sayrer

It sounds like you built the right product on the wrong technology stack, at the lower levels. For example, I have never heard a Nomad success story, but this might be colored by interviewing engineers desperate to escape it.

Something like linkerd on Kubernetes would be stronger, I suspect. But I don't know the exact nature of your problems.

I can recommend the book “American Girls” by Nancy Jo Sales.

I don’t have any daughters, but I’ve worked at some big social media companies, and this was required reading at one of them. The boss there did have daughters, but it wasn’t only so we would understand. It was so we wouldn’t build Trust and Safety headaches in the first place.

No, they’re trying to get to Singapore. Business in Vietnam is highly dependent on things like Facebook and Google. I was there last year, and have some friends that run businesses there.

https://vietnamembassy-usa.org/news/2001/02/vietnam-puts-reg...

I have seen "meaningful, revenue-driving conversation happen", and I do find Zoom calls kind of irritating. I think the real issue is that this tech has exposed the fact that urban planning in the US is awful unless you're in New York City or something similar.

People just don't want to sit in traffic on the freeway for two hours a day for a serendipitous conversation that probably won't happen most of the time. If you think about it that way, it makes perfect sense.

Here's the Twitter edge server from years ago: https://courses.cs.washington.edu/courses/cse551/15sp/notes/...

Here are some pointers:

"Our approach to blocking links" https://help.twitter.com/en/safety-and-security/phishing-spa...

"The Infrastructure Behind Twitter: Scale" https://blog.twitter.com/engineering/en_us/topics/infrastruc...

"Mux" https://twitter.github.io/finagle/guide/Protocols.html#mux

I do agree that some of this could be done better a decade later (like, using Rust for some things instead of Scala), but it was all considered. A single machine is a fun thing to think about, but not close to realistic. CPU time was not usually the concern in designing these systems.

"web previews: I'd do this by making it the client's responsibility."

Actually a good example of how difficult the problem is. A very common attack is to switch a bit.ly link or something like that to a malicious destination. You would also DoS the hosts... as the Mastodon folks are discovering (https://www.jwz.org/blog/2022/11/mastodon-stampede/)

For blocks/mutes, you have to account for retweets and quotes, it's just not a fun problem.

Shipping the product is much more difficult that what's in your post. It's not realistic at all, but it is fun to think about.

Userspace networking is pretty common. The chair of the IETF even wrote one: https://github.com/NTAP/quant

"Quant uses the warpcore zero-copy userspace UDP/IP stack, which in addition to running on on top of the standard Socket API has support for the netmap fast packet I/O framework, as well as the Particle and RIOT IoT stacks. Quant hence supports traditional POSIX platforms (Linux, MacOS, FreeBSD, etc.) as well as embedded systems."

It's a neat thought exercise, but wrong for so many reasons (there are probably like 100s). Some jump out: spam/abuse detection, ad relevance, open graph web previews, promoted tweets that don't appear in author timelines, blocks/mutes, etc. This program is what people think Twitter is, but there's a lot more to it.

I think every big internet service uses user-space networking where required, so that part isn't new.

1914 and 2016: https://www.rome2rio.com/labs/isochronic-travel-times/

This one is excellent. I've eaten Wattana Panich, Ekkamai (วัฒนาพานิช-ก๋วยเตี๋ยวเนื้อ) as well. Pujol might be my favorite restaurant. It's not that hard to get a taco omakase seat at the bar.

Usually it's because there's a database license or something that costs even more than the OS. No one wants to be there, but it possible to get trapped.

It's helpful in learning nearby languages. I am a native English speaker, and my French can approach fluency given some immersion (it's pretty rusty right now, though). DuoLingo helped me improve transactional Spanish and Portuguese (BR) very quickly, meaning I can get through airports, grocery stores, restaurants, and read news articles. I'll need something more to have free-ranging conversations or read novels, though.

I've never really bought this argument. There might be unethical growth hackers in the company, but A/B tests are not the problem in that situation.

In my experience, they mostly just catch bugs. Stuff like "hmm, our much better looking signup flow underperforms... oh, the form is broken on Safari." That kind of thing.

A/B tests are pretty problematic for social networks, since anything effective can also influence the control group. They are excellent for single-player activities like shopping carts and signup flows.

"Network Experimentation at Scale" from Facebook describes how difficult this problem is. Most A/B test frameworks don't reach this level of sophistication. It does make some sense to just ship things if you don't have time to build out something like that. (disclosure: I worked at Twitter long ago)

The early ones are just glossier versions of the NextStep preferences, too. See: http://nixon-development.com/guis/nextstep3/prefs.png

Most of the lakebed is dried up now, though. There are even two huge drainage tunnels to prevent the basin from flooding again (https://es.wikipedia.org/wiki/Túnel_Emisor_Oriente). These canals used to take farm products from Xochimilco to the city center.

An interesting fact is that the big freeways and streets, like the periférico, usually follow the exact path of the causeways built for Tenochtitlan.

The former lakebed is also significantly more susceptible to earthquake damage than the land that formed the original banks or islands.

Well, I'm sure there are "growth hacker" types out there abusing the ability to observe browsing. But I think the real reason they don't bounce you to Safari, Chrome, etc is because users don't stay in the app if they do that.

I think all of the various bad things people talk about here must happen sometimes, but it's mostly just retention I'd guess.

It gets worse the longer the IPv4 address is up, in my experience. Eventually, your address ends up in some log that is inadvertently made public.

So weird, because they actually had the best one for some time. The Windows Google Talk client was amazing.

How did that work when JFK was elected?

This is why you'll find the "© 2022 Twitter, Inc." footer in the right column of the twitter dot com website.

Get a job at a big company, like Google or Microsoft etc. There is always someone smarter.

This isn't really true anymore, in my experience. I've used Bazel with teams of 30-50 and no full-time maintainer, let alone a team.

Once the migration is done, all you need is a few people that do some Bazel gardening every few weeks, and it's certainly not a full time job. This can be someone that does operations (CI, deployments, etc) or a product/infrastructure engineer, or one of each. Github / Gitlab scale to all but the largest projects, and even then, you can just split into two or three "monorepos" and kick the can down the road. With things like BuildBuddy, it's even easier.

As the article states, there are a lot of little of hidden costs and paper cuts when using a many-repo layout. The one that I've seen that's most prevalent is that it obscures copy/paste behavior, since it's much more difficult to detect in a many-repo setup.

Going to Bazel or equivalent is a bit of a mind adjustment, and some languages are better supported than others, but it really starts to pay off in larger projects. Especially if there's more than a few languages in use.

Another problem here is the initial HTTP request on the web, if the site isn't in the HSTS preload list. It's easy for this to happen when you click a link, and there are lots of HTTP links generated by protocol-less highlighting too (think foo.com with no prefix).

What usually happens these days is a redirect to the HTTPS site, but that first request can still be attacked.

You can't paste your password on United's inflight WiFi portal. I eventually got it to work by making Safari share it from my laptop to my phone.

I've never tried to go to Russia, but the process you describe is the way to go if you can spare an extra $100.

For example, it's much easier to get a Chinese visa by paying this company across the street from the Chinese consulate in SF than it is to go there. I think everyone that speaks Chinese knows how to work the system, but I don't speak it at all.

They turned me down for having a bent corner on my passport, lol, after taking my money. I just gave up and used the company across the street for my second try.

Everything went fine for me too, but it was difficult to figure out what to do from California. Sending my passport to them that way made me nervous, though.

It's very difficult for Americans to travel there. But I got the visa once, and went to Oran and Algiers.

I had to send my passport via Fedex to the Algerian consulate in NYC and cross my fingers. There are strange rules about whether you have to send it there or to Washington DC, based on which US state you live in (no one in the US cares about this).

Beyond that, it can be difficult for Americans to understand French with any unconventional accent, let alone understand Arabic at all. I know enough words to count and understand what the police say.

The article makes some good points, but it also shows that the author doesn't have a lot of hiring experience on the closing end (after the candidate has passed interviews).

"For example, anyone who goes by a name that doesn’t match their government ID could be forced into an uncomfortable explanation"

I've seen this happen more than a few times, and the explanations are seldom anything you'd have sympathy for when discovered during a background check. People with good reasons will tell you up front.