[DoneWithAllThat]

The claims in this article assume a lot. To pick just one, regarding stability/uptime: without knowing the distribution of layoffs by department, this may not just be incorrect, but the opposite might be true: if most SREs were retained, and engineers who know the systems well are still around, stability will likely improve as there will be fewer changes to the systems (usually the number one source of outages). If PMs were heavily cut then it’s possible they have infra that was slated for future projects that will no longer happen which gets them extra capacity.

Not saying any of what I spelled out above is what happened. But it is entirely plausible that thousands of jobs across the organization could be cut, and stability and uptime could improve.

[summerlight]

https://www.cnbc.com/2022/11/03/musk-orders-twitter-to-cut-i...

Musk wants to save $1B from infrastructure costs, so the service quality will very likely degrade. The current business focus is money saving at all cost so I don't think SRE or whatever engineers are not very safe from layoffs.

[paxys]

Services that are on the hot path of serving requests and generally keeping the site up and running are a tiny, tiny fraction of overall infra costs at companies like Twitter. The majority of spending is on storage and processing of logs, metrics and analytics data. Now you can argue that Twitter will be worse without this data, but it isn't as simple as "Twitter will go down more if you cut infra costs".

[summerlight]

I don't think you understand what you're talking about or Twitter's business model. Those log and metric, analytics are the reason for advertisers paying Twitter. What's the point of saving $1B if you lose $2~3B of advertising revenue?

Oh and good luck with cutting dependencies on those storage while firing all those engineers with knowledge on those infrastructure. From my experience, migration and deprecation is usually much harder (I would say 2~3x difficulty) than launching a new stuff especially in an established big company.

[BryantD]

True, but given that he's looking for $3 million a day in savings from servers and cloud services[1], I'm not wildly optimistic about how busy the remaining SREs are going to be. No matter how many of them there are.

There's always money in the AWS bill but that's a lot of money to expect to be able to find. And, um, this kind of thinking is one of the things Mudge was talking about.

[1] https://www.reuters.com/technology/musk-orders-twitter-cut-i...

[smarcellus]

Doesn't seem likely considering Musk's intent to push a ton of new experimental features within a week. https://twitter.com/RMac18/status/1588304086792667136