Datadog | Software Engineers | ONSITE (Boston, Denver, Lisbon, Madrid, NYC, Paris, Tel Aviv) and REMOTE | Full-time
Datadog is a monitoring, tracing, logs system, and more, for your infrastructure and services. We build our own tsdb, event store [1], distributed tracing tools, cutting edge visualizations, and more. We love shipping great experiences for customers just like us and are growing fast! We write a lot of Go, Java, Python, Typescript (with React), and a bit of other languages. We run on k8s, and are multi-region and multi-cloud.
We're looking for people who can build systems at scale as we process trillions of events per day. Let us know if that's you!
I feel the need to defend Finnish cases. Yes, they have a gazillion, but unlike other languages you might know (eg Latin or German), there’s nothing difficult about Finnish cases.
Most cases are simply used where English would use prepositions. In Finnish those are postfixes instead, a bunch of letters tacked onto the end of a word. It’s a case and not a word because there’s no space between the two, that’s it.
Eg “talo” means house. Talossa means “in the house”. Talon means “of the house” (actually, “the house’s” - omg English has cases too, super difficult). Talolla means “on (top of) the house”. That’s not harder than prepositions is it?
I did cheat a bit, because with some words you first got to find the root before you can tack on “-ssa”. The root of talo is also talo, but for some words you got to apply a (simple, purely letter-based) rule. Eg the root of “ankka” (duck) is “anka” so “the duck’s house” becomes “ankan talo”. There’s a bunch of rules to find the root of a noun and you can learn them in half an hour or so.
There’s plenty stuff that’s harder about Finnish (notably the vocabulary), but the cases are peanuts.
I read Finite & Infinite Games by James P. Carse after coming across it in a comment on HN. It's one of those books that changes how you view interactions with others, and puts things into perspective.
Man's Search for Meaning by Viktor Frankl is a close second. I read this when I was going through a particularly difficult time in life, and it has stuck with me since.
David Deutsch's The Beginning of Infinity definitely changed how I look at skepticism and pessimism in general (a weird take-away, I agree).
There's only one book that I often find myself going back to, which is Hermann Hesse's Siddhartha. It's a very profound work IMO, and provides a sense of grounding to me.
The implementation of the __cos kernel in Musl is actually quite elegant. After reducing the input to the range [-pi/4, pi/4], it just applies the best degree-14 polynomial for approximating the cosine on this interval. It turns out that this suffices for having an error that is less than the machine precision. The coefficients of this polynomial can be computed with the Remez algorithm, but even truncating the Chebyshev expansion is going to yield much better results than any of the methods proposed by the author.
[1] has a potential use case. Resistor networks are handy when modeling divergence-free flows through arbitrary networks with linear "costs" from a predefined origin to a predefined destination.
Kirchoff's law and Ohm's law makes things linear, and turning the optimization problem into a linear problem is very advantageous for efficiency.
Diodes are asymmetric resistors with R=0 for one direction and R=inf for the other, so represent one-way streets in the network. A voltage source would be how you define the origin location (destination location is set at ground) for recognizing different routes for flows to take. Inductors are out of scope I think since this should find the equilibrium for voltages at each node and currents across each resistor and inductors' dynamics are only interesting over time.
Spotify knows enough people won't churn to make a huge difference (and honestly, the people who sign up because of this might even balance it out). However, what they can't lose is employees. If they become a place nobody wants to work at, they're actually screwed.
It's weird to see how inevitable the ending feels (to me, as an outsider). I'm 100% confident Joe Rogan will be removed from their platform; it's just a matter of how long it takes for Spotify to realize it's past the point of no return.
Tangentially, the physical labor jobs I had when younger never really tired me out the way that the mental/bureaucratic/political job I have now does.
Physical work actually energized me, I mean this is all confirmed today as well, we know all the benefits exercise brings on well being.
There's something about desk jobs that is frankly soul sucking and literally mentally draining in a way where when the day ends, it's as if you suffer from temporary depression. Even getting motivated to do things you want to do is hard, resorting to the laziest activity is often what happens, phone, social media, television. Sometimes I can't even get myself to play a video game and I love video games.
And when the night comes, you'd think sleep is what you need, but that same day of desk job actually gives you insomnia, falling asleep is hard, and while you sleep it's as if all of that mental activity is still happening in your head from the work day.
If physical labor work paid me as well and provided the same benefits, I'd probably switch back to it honestly.
Yes, it's extremely powerful once you truly understand Nix.
There are efforts to improve documentation, but it still is lacking (I think the biggest problem is that Nix is so big, not just the OS but it can be utilized as a build system).
Just with NixOS is not exactly clear how can you for example build your custom image.
I think https://nix.dev/ is approaching the documentation from the right direction.
There are also many pieces that people built that you need to find.
For example some things that I found accidentally:
I came across this site a few weeks ago. And it's quite unlike anything I've seen. And try to inspect the site using browser inspector. It's a very different ui than I've ever seen before.
Makie.jl [1] does it's plotting on the GPU (like a video game engine), so can handle millions of datapoints just fine.
Also, note that for people to whom plotting is really important, it's quite easy nowadays to just AOT compile your plotting library to your sysimage with PackageCompiler.jl [2] for instant plots.
I am just guessing here, but in case the author had their service compromised, maybe he can't disclose the information. Feels like they know what they are doing, and at least to me, reading between the lines, it looks like they fixed their problem and they advice people to fix it too:
> If your site has actually been hacked, fix the issue (i.e. delete offending content or hacked pages) and then request a security review.
I was looking for Twitter alternatives and found this one, which is old and stale but looks like a very interesting experiment! Quoting its frontpage:
But what's wrong with Twitter?
I'M GLAD YOU ASKED. There are two aspects of Twitter that just bug me as an engineer:
Ruby on Rails - Using rails to prototype a system is fine — scaling up to a million hits a day with it is just a bad idea. As the service grew, I'm sure it cost them a lot more time than it saved.
140 characters is not enough - I routinely write sentences longer than 140 characters, so I can't even begin to imagine making a point in such a small space. This textual confinement has led to the rise of URL shorteners, which are breaking the internet.
Blërg solves these problems by applying absurd reactionary engineering. Blërg's database backend is a custom C program that handles requests over HTTP and stores data in a very small and efficient indexed log-structured database. The frontend is done entirely in client-side Javascript. A single post can be up to 65535 bytes in length.
Which is not to say that I believe writing your service in C is the solution to all your problems. Clearly, this approach has just as many hairy problems that will bite you in the ass sooner or later. The best way, as with most things, lies somewhere in the middle of high-level abstraction and ZOMGHARDCORE OPTIMIZATION.
Or more politely described in the documentation page:
Blërg is a minimalistic tagged text document database engine that also pretends to be a microblogging system. It is designed to efficiently store small (< 64K) pieces of text in a way that they can be quickly retrieved by record number or by querying for tags embedded in the text. Its native interface is HTTP — Blërg comes as either a standalone HTTP server, or a CGI. Blërg is written in pure C.
I enjoy how everyone discovers this "feature." It really is a lot of fun to play with.
True story, back in the cave person days, when television only came over the air waves and if you wanted to "own" a computer you shelled out anywhere from $500 to $5000 for something with an 8 bit processor that ran between 1 and 4 MHz and had, usually, much less than 65,536 bytes of RAM, peripherals for these things were few and far between. A couple of serial ports and a couple of parallel ports was considered a "lot" of I/O. You might have stored your programs on an 8" or 5" floppy disk, but if it was before 1978 you most likely stored programs on cassette tape. Anyway ...
One of the things you could always count on was that these things radiated all sorts of spurious energy and being relatively low frequency beasts to begin with, finding a harmonic that you could pick up with an AM radio held "near" (where near could be 1' to 10' (.3 - 3m)) that interference and with the judicious use of delay loops cause that interference to modulate to "tones" and thus create recognizable "music."
As a kid in high school with just such a computer I can tell you that "playing music" (no matter how poorly) on a nearby AM radio was orders of magnitude more impressive than pointing out by the pattern of lit and unlit LEDs that your amazing machine had just calculated a logarithm. (even though the latter was significantly more difficult from a programming perspective!) Once I got a printer, being able to print out NSFW pictures was a neat trick but for crowd appeal and cheap accessories, music over an AM radio was still the top demo of the time :-)
