applications are not so performance-critical that a 1ms arithmetic operation would be noticeable.
>>26 Bignum rationals are more than efficient enough for most programs. Even if they weren't you could go for fixed-point numbers instead or have fixnum rationals with proper overflow handling. You do check for overflow, right?
Name:
Anonymous2021-04-09 13:36
>>28 How many ROPS vs FLOPS would your computer perform?
Name:
Anonymous2021-04-09 13:39
>>27 Most apps need only integers, and when they need floats, they use them for GPU or some engineering simulation, which usually needs to be as fast as possible. The only use for rationals are the financial apps, where any imprecision means lost money.
Name:
Anonymous2021-04-09 13:40
>>29 Doesn't matter when floats are simply the wrong data type for your problem, which they are most of the time.
Name:
Anonymous2021-04-09 13:42
>>26 Efficient implementation of rationals is more a question of memory management than execution speed, and virtual memory solved that issue a long time ago with 64-bit address spaces (well 47 bits for userspace today) picking up the rest. Furthermore the parallelism issue can be solved by storing the limbs of an arbitrary-precision integer in a "long carry" format, so that there's (say) 12 bits of carry and 52 bits of body, allowing operations to produce denormal limbs, and catching carry overflow with a slowpath which normalizes the operands first.
This isn't common practice because libgmp is the practical standard, and it fixes a carryless format. But perhaps one day.
Name:
Anonymous2021-04-09 14:02
>>32 FrozenVoid, is that you? Just compress these rational into single bit values with your infinite compression algorithm.
Name:
Anonymous2021-04-09 22:02
Speed trumps everything. Rationals rarely are absolutely necessary.
Name:
Anonymous2021-04-09 22:09
I present the ultimate program. It can perform every imaginable task at the highest possible speed. int main(void) { return 0; }