HNHacker News
TopNewBestAskShowJobs

chrchang523

591 karma · joined March 8, 2017

submissionscomments
chrchang523··on Sorry everybody, I failed with you
There is an interesting "converse" to this observation for open-source projects, though. Namely: if a nontrivial abandoned project is still widely used, it probably represents a major career opportunity for anyone willing to take over maintenance.

This isn't purely hypothetical: it's the foundation of my own career.

chrchang523··on Ideal divisors: when a division compiles down to just a multiplication
See the "Labor of Division" blog posts by the author of libdivide:

https://ridiculousfish.com/blog/posts/labor-of-division-epis...

https://ridiculousfish.com/blog/posts/labor-of-division-epis...

chrchang523··on Ideal divisors: when a division compiles down to just a multiplication
See also https://libdivide.com/ , when you need to repeatedly divide by a constant that isn't known until runtime.
chrchang523··on Embrace the Grind
I had forgotten that the schlep essay was distinct. Interesting that they don't refer to each other.
chrchang523··on Embrace the Grind
Alternate framing: be prepared to "Do Things that Don't Scale" (http://paulgraham.com/ds.html ) when your organization is small.
chrchang523··on Wuffs the Language
The purpose of the language is clear to me: make it practical to prove that >90% of a C file-munging library is safe from several common types of errors. I will be looking into reorganizing some of my existing C library code of this type into large Wuffs components and small interaction-with-outside-world components.
chrchang523··on Summary of C/C++ integer rules
Also, go to the Compiler Explorer and compare the generated code for C++ "num / 2" when num is an int, and when num is an unsigned int.

While there are a few cases where the compiler tends to do a better job of optimizing signed ints than unsigned ints (generally by exploiting the fact that signed integer overflow is undefined), they are not as fundamental as "num / 2". Being forced to write "num >> 1" all over the place whenever I care about performance is basically a dealbreaker for me in many projects; and I haven't even gotten into the additional safety issues introduced by undefined overflow.

chrchang523··on No one has been murdered in Norway so far this year
The Svalbard archipelago is governed by Norway, and does have very open borders.
chrchang523··on How to implement a hash table in C
Not an issue since the following is at the top of the function:

  if (size + size < size) {
    return NULL; // size too big; avoid overflow
  }
chrchang523··on How to implement a hash table in C
Well, he did say "simple hash table". In my experience, many applications have no need for single-element deletion.
chrchang523··on Are You Trading or Gambling?
I assume you've heard of index funds; out of curiosity, what makes that investing choice unappealing to you?
chrchang523··on Gridsort: A stable sort faster than std:sort
Interesting work.

License is worth noting. I am considering an "#ifndef FREE_MARS" std::sort/qsort fallback for any code I write that uses this.

chrchang523··on Weird compiler bug – Same code, different results
I currently build Windows binaries via cross-compilation on Linux using gcc-mingw-w64; I assume that is affected by the bug?
chrchang523··on GameStop Is Rage Against the Financial Machine
The inferior tax treatment of put options relative to index funds also matters. Not tempted.
chrchang523··on GameStop Is Rage Against the Financial Machine
I've been keeping an eye on the far-out-dated put options, and I haven't seen anything that's more compelling than simply leaving money parked in an index fund. The expectation that GME will drop back down to Earth over that timeframe remains priced in.

The main takeaway from this incident is that margin-call-constrained short selling is even more dangerous than previously understood.

chrchang523··on The platforms have acted, raising hard questions about technology and democracy
My primary prediction is that the main US platforms will fall behind homegrown alternatives in India over the next 1-2 decades, >80% likelihood.
chrchang523··on How Silicon Valley destroyed Parler
If this was primarily about "[objecting] to the US not having laws that require this behavior", their emphasis would not have been on the platforms being out of line.
chrchang523··on How Silicon Valley destroyed Parler
The primary message from all of these governments is that the platforms are out of line.

Your original comment was in support of/elaborating on "It would be governmental overreach [to set the rules]." That original comment had an overly permissive rule in place of what I've bracketed, but that's beside the point since all of these governments are specifically objecting that the recent actions are problematically restrictive.

chrchang523··on How Silicon Valley destroyed Parler
It directly contradicts your use of the statement as justification for the platforms' behavior.
chrchang523··on How Silicon Valley destroyed Parler
Did you actually read what Merkel or AMLO said? The first sentence of your second paragraph ("So France and Germany could be seen as asking the US government to take a stronger stance on hate speech.") can be immediately verified to be false in the sense that you are stating it (as justification for the platforms' behavior).
chrchang523··on How Silicon Valley destroyed Parler
The issue is that the corporations are acting as the government. As you noted, Germany restricts speech more than the US for obvious historical reasons. So the fact that even their government objects to this behavior is evidence against your position, not mine.

In your linked comment, all of these governments recognize (3) as a scenario that should be governed by laws (which vary between countries).

chrchang523··on How Silicon Valley destroyed Parler
Let's set aside the distinction between free speech rights guaranteed by the First Amendment vs. the broader free speech ideal that's foundational to our society, since it isn't even needed: as suggested by the objections from Germany, France, and Mexico noted by Greenwald, these corporations are effectively acting as the government, so the reasoning behind the First Amendment's existence directly applies here.

(This is not an endorsement of the delusional presidential behavior that created the leadership vacuum filled by the corporations.)

chrchang523··on Why mmap is faster than system calls (2019)
Yeah, this is the biggest reason I stay the hell away from mmap now. Signal handlers are a much worse minefield than error handling in any standard file I/O API I've seen.
chrchang523··on Tech giants will block Kazakhstan’s web surveillance efforts again
Correct, they couldn't keep millions of people in camps, but the leading 13th century power just went ahead and killed millions of people instead (at a time when world population was much lower).
chrchang523··on Biden wins White House, vowing new direction for divided U.S.
"One man's modus ponens is another man's modus tollens."

Someone who was e.g. raised as a Christian, and then concluded that some of the stories they were told in Sunday School were lies, is likely to treat an appeal to religious authority as evidence against a claim. Even in domains where Christian groups have been effective in practice.

The same can be said about mainstream media or science.

In the short term, it's practically impossible to reach such people without identifying somewhere they're actually right in their rejection of the mainstream narrative, and gaining their trust by recognizing that. Only with that trust do you have a shot of being listened to when you point out where they overcorrected against the mainstream narrative.

In the longer term, it would really help if some key mainstream narratives became more accurate. Without that, "getting out of the country" may actually become the best option for an increasing number of people.

chrchang523··on Where Are All the Successful Rationalists?
Any toolbox can be misused. Anything can be made into an idol.

The fact remains that the "lesswrong rationalism" community was the first to popularize several rather sophisticated reasoning concepts. They have provided common reference points that have made it easier for me to communicate my thinking to others and vice versa, and I am far from the only writer/reader who has benefited in this way. As a consequence, I'm generally sympathetic to this group, and personally identify as a rationalist.

I agree that the heuristic "internal consistency is always worth actively working on" can go very wrong. It looks like you're defining rationalism as structuring of one's life around that heuristic; I'm guessing you must have had bad experiences with people who actually do that.

chrchang523··on Where Are All the Successful Rationalists?
It is not necessary to resolve all contradictions in one's beliefs and values to gain some benefit from rational introspection. Even professional physicists haven't arrived at a consensus on how general relativity and quantum mechanics fit together; that doesn't make the internal consistency of those two models worthless.
chrchang523··on Where Are All the Successful Rationalists?
I don't see how the disagreeing students in Varoufakis's classes are irrational. There are at least four reasonable explanations I can think of for their behavior.

1. They perceive a tail risk that at least one student is willing to spite the others. Possibly a student that has less to lose from bad grades than their classmates do, but that's not the only context where it could happen. While an assumption that everyone has the same values and will make the optimal choice under those values often leads to an effective approximation, modeling human behavior isn't always that simple... especially when only one person has to deviate for your model to make a catastrophically wrong prediction.

2. Grades aren't quite on an absolute scale; at least when pushed to extremes like "everyone in the class gets an A+", the most sophisticated readers of the transcripts will interpret that A+ differently than an A+ in the same class where the distribution of grades was more typical.

3. Getting yourself to behave in your long-term self-interest isn't always trivial. People do things like publicly commit to donating to a cause they hate if they don't achieve goal X, to try to increase their chance of reaching that goal when they've wanted to but failed in the past. Meaningful grades can be better than their absence if you struggle with self-control and trust the professor enough.

4. If you have a clear interest in seeing your classmates put effort into the course, you might vote against this.

Note that, in some of these cases, the student has an incentive to be quiet about their primary motive and instead give the cover-story explanation Varoufakis heard. I don't think those cases are likely (I think it really is mostly #1), but I wouldn't rule them out.

chrchang523··on Where Are All the Successful Rationalists?
Rationality is about internal consistency of beliefs and values.

"Rationalism is inherently flawed in that it lacks a value system" strikes me as a category error. Giving you a value system isn't its job; instead, it provides a toolbox for identifying and resolving contradictions in your preexisting beliefs and values.

(Since we all work with imperfect information about the world, internal consistency is far from a guarantee of accuracy; but it improves the odds.)

chrchang523··on Custom Allocators Demystified
I've gotten a lot of mileage out of the following two extensions to the linear allocator:

1. A "pop-to" operation. If you're ready to free ALL memory allocated at or after allocation X, you can do this by moving the start-of-free-space pointer back to the start of allocation X.

2. An end-of-free-space pointer, with its own push and pop-to operations.

In combination, these are often enough to get maximum value out of a fixed-size memory workspace. A function can allocate all returned data structures on one side of the workspace, and use the other side of the workspace for all temporary allocations. This approach does require tight coupling so there are a lot of settings where it clearly doesn't belong, but I've found it to be very effective for high-performance scientific code.

← PreviousPage 5 of 9Next →