whorleater

unserious article written by inane consultancies feeding at the trough of linkedin. The diagram itself is hideous.

Software consultant, solution architect, solutions eng, all do what an "FDE" does without having the weirdly cringe palantir branding.

Honestly I'm kinda down on most of FB's tech, but them pioneering Scuba was pretty sick. It has all the wide logging/apm/tracing/etc, and it's hard to describe how useful it is if you haven't seen it in action.

A fork of a fork I guess, Organic Maps was originally `maps.me`, and I suppose we're forking it again

One thing I haven't seen being discussed is the BEAM internals becoming a little long in the tooth. We still have static reductions before the scheduler switches to another task, the priority system in scheduling is a bit dodgy, flipping vmargs is kinda complex, lock counting and crash dump tooling kinda suck, etc.

BEAM is great, although it's definitely missing something like pprof for go or java flight recorder.

Yeah, I don't know why this falsehood continues to persist. WhatsApp and Ericsson engineers continue to work together to evolve Erlang, alongside a bunch of other people across the industry.

Source: I work at WhatsApp

WhatsApp very long ago used to hot reload across all nodes with a ssh script to incrementally deploy during the day

Huh, kinda funny, I feel the exact same way. Few people IRL know about my website, 2-3 people occasionally write in a year about something, but weirdly it feels like the idea of a personal, non commoditized internet space has become so rare it's seen as odd.

I have consulted for UNDP and DPPA during my time in Iraq, although that left me with complicated feelings about that industry.

I would be curious why kube cron jobs didn't seem to fit the bill, my favorite part of these posts are when they have a section hinting that they explored other options picked specific tradeoffs

Physical notebook and data entry later.

Why is this news? Iraq does this, Jordan does this, it's common throughout the region to block internet. It's not effective, and usually there are carveouts (office internet still works, usually it's only cellular and public internet, etc). When I was living in Iraq this was just something you dealt with, like a rolling blackout once a while.

If we buy into the fact that we can account for "culture" onto this distribution, this article assumes that the distribution has remained the safe size and shape, and so he draws the conclusion that the long tail is declining or disappearing.

But the nature of the long tail is that it is comprised of a series of outliers. To use the author's analogy, what might be happening is that there's a distributional shift occurring at the same time the general amount of "stuff" under the curve is increasing, which gives the appearance that things have become more alike, even though what has actually happened is that the bump has just gotten bigger.