Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
camccann
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
25 ms
·
241.
▲
by
camccann
17y ago
Since the overall reaction to this seems pretty negative, a moment of "devil's advocate": Yes, he sold MySQL--to Sun, not Oracle. I imagine he expected that Sun would continue MySQL's business model. He got lots of money, Sun got a new reve
242.
▲
by
camccann
17y ago
The similarity isn't entirely coincidental: http://en.wikipedia.org/wiki/Halting_problem#Relationship_wi... I wouldn't characterize either as "trivial", though. May seem obvious in hindsight, but both were a pretty big deal at the time.
243.
▲
by
camccann
17y ago
No, the problem is the dual-license strategy combined with changing business models. If it had been pure GPL from the start, anyone who couldn't work with that would never have used it in the first place. But because they sold licenses for
244.
▲
by
camccann
17y ago
Well, for a program of sufficient complexity there'll probably always be some aspect that can only be checked easily by unit testing--if only stuff where the program works as intended but the programmer's intent was mistaken.
245.
▲
by
camccann
17y ago
But it is easier when you're dealing with a "pure" function. In something like Haskell, a function that doesn't touch any of the stateful sequencing monads is guaranteed to neither alter nor depend on any sort of global state. In my exp
246.
▲
by
camccann
17y ago
You may want to reread my post with the sarcasm bit flipped to "on"; I was trying to make fun of static-typed languages that don't have type inference. Sorry about that. I was actually hacking on some Haskell just now, in fact, trying to fi
247.
▲
by
camccann
17y ago
Satisfying the type-checker can take a huge amount of extra code if you use type annotations everywhere. Even more so if you don't make good use of polymorphic functions. Which is of course ludicrous--no one would create a static-typed la
248.
▲
by
camccann
17y ago
It's just an anecdote, but I notice less complaining about parentheses, too. I still like the idea of "Lisp with syntactic whitespace". A few simple transformation rules to add implicit parentheses based on indentation would make it a lot
249.
▲
by
camccann
17y ago
You write unit tests and ensure 100% code coverage. That sounds like a lot of extra effort; it would be nice to automate the testing a bit. Maybe some sort of simple declarative system to check that the code obeys certain constraints. Ide
250.
▲
by
camccann
17y ago
You're implying nobody has any real control over what their actions and thereby assigning killers and rocks about the same level moral culpability. No, I'm just saying that the default state is to have neither control nor awareness of dec
251.
▲
by
camccann
17y ago
See here: http://en.wikipedia.org/wiki/Active_noise_control The basic idea of dual microphones to filter background from signal is pretty common.
252.
▲
by
camccann
17y ago
My point is that you're calling people jerks for something they probably have very little control over and are helplessly unaware of. If you're walking behind someone with a broken leg, do you complain that they should be walking faster? Pe
253.
▲
by
camccann
17y ago
To be fair, almost everything that happens in the brain has to do with dopamine one way or another.
254.
▲
by
camccann
17y ago
If memory serves me, the operating hypothesis for ADHD is actually that it's due to reduced activity in a portion of the brain that regulates and directs the rest, the theory behind most ADHD drugs being to increase activity in that porti
255.
▲
by
camccann
17y ago
I think you, like Aristotle, wildly overestimate conscious awareness. I'm not talking about doing wrong and believing it to be right, or even doing something without realizing it was wrong. I'm talking about just doing something without c
256.
▲
by
camccann
17y ago
It's worth noting that studies have shown that applying "willpower" works exactly that way. Specifically, making conscious decisions to override subconscious impulses depletes cognitive resources in a measurable way, with some evidence that
257.
▲
by
camccann
17y ago
You should have pointed out to them that fuel efficiency and pollution from automobile engines are worsened by having to decelerate/accelerate. In other words, if you force someone in a car to slow down (or, heaven forbid, stop) you're tang
258.
▲
by
camccann
17y ago
I think the reason that is so is that the HN system encourages certain behaviors which lead to better discussions whereas the slashdot system is more of a passive filter on a noisy feed, which is inherently a losing battle. HN survives mo
259.
▲
by
camccann
17y ago
You must have a tough time reading C code.
260.
▲
by
camccann
17y ago
I'm not impressed as much as jealous . Sure, maybe your program was very crude by most standards, but the amount of stuff you did figure out on your own was great! It sounds like you had a lot of fun in the process. Figuring things out mos
261.
▲
by
camccann
17y ago
I'd actually go so far as to say that anyone not impressed either doesn't understand what he did, or has a badly miscalibrated scale of what counts as impressive.
262.
▲
by
camccann
17y ago
You don't think people would buy and use such an app, if it existed?
263.
▲
by
camccann
17y ago
If memory serves me, Skeet works for Google--perhaps SO karma is his 20% project?
264.
▲
by
camccann
17y ago
I'd actually be inclined to get C++/C#/Java off the list entirely due to not showcasing any one perspective particularly well, to be honest. Smalltalk or Ruby would better illustrate OOP, and between them the OOP language and the assembly s
265.
▲
by
camccann
17y ago
Hmm, after looking at your rather nice list, six was probably too low; I think I'll revise my estimate to a more conservative eight. There'll always be some gap, of course, but eight ought to get closer to the point of diminishing returns
266.
▲
by
camccann
17y ago
Civil disobedience is appropriate for individuals, not institutions. You say this, yet oddly you you seem to put little moral weight on individual actions. If an institution collectively performs an act that is evil, and the law does not
267.
▲
by
camccann
17y ago
They're orthogonal in that both can vary independently of the other. Making an act legal or illegal does not alter the morality of the act, nor does the morality of an act in a given context change the legality of performing it. To the exte
268.
▲
by
camccann
17y ago
I think the advice to learn a new language is more incomplete than anything else; it's not learning a new language that's valuable, it's learning new ways to think about problems. Picking up a language that's very similar to one you already
269.
▲
by
camccann
17y ago
Unfortunately, there are also large classes of interesting problems which are intractable even with quantum computation.
270.
▲
by
camccann
17y ago
Following evil laws is evil. Legality and morality are orthogonal, not isomorphic, concepts.
More ›