Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
kwillets
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
32 ms
·
301.
▲
by
kwillets
9y ago
Yes, look up the MSDS on anything like this -- BLO has Cobalt-Manganese salts.
302.
▲
by
kwillets
9y ago
My kids figured out that the online target.com prices are much lower than the local store, so they still shop there but price-match everything at the register.
303.
▲
by
kwillets
9y ago
It adjusts to the price on that day far in the future when employees get out of lockup, but if it spans tax years you're still screwed. Basically, when you have gains, you only ensure realistic taxes when you exercise and sell within
304.
▲
by
kwillets
9y ago
What types of compression do you use? I did some work on streamvbyte, and it seems relevant.
305.
▲
by
kwillets
9y ago
How is overall throughput on GPU systems? It looks a lot faster than the last time I looked at GPU bus interfaces; how well does this pull in data from external sources? Viva Peru!
306.
▲
by
kwillets
9y ago
Here's the bill for the parcel in question: https://ttxonlineportal.sfgov.org/content/San-Francisco-Form... As with other public notices, the tax collector doesn't normally take "you didn't tell me
307.
▲
by
kwillets
9y ago
The tax bill is a public record, and any responsible property owner would make sure it's paid.
308.
▲
by
kwillets
9y ago
His regex golf seems kind of strange; he never mentions suffix trees. It seems like a fairly straightforward process to go through and mark the shortest distinguishing infix for each string in a generalized suffix tree.
309.
▲
by
kwillets
9y ago
Side-to-side. There's a certain design range where it's not too fast and not too slow, and some inspectors check for it. https://en.wikipedia.org/wiki/Metacentric_height#GM_and_roll... Also, there's ran
310.
▲
by
kwillets
9y ago
The Saewol used a tactic of overloading with cargo and compensating by emptying ballast tanks so that the hull depth would be the same for inspection. There's already a need to overcome malicious actors.
311.
▲
by
kwillets
9y ago
This is essentially what capsized the Saewol and killed hundreds of people. It can be detected with empirical tests however. Almost every ship should right itself from a roll within ten seconds or so, so anyone with a stopwatch can detect
312.
▲
by
kwillets
9y ago
Just traverse the tree forwards and backwards simultaneously and match the moves. I hate peevish interviewers. If you want to show off, write a paper. bool is_mirror(left_tree,right_tree) { if( !left_tree && !right_tree)
313.
▲
by
kwillets
9y ago
There have been a few local inventory search startups, but there are a few gotchas: An in-stock item may be bought by another party before order or pickup. Inventory may show more inventory than actual. Shoplifting is one reason for this.
314.
▲
by
kwillets
9y ago
So do the engineers work on finishing the building too?
315.
▲
by
kwillets
9y ago
To be fair, neither does Vertica. The big data warehouses are expecting fairly flat predigested fact tables.
316.
▲
by
kwillets
9y ago
One of the problems with SIMD is that we can't do this kind of shuffle at bit granularity, at least not in one instruction.
317.
▲
by
kwillets
9y ago
This is indeed fast and simple, since the pshufb instruction allows unpacking of bytes into multiple zero-padded words in one step. This technique uses a precomputed table of shuffle patterns, so it's a short sequence of instructions
318.
▲
by
kwillets
9y ago
Thanks for that link; this is the kind of re-engineering that I'm curious about. While we're at it here's a proposal for normalized keys in pg: https://wiki.postgresql.org/wiki/Key_normalization .
319.
▲
by
kwillets
9y ago
Technology advances, but human nature does not.
320.
▲
by
kwillets
9y ago
Postgres is going through some of these issues as well with ICU support. They don't necessarily have room for normalized strings in indexes, conversion is slow, and hashing also has trouble deciding if strings are identical. I've
321.
▲
by
kwillets
9y ago
His work on Base64 decoding might be a start. In many cases it's a matter of how small an FSM or transducer is. For multiple string match I was looking at doing Aho-Corasick for long strings in a SIMD-friendly way, but it's more
322.
▲
by
kwillets
9y ago
There's a pretty good discussion and benchmarks here: http://0x80.pl/articles/simd-strfind.html and here: http://0x80.pl/articles/simd-friendly-karp-rabin.html
323.
▲
by
kwillets
9y ago
It's 16/3 bytes per cycle. The newer SIMD stuff seems to be factoring it out into bytewise compare and CLZ on a mask.
324.
▲
by
kwillets
9y ago
PCMPISTRI, what an instruction. It's supposed to be a comparison, but it doesn't set the flags with comparison results. Instead it uses them to signal if it's found an unequal byte, or hit the end of the strings. The latter
325.
▲
by
kwillets
9y ago
That's fine until you get numbers with decimal points.
326.
▲
by
kwillets
9y ago
Another name for sign + short representation of magnitude + mantissa is Floating Point, which also sorts lexicographically for positive values.
327.
▲
by
kwillets
9y ago
A conventional sorting algorithm such as Multikey Quicksort. http://people.mpi-inf.mpg.de/~sanders/courses/algdat03/salgd...
328.
▲
by
kwillets
9y ago
The bound on string sorting is typically written as O(n log n + D), where D is the sum of distinguishing prefixes, ie input length minus some fluff. Since D >= n log n we already have linearity on input length.
329.
▲
by
kwillets
9y ago
You are correct. Apologies for misremembering. Looking it up shows (d+1)^n points are needed for multivariate.
330.
▲
by
kwillets
9y ago
True, I ignored division. I believe this is used in symbolic solvers to prove identities; that's the context where I learned it.
More ›