back
user profile

kristjansson

4,853karma·1,566submissions·February 17, 2015
recent activity (1,566 total)
comment
Interestingly, looks like that's presented as a checkbox in Monterey, rather than an option-click (never Apple's most discoverable UX).
4y ago·view thread
comment
> In 2003, the entire editorial board of the Journal of Algorithms resigned ... at the suggestion of Journal of Algorithms founder Donald Knuth That's a great detail for that story.
4y ago·view thread
comment
This has actually succeeded before though. Journal of Machine Learning Research (JMLR) is one of the premier journals in its field, is open access, and was formed by the mass resignation of editorial…
4y ago·view thread
comment
There's also the alignment of interests between you and the site operators, even for a small search engine. Cooperation improves outcomes for both sides. Whereas the more aggressive scrapers al…
4y ago·view thread
comment
The argument against p-values is part of the argument against any bright-line single-number rule for identifying truth. The job of the researcher is to demonstrate (at least) that 1. there is (isn…
4y ago·view thread
comment
It's worth pulling the principles from the ASA's statement [2] as well: 1. P-values can indicate how incompatible the data are with a specified statistical model. 2. P-values do not…
4y ago·view thread
comment
Like all controlled experiments though, the experimenter wants to hide that information from the subject (user in this instance) to measure how they respond to the change itself, rather than the chang…
4y ago·view thread
comment
Good to have everything under one roof in Santa Barbara, a city where you famously can't go outside 9 months of the year
4y ago·view thread
comment
Lots of confusion derives from the best-of-breed parquet readers for Python and R residing in the Arrow packages, mostly because Arrow is (and does) a lot of things. There's: * arrow, a in-memory…
4y ago·view thread
comment
The memory profiling only measures allocations on the R heap (i.e. objects managed by R's GC). It doesn't measure allocations by other libraries running in the the same process, unless they…
4y ago·view thread
comment
Of course, since the memory allocation comes from Rprofmem (via benchmark::mark), this only measures allocations of memory for objects on R's heap. Allocations by C extensions (like DuckDB and S…
4y ago·view thread
comment
People will act to maximize their short term benefits vs costs. They’ll buy American made, Union made, imported, whatever is cheapest. Of course at the margins some people will choose (place value o…
4y ago·view thread
comment
NB updated as of two days ago, despite the 1809 title.
4y ago·view thread
comment
And at least an order of magnitude more expensive than an A100, if not two orders
4y ago·view thread
comment
AV/Photo import?
4y ago·view thread
comment
Isn't that what one already does with a successful referral bonus, buy beer for all parties involved?
4y ago·view thread
comment
There are referrals and there are referrals. One is basically a a lead generation product for internal recruiters - internal referral bonuses at the like. The other is a professional and social relat…
4y ago·view thread
comment
Briefly (like, minutes to hours) and for very specific structural reasons. Negative oil prices were never an accurate reflection of the actual supply/demand equilibria for oil.
4y ago·view thread
comment
That was underspecified, apologies. You own the vault. Everyone has a vault. The vault is your key. Would've been clearer to say 'pocket' or something, but that doesn't imply the …
4y ago·view thread
comment
Bitcoin is trustworthy in the way that gold or beads or shells in a vault are trustworthy - if you have it, you have it, and if you give it to some else, they have it and you don't[1]. That doesn…
4y ago·view thread
comment
_Any_ database can be foundational, digital money if you get the enough people and their governments to agree that it is. The database part is not the hard part.
4y ago·view thread
comment
> Natural immunity is strong, but for most cohorts is more likely to have side effects That’s a very generous and evenhanded way to say that
4y ago·view thread
comment
BTC/USDT (on Binance) goes moonward as the denominator collapses, BTC/USD goes doomward as the numerator falls.
4y ago·view thread
comment
Bitcoin and its ilk are explicitly pseudonymous not anonymous. Others attempt to provide actual anonymity, but lack the usage and mindshare of bitcoin
4y ago·view thread
comment
Yeah, knocking out their backup IRC (or maybe it was the authentication to the backup IRC?) along with everything else was the key.
4y ago·view thread
comment
Yup, it's not really about remote hands in datacenters (which of course have been staffed continuously), it's about losing 'run over to their desk' and 'shout across the offic…
4y ago·view thread
comment
The causal relation to WFH is ... tenuous ... but now can see how losing all internal comms to an incident could make managers more nervous about WFH. Of course announcing a policy change now is closi…
4y ago·view thread
comment
There’s an infinite number of each, so we’re already stretching colloquial definitions by comparing them
4y ago·view thread
comment
It’s interesting to read others reasoning on this. I find it hard to follow and much harder to generalize without some notation (and a diagram which this comment box struggles to reproduce)
4y ago·view thread
comment
I don’t think that’s right. It if the radius of the inner circle is one third the outer, it only rotates twice rolling around the inside, which makes sense as it’s center traces a circle that’s the t…
4y ago·view thread