vladf

10 ms·

61.
by vladf 5y ago
Funny, I was just complaining about the exact same phenomenon (complaining with a fix!) a few days ago: https://twitter.com/FeinbergVlad/status/1475148590423023619
62.
by vladf 5y ago
What about adding a captcha?
63.
by vladf 5y ago
Numba is great! Whenever you’re using CPU and have very simple parallelism patterns, it’s your best bet for speeding up numpy. But if you needed to do this on a GPU or TPU, ideally with native and transparent SIMT, such as the case for the
64.
Vectorizing Ragged Arrays(vladfeinberg.com)
28 points by vladf 5y ago|3 comments
65.
by vladf 5y ago
How would such a thing like this exist in practice? Who would watch this watchman? When your goal is to "discover truth" as Chief Truth Discoverer of the DOJ the only way to audit you is with grander truth discoery.
66.
by vladf 5y ago
Surely the hiring issue is a market inefficiency, then, no? There's got to be plenty of engineers out there, but maybe not for typical startup cash/equity structures.
67.
by vladf 5y ago
> Sure, you could crank up the difficulty until it makes the cost of the attack too high, but then all your users will get mad - that's basically what CloudFlare does and it attracts non-stop criticism. This is interesting, where ca
68.
by vladf 5y ago
On your somewhat off-topic meta-comment about my lack of due diligence: lucky for me I don't need to break out Coq to validate an RFC draft to post a comment on HN, which despite my mushy-gushy feelings has resulted in a productive, cu
69.
by vladf 5y ago
You do bring a good point. Maybe outlandish, but see my sibling comment ( https://news.ycombinator.com/item?id=28578832 ) where the server overhead for rejecting invalid packets which the attacker made for free goes down to p
70.
by vladf 5y ago
Hrm, I don't think you're engaging with the spirit of my proposal, which does require resource-symmetry to establish auth in the protocol. I admit I haven't specified a full counterfactual protocol here, but see my edit for
71.
by vladf 5y ago
Right, under the previously unspecified constraint that you need a response in one round trip I agree there's a natural resource asymmetry.
72.
by vladf 5y ago
I guess the problem is "simply" the use of these resource-asymmetric protocols. As a server-owner, if there was a robust alternative protocol which ensured resource symmetry, wouldn't you be naturally incentivized to adopt it
73.
by vladf 5y ago
Why is this resource asymmetry unavoidable? It seems reasonable that a new protocol could solve it. Consider augmenting TCP with a "pre-SYN" and "pre-SYNACK". Suppose you have to send a pre-SYN which declares your identi
74.
by vladf 5y ago
> Hackable is the default for any test imposed by an authority. The reason the tests you're given are so consistently bad — so consistently far from measuring what they're supposed to measure — is simply that the people creatin
75.
by vladf 5y ago
Hrm, is it really that easy? If you only have a placeholder in one of your arrays that points to the moved object in the other, then you need to do the actual move back if you have a sequence of pops/pushes which force you to delloc th
76.
by vladf 5y ago
Ah, I was in a rust/c++ frame of mind here—it seems you deeply rely on a copy operation being available, but unless you use indirection and weak pointers it might not be. I think with copy available this certainly works
77.
by vladf 5y ago
I challenge you to try to implement it this way and test that (1) an arbitrary sequence of push/pop is valid and (2) doesn’t use more than linear space.
78.
by vladf 5y ago
One thing I’ve always been curious about, can you guarantee worst case linear memory use with this setup while supporting pop() at the same time?
79.
by vladf 5y ago
Often, especially when analyzing production logs, you might be interested in questions like "how many users were active in the last hour?" or "how many active connections in the past hour for each of my servers during high lo
80.
sort -u | wc -l but faster, with less memory, and slightly wrong(github.com)
9 points by vladf 5y ago|1 comments
81.
by vladf 5y ago
Why would a publisher narrow its target purchasing demographic preemptively? At the end of the day, the point is book sales, not to actually make you successful (even if that’s the effect for some small slice of the people who actually buy
82.
by vladf 5y ago
I do quite a bit of data munging on multiprocessor Unix systems, and found (some remarkably lightweight) homemade Rust/py tools invaluable for getting stuff done. A low-memory Unix map-reduce combinator - https://github.com&
83.
by vladf 5y ago
Perhaps my info on these are out of date. Why would you say those are successful examples?
84.
by vladf 5y ago
Is this really true? To be able to build a track you need to have government-level authority (e.g., in GP's example, to build something on its own lane in the road, or to displace people in homes in the way of the track). However, the
85.
by vladf 5y ago
I solved a similar problem recently: given a stream of data, how should you choose packet size in an online way to minimize regret (a linear combination of spare capacity of last packet and total packets used). Turns out doubling isn’t the
86.
by vladf 5y ago
Tutorial and git repo: https://github.com/vlad17/slb
87.
Beating Spark with a Better Map Reduce(vladfeinberg.com)
24 points by vladf 5y ago|1 comments
88.
by vladf 6y ago
I think the problem with observing interventions is it will result in impractically large sample complexity to derive the same causal conclusion as causal calculus armed with assumptions of DAG structure. I think metaphysically speaking the
89.
by vladf 6y ago
I think parent’s construction works just fine if you disallow removes, so it grows monotonically, regardless of how dramatic the increase in keys, so long as you move over 2 keys for every update to your map. My comment was a description of
90.
by vladf 6y ago
What happens if you have a sequence of puts that triggers an expansion and some keys get migrated, but all of a sudden you get a series of removes, such that you need to perform a shrink to keep memory usage linear? Do you now end up with t

More ›