Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
dspeyer
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
44 ms
·
421.
▲
by
dspeyer
17y ago
Google recruiting is notoriously slow. A few weeks to first acknowledgment is not unusual. Also, if you've just submitted your resume, there are multiple stages of interviews between you and an offer.
422.
▲
by
dspeyer
17y ago
More like the day that Google admits defeat and begins preparing for its own death. Not that I expect that day to come anytime soon. Alternatively, it'll be the day all Googlers depart their biological bodies once and for all and live as p
423.
▲
by
dspeyer
17y ago
About a year ago the New York microkitchens were shrunk. A lot of variety was removed, though they remained well-stocked and high-quality. Since then they've bounced back a bit. AFAIK, no perks have been cut recently. I had to triple-che
424.
▲
by
dspeyer
17y ago
Absolutely. Current latency from New York to Tai Pei is about 210ms. With neutrinos it would be under 80ms. And immune to earthquakes and undersea mudslides. It would also be a lot easier to build emitters and detectors on land than in s
425.
▲
by
dspeyer
17y ago
It's worth noting that all the pro-retail-clinic claims in this article come from a study done by the RAND Corporation. This is not an entity known for single-minded pursuit of truth. It is closely (though informally) tied to the republic
426.
▲
by
dspeyer
17y ago
Google Books cut off before he got to explaining what the fraud was. Does anyone know? He seemed to be complaining about assuming distributions were normal without checking, a simple mistakes that is warned against in any introductory stat
427.
▲
by
dspeyer
17y ago
Sounds like you've got interest.... You probably ought to think a little about where you land after this if it doesn't become profitable, or decide that when your bank account drops to something ($200?) it's time to stop. On the other hand
428.
▲
by
dspeyer
17y ago
A defining characteristic of academia is publication. Everything an academic discovers should be revealed to the public as soon as possible, and genuine questions should be answered thoroughly. This ideal isn't 100% lived up to, but it is
429.
▲
by
dspeyer
17y ago
I believe the original problem definition called for the data to be randomly spread over 1000 computers. Bringing together 32GB of data from 1000 nodes is going to stress your network, and pre-emptively sharding would be worse. I think the
430.
▲
by
dspeyer
17y ago
Wait, he's saying that recognizing a proof in polynomial time makes finding a proof easy? The search space is exponential, and there's no adequate heuristic for a best-first search.
431.
▲
by
dspeyer
17y ago
Oversubscription based on expected use is no excuse for routinely not providing what was promised. It's the ISP's job to measure actual patterns and make sure enough capacity is there. How would these ISPs feel if their banks told them "s
432.
▲
by
dspeyer
17y ago
Nope, Google saw this coming first and stopped them. Remember the 700MHz D-Block?
433.
▲
by
dspeyer
17y ago
> Passing MyArray[x+3] as a macro argument would lexically copy this expression in multiple locations in the macro expansion, causing the generated code to needlessly evaluate *(Myarray+x+3) again and again and again. Wouldn't any decen
434.
▲
by
dspeyer
17y ago
There was a big media frenzy back when Google first announced its edge-caches physically located in ISP server rooms: http://techliberation.com/2008/12/16/edge-caching-vs-prefere... http://www.dslreports.com/shownews/The-Wall-Street-Jour
435.
▲
by
dspeyer
18y ago
Can anyone point out the "report as inappropriate" he claims to have used? I can't find it on either current Amazon pages or Google Cached ones. If it indeed points to an unsecured GET statement, that supports the claim. If it doesn't, th
436.
▲
by
dspeyer
18y ago
From my experience, Cornell Alumni have a much greater presence in high-end cs industry than any of those other schools. It's probably worth trying to squeeze more money out of them (what do you have to lose by trying?). Tufts has a genera
437.
▲
by
dspeyer
18y ago
University of Washington has the best overall CS department of those three by a fair margin. At least, that was the general impression of grad schools at UMaryland undergrad, and UWashington alumni are noticeable at both Microsoft and Googl