Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
wendyshu
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
25 ms
·
271.
▲
by
wendyshu
7y ago
Yeah I don't know. I guess I'd just prefer to give to the worthy causes out there that don't spend as much of the donations on the leader's salary. Also, there's an abundance of free educational resources on the web
272.
▲
by
wendyshu
7y ago
If he thinks other people should donate a whole bunch of money, shouldn't he donate a whole bunch of money if he has it? Maybe he already does this, I don't know. If he took a lower salary he would effectively be doing this -- in
273.
▲
by
wendyshu
7y ago
I'm not asking for donations.
274.
▲
by
wendyshu
7y ago
Salman Khan takes $800k/yr from Khan Academy donations for himself, maybe he should donate.
275.
▲
by
wendyshu
7y ago
Why not test on a set of problems that come up in practice rather than generated by an artificial distribution?
276.
▲
by
wendyshu
7y ago
Social anxiety? They'd make communication more difficult not less.
277.
▲
by
wendyshu
7y ago
https://predictionbook.com/predictions/198425
278.
▲
by
wendyshu
7y ago
Not quite sure why everyone needs to learn so much math.
279.
▲
by
wendyshu
7y ago
I don't think it's because of handwriting. Math notation (for math) is just as useful when reading and writing it on computers or printed pages. I think it's just optimized for its subject matter, e.g. a small set of names an
280.
▲
by
wendyshu
7y ago
Sure ∑ would pass code review. And in Maple and Mathematica it is code.
281.
▲
by
wendyshu
7y ago
Yeah I guess cards can be slower than cash
282.
▲
by
wendyshu
7y ago
Did they not have the equipment?
283.
▲
by
wendyshu
7y ago
Why are cards worse?
284.
▲
by
wendyshu
7y ago
There's https://www.metaculus.com/questions/3530/how-many-people-wil...
285.
▲
by
wendyshu
7y ago
Generally no amount of new information will change someone's mind about a date but in hiring that may not be true.
286.
▲
by
wendyshu
7y ago
Sounds like managers are risk averse because they'll get the blame if something goes wrong but not the credit if it goes right. But I don't think the risk has to be so big. By the end of an interview process, candidates I talk to
287.
▲
by
wendyshu
7y ago
Nothing wrong with being transparent. And maybe they really can do the thing you're assuming they can't.
288.
▲
by
wendyshu
7y ago
Yup
289.
▲
by
wendyshu
7y ago
TAoCP is a terrible resource for a beginner.
290.
▲
by
wendyshu
7y ago
For data structures and algorithms, work through Introduction to Algorithms by CLRS and follow the video series from IIT Bombay, available here: https://www.youtube.com/playlist?list=PL7DC83C6B3312DF1E . ...or use any othe
291.
▲
by
wendyshu
7y ago
Al who?
292.
▲
by
wendyshu
7y ago
Maybe they should make their products as noteworthy as their hiring practices.
293.
▲
by
wendyshu
7y ago
Huh?
294.
▲
by
wendyshu
7y ago
Summary: * contrition * the moderator dismissal was too hasty * they will create a process for reinstating moderators * new Code of Conduct announced Oct 11
295.
▲
by
wendyshu
7y ago
I suspect they've essentially decided on the new Code of Conduct and any solicitation of feedback will be perfunctory.
296.
▲
by
wendyshu
7y ago
The quote: https://i.imgur.com/OA5K2Ty.png
297.
▲
by
wendyshu
7y ago
Stack Overflow hired a highly political, sanctimonious Community Manager and this is the result. E.g. he suggested users who didn't like the rules require therapy https://meta.stackoverflow.com/questions/389935
298.
▲
by
wendyshu
7y ago
In terms familiar to users of the religion stack exchanges: zealotry -> schism -> excommunication.
299.
▲
by
wendyshu
7y ago
I don't think it's unreasonable to say it has a anti-Beijing bias.
300.
▲
by
wendyshu
7y ago
Oh for Pete's sake, what next.
More ›