18 comments

[ 3.6 ms ] story [ 50.1 ms ] thread
This was a fun read, a few additions:

"Everything fails, all the time" --Werner Vogels, CTO of AWS.

They are in a code freeze right now. Which means they're not changing the system. And as every SRE learns in SRE Kindergarten, the biggest cause of failures are system changes. And a common lesson I learned was that the biggest problem with code freezes was that everyone starts shipping code when they get back. And that bulk set of changes causes increased risk to the system.

This probably won’t be a problem here, I doubt anyone is working on features during this code freeze. They’ve either left or are frantically trying to get to SF and justify their existence.
Scenario 28:

Genocide. People use your platform to orchestrate mass murder, the machetes in churches kind. And fast. Lightning fucking fast. You need to be prepared before. If you don't have a team who knows how to detect and stop this ASAP, your ass is getting dragged to The Hague

This is an excellent list of some shit that can go wrong in any big IT department or company.
all it takes really, pressure, and time
Sherman? Wayback machine...

But Mr. Peabody, that article is still on the web...

https://www.datacenterknowledge.com/archives/2010/06/16/twit...

> The 2010 World Cup has lived up to its billing as a watershed event for the Internet, with web traffic from early game easily exceeding previous records from Election Night 2008. The leading casualty has been Twitter, where Tweetstorms during World Cup games have exposed deeper scalability challenges.

> Traffic spikes over the past week have overwhelmed Twitter's internal network capacity, and the performance problems are likely to continue until a long-term fix is found, the company said last night, meaning outages and maintenance downtime are likely to continue through the remainder of the World Cup.

And when is the next World Cup you ask? It starts November 20th with Qatar vs Ecuador.

And November 20th? That's Sunday. Offices will be closed.

Prediction? The lettuce wins again.

Something tells me this story won't end with "I hope."
Hope is a dangerous thing, my friend, it can kill a man...
While those would all be fun, I'm not sure any of them could doom Twitter.

Maybe if several of them happen consecutively, and people have other places to go and read about the latest Twitter outage/war-crime.

When the foundation to almost every company’s long-term success is the people it has hired, eviscerating those ranks of people indiscriminately is a really great way to have loads of institutional knowledge flee and the job become much, much harder for anyone who remains.

At this point, my only immediately rational conclusion is that Elon is acting out some sort of a revenge fantasy, trying to create as much harm to Twitter as possible without any care that the company might actually collapse. Or maybe even to that end, explicitly, just with plausible deniability.

These tweets makes some great arguments that give legs to my supposition - anyone with any minimal amount of business sense would know of these arguments, and yet Elon is ignoring them wholesale to the detriment of Twitter. What remains is the determination of whether it is ignorance or intentional malice.

OP of that thread here. Wanted to say that the first paragraph of your reply is the perfect encapsulation of what I wanted to get across in that post. It's not about the disaster scenarios, it's about the fact that these scenarios every web-scale company faces every day become a hell of a lot scarier when you just lost >50% of the company without an orderly transition
The other Twitter threads on hn have been extremely bandwagony and contemptuous, and after reading hundreds and hundreds of replies, almost none were substantive, thoughtful, or informative.

Which is why it was so nice to see somebody talk through the technical risks rationally like this person. Interesting write up! I definitely learned something

Here’s hoping the discussion will be of similar quality as the referenced tweets!

I don't beleive you need 5000 people to maintain the servers, even not 100.

If anything it makes it harder to coordinate.

I don't think twitter will die, I mean myspace never truly died. But I did hear its future being compared to AM radio, where the advertisers pulled out (already happened) and so needing a source of income more or less sold to those with agendas (political talk radio, religious radio, minority languages radio). I see a similar future for twitter, where only those with agendas will invest in the platform (since lack of moderation and verification can't be deemed safe for normal advertising), and those with agendas will become the major source of income and therefore the future of the platform will form around them thus alienating the majority of the population and current user base.

More or less it will become a place riddled with propaganda, more misinformation than it currently has and will likely never recover as a place for advertising or as a public forum of sorts.

So it won't physically die, but probably soon it will have sold it soul.