oivey

37 ms·

481.
by oivey 6y ago
I’ve generally found they can be interchanged in the way you would expect as a Python user: duck typing. I have more experience using SubArrays, but I think the underlying machinery in the language is basically the same. Adding performance
482.
by oivey 6y ago
I’m a huge Julia proponent (and before, a huge Python proponent over Matlab), but I would be careful about claiming that Julia will be the coming standard. Python has a crazy amount inertia and excellent projects still in the pipeline. I th
483.
by oivey 6y ago
Yes, you’re right. At least partially recipes call for items in state because prep time will vary significantly by skill level. More cynically, with blogger recipes there is also a potential for SEO games. It isn’t really a shady practice,
484.
by oivey 6y ago
This used to be very discouraging for me, but I recently learned that usually the time given for recipes doesn’t include chopping, measuring, etc. Because how much time that takes people is so variable, recipe authors leave it out. If it in
485.
by oivey 6y ago
I haven’t read through it, but the essence I believe is an emphasis on fitting components and techniques together rather than slavishly following recipes. Similarly good (and maybe a good companion) is J. Kenji Lopez-Alt’s YouTube channel.
486.
by oivey 6y ago
We’re probably not that far apart. I do think in the future that more people will be concerned at least with basic low level optimizations because some “next gen” languages make it so much easier. Part of what makes it hard now is that C la
487.
by oivey 6y ago
In hindsight it makes a lot of sense. The cost of a transaction for Bitcoin is always going to be higher than just updating a value in Visa’s database. Visa may have some increased costs due to consumer protections, but consumers are pretty
488.
by oivey 6y ago
Low level programming isn’t like some disused area of mathematics that was popular in the early 20th century; it underpins everything we do. Your OS, your network stack, your graphical desktop, your browser, probably your favorite language’
489.
by oivey 6y ago
It’s pretty useful to know some low level computer engineering if you’re writing high performance software. Having some understanding as to how memory, cache, and threads work for your target platform is pretty important in a lot of applica
490.
by oivey 6y ago
If something like that becomes mainstream, data scientists who just glue together canned ML algorithms from libraries should also be concerned. The kind of automation that enables eliding finding and cleaning the right data probably can als
491.
by oivey 6y ago
Sure, but depending on what you’re doing, “quick” might be years. You can get a PhD in understanding the theory, a PhD in designing fast numerical algorithms, or spend many years becoming a strong software engineer. I think the willingness
492.
by oivey 6y ago
Nothing precludes you from doing integer arithmetic in a dynamic language. I’m not a quant and this isn’t my area of expertise, but, for example, I’m pretty sure various differential equation solving methods depend on variables taking on co
493.
by oivey 6y ago
This is dogmatism swung too far in the other direction, IMO. There are many, many successful production code bases written in dynamic languages. In my own experience as a vision scientist/engineer, there is tremendous value in being ab
494.
by oivey 6y ago
Yep. The key is really software skills. If you’re unable to even filter the data yourself, you’re also probably unlikely to be able do implement novel analysis techniques, especially if the analysis algorithm has many complicated steps or i
495.
by oivey 6y ago
I think this insight exists across a lot of fields. Basically, if you want to be a really excellent magician you also better be a decent blacksmith. More concretely in this case, if you’re unable to do the data “engineering” yourself then
496.
by oivey 6y ago
I wouldn’t call that perfect because if your private key is ever compromised, all of your communication ever is compromised. Ideally data would be destroyed after a while, which isn’t really possible to do if everyone ever has the data. Bro
497.
by oivey 6y ago
> - Removing it in any capacity should be done with extreme care and free of any bias one way or another, with a lot of thought put into any future consequences I agree with basically your whole post, but I think it should be emphasized
498.
by oivey 6y ago
The percentage of people who became rioters is irrelevant. Trump talked about marching to the Capitol building and “showing strength” at the rally. His employee, Giuliani, was making comments about hand to hand combat at the rally. Before t
499.
by oivey 6y ago
Teams in particular recent had a massive security breach where the client could be drive through exploited to own whole orgs and any org connected via guests, right?
500.
by oivey 6y ago
One of Trump’s children perjured themselves to lie about the purposes of a meeting with a Russian lawyer (later proven to be for Clinton dirt), his campaign manager was convicted for laundering money with Russia connected people, an advisor
501.
by oivey 6y ago
That’s a pretty false equivalence. To cite the most recently example, Fox repeatedly propagating Trump’s lies about the election has no analog on the left.
502.
by oivey 6y ago
That doesn’t work well in practice. The US government pays insane sums for those things. The best direct comparison is roads and other transportation infrastructure, where the US’s costs per unit are way, way higher than the rest of the Wes
503.
by oivey 6y ago
There's definitely a focus on cost. It is just misguided. Short of not having security clearances, I'm not sure how you can avoid paying people to wait for them. Simple projects, like properly paying people unemployment benefits,
504.
by oivey 6y ago
Not everyone is driven by money, but I think its fair to say that most people would take a job in industry where you can make 2-3x over a government job. I think this is born out in your own admission that it is very rare for government to
505.
by oivey 6y ago
Just to be clear, GS-15 is the absolute highest pay grade for regular federal employees. A GS-13 is supposed to require at least a Masters and a GS-14 a PhD. If you got 13 after a year, you have very little potential growth for the rest of
506.
by oivey 6y ago
GS-13 is a significant amount of experience and 107k isn’t a competitive wage for an engineer of that experience level, especially in tech. How can the government provide adequate technical oversight of its contractors if it doesn’t hire en
507.
by oivey 6y ago
Most tech employees in even in the valley aren’t founders or early employees risking it all. There is probably some element of drive for commercial products or something but I think you’re vastly overestimating the degree to which that is a
508.
by oivey 6y ago
Government is competing with FAANG for employees. There isn’t a good reason to exclude them from comparison. Even so, working for a startup without equity is often more lucrative than working for the federal government. The situation is eve
509.
by oivey 6y ago
I think the public focus on cost has paradoxically led to this situation. If you’re a good engineer, why work for the government? The compensation is extremely poor compared to industry, especially in tech. The lazy/incompetent people
510.
by oivey 6y ago
It’s unclear how this is true. You would still need tools to search the blockchain for the content you’re looking for. Someone would also need to curate and coordinate the format for embedding this information. Anyone involved with either o

More ›