Clearly these engineers have never done customer feedback for developing nations, twitter interactions clearly provide evidence for what musk said, but some engineers seem to be in denial, imho the firing is cause worthy.
Performance has been bad forever. Twitter has had forever to fix it. They haven't. By definition, the existing developers aren't smart enough to do it. The developer here is trying to throw smoke because he's embarrassed. He should be.
Or it could be that previous management was constantly prioritizing new features instead of performance gains and now he's getting unnecessary flak for it.
The worst thing is that by firing and laying off too many people on a team with so much legacy code, they might end up scrapping the app and doing a rewrite because it'll be easier than fixing it. So inevitably, it will run faster, and people will hold it up as an example of Musk being a "genius" . When really, it's exactly what the guy suggested doing, but instead of listening they fire him and force a situation where they have to do it anyways.
Blind assumptions count for nothing. If you actually read the tweet thread Eric (the former employee) gives a summary of the trade offs made that have given Twitter bad performance and suggested removing specific features to improve perf.
It doesn't matter if it's "thousands of RPC calls" or not. Eric's response is, "It's definitely not that! It's... well... we don't know... so just spend more money on servers." That response alone would have me fire him.
> By definition, the existing developers aren't smart enough to do it.
You have no evidence to support this. Performance sucks in other countries because product leadership with their hands on the wheel of the company never prioritizes "unsexy" investments in performance and infrastructure work (same as the new guy btw). That is exactly what Eric was advocating in his thread.
I think that's misreading the exchange. He was challenging Musk on the fact that "1000 RPC calls" isn't what's happening and isn't the reason for the slowdown in his opinion, but did not deny that there are performance issues - and even cited other reasons it's slow. I don't see the denial you see.
Said engineer, said 1000 RPC calls are not the issue, and when questioned further on proving a tracepath / perf result casually exited the conversation, to me it looks like they are not ready to do better or listen to valid criticism, leading conversations on semantics is just creating noise instead of acting on signals.
Also 1000 RPC's are a very valid performance bottleneck, 1000 * latency caused by serde+ network can compounds easily based on data locality + container network orchestration, no data was provided otherwise even when asked for.
Not to defend any of the catastrophe of management that just happened but can't both be true? Eric does say in his tweets that a lot of time is spent waiting on the network. If that is because server side, the endpoint handling a request is spamming other backend services with 100s of RPC calls it would be slow. Now, why you would call out your android dev. on why your backend is not responding fast, I dunno.
That's client-side requests. Musk was talking about internal RPC calls to microservices needed to ultimately respond to the client-side requests.
If the internal arch is heavy on microservices and you're putting together, say, 50 tweets you're at 1000 RPC calls with just 20 per tweet. It's not inconceivable.
---------------------------------
Applied Machine Learning Engineer with 11+ years experience in backend engineering and machine learning.
I have experience in solving for end to end machine learning problems, robust data collection & cleaning, distributed training on GPU clusters, building custom model & training recipes on domain specific problems, model calibration, scalable and low latency inference pipelines in C++ / python (GRPC, k8), monitoring and hard negative mining, Model specific inference optimizations (Quantization, Pruning, Torch Dynamo etc) and have worked on various domain problems like: Search & Ranking, Cold start recommendations, Constraint based optimization, Speech (Speech to text, text to speech), Vision (Segmentation, Classification, OCR), NLP (Doc QA, Classification, Entity Recognition) etc
TBH high levels skills on C++-20 is not what these companies are looking for, instead on C++-14 or so, meanwhile anyone learning C++ in 2022 would rather want to pickup C++-20 ++ with its niceties, old C++ is too tiresome without extra rewards.
Twitter owns the debt. Elon owns Twitter, but can let Twitter go bankrupt on its own and cut his losses.
As far as I can tell, Elon Musk is just invested at "getting back" at everyone who forced him to buy this company at way overpriced values. He's fired the CEO, CFO, Legal Team, and now the data-engineers who supported the legal team's arguments.
Without digging, I'm guessing you thought Musk was just doing a 'pump and dump' when the news broke in April, and that he would never acquire the company.
This assumption of bad faith is just terrible for HN, and the fact that you can't substantiate your argument doesn't breed much confidence in your assertions.
Musk's reputation is on the line without a doubt, and we can all point and say, "ha ha" if he fails. Otherwise...
Same, "We knew that C++ was harder to scale and maintain high quality as you build a dev team" this just sounds arbitary and a weak excuse to use rust, C++-20 is as scalable as rust with a very rich ecosystem.
August 15th 1971, President Richard Nixon announces that the United States will no longer convert dollars to gold at a fixed value, effectively ending the Bretton Woods system. He also imposes a 90-day freeze on wages, prices and rents.
Meh, trump's gonna whine democrats stole the election, media is constantly talk about him for next 4 years, he will win again in 2024 and nothing is going to change. Meanwhile climate change, economy and social media nonsensence will keep getting worse, why ? because nobody wants to really fight for the little guy.
