We ditched the dial-up modem quite awhile ago, Grandpa. While yes, I guess it’s possible, the idea that Twitter is using remote procedure calls instead of REST immediately felt wrong. It also felt wrong to blame the codebase for an intermittent problem without mentioning an accompanying change.
I think the author has some idea of SOAP or CORBA in mind for RPC rather the more modern Thrift or gRPC solutions that are used extensively everywhere. I don't find it that unbelievable that Twitter would do 1000s of RPCs on the backend to render the home page, but that says basically nothing about if it's the bottleneck or not, after all.
I don't want to defend Elon – far from it – but I think the author doesn't have their ducks quite in a row here.
Or basically, they have a constrained and very mistaken belief that "RPC" is a specific technology. RPC literally means "remote procedure call", and it's typically used to mean just that. It's, for example, a perfectly reasonable observation (and common in HATEOAS threads) that REST APIs these days are typically just RPC over HTTP.
I'm not going to defend Musk, but I find it hard to hate him either, given that every single criticism I see on the Internet is at this level of stupid, or worse. It's really hard to tease out real issues from the torrent of asinine soundbites of critics trying to one up each other.
> every single criticism I see on the Internet is at this level of stupid, or worse.
That's not at all true. There's plenty of good criticism of Musk out there. It's not hard to find valid things to criticize, from bullying behaviour to overpromising (Tesla FSD) and lots of other stuff.
I qualified it as "every single criticism I see" for a reason - if there are good, strong critical points, I haven't seen them, because they're drowned in a sea of nonsense. Honestly, even recent HN threads have been suffering from the same problem of people just repeating trivially disprovable or purposefully misrepresented nonsense without a second thought. I think we may just need to wait until Musk stops being Internet's Public Enemy #1 to see anything clear.
> It's not hard to find valid things to criticize, from bullying behaviour to overpromising (Tesla FSD) and lots of other stuff.
Also, I should perhaps have qualified more clearly that I meant criticism on the level justifying the current level of dislike and hatred. Of the things you mentioned:
- Tesla FSD overhyping is a valid one, though it's been exaggerated by the media and then dumbed down to a simple gotcha on social media. As far as I know to date, he didn't go beyond what's common in marketing in general.
- Bullying behaviour - I believe he did things that could be construed as bullying, but back when there were stories on it years ago, I couldn't dig into anything solid - everything I read had clear markings of usual journalistic misrepresentation. I'm giving it a big "shmaybe".
- "Lots of other stuff" - other than that diver debacle, can't think of anything. All I see people post is trivially disprovable bullshit like "he's not an engineer" (he is), "he isn't doing engineering at SpaceX" (he is), "his money comes from his father's apartheid emerald mine" (it doesn't, the mine was in literally the most anti-apartheid country there was, and his father IIRC was an anti-apartheid activist). Or "RPC in 21st century, he must clearly be an idiot". It's an unending stream of this kind of "arguments", posted by supposedly smart people. All it makes me is to lower my opinion about people posting this stuff.
Well, as engineers, the only criticism of mr. Musk that matters is that all his companies are authoritative, very badly paying companies that work their employees to the bone. Very bad places to work. At the very best they're like a doctorate: you work your ass off for a minimal number of years not having a life, but you try to learn a lot and go anywhere else.
Presumably you haven't read this article, which seems to scratch a few different surfaces. If you can read that and still come away thinking "every single criticism [you] see on the Internet is" stupid, well, nobody can help you.
I haven't until now. Yup, this article is stupid too - most of the list there is noise, or reaching. There are a few good leads, but even taken together, they're hardly enough to justify the current level of vitriol.
There's a class of bad leader/manager/executive/person-in-charge who has mastered the ability to generally not do or say anything on the record that would be a red flag, but leaves a wake of orange and yellow flags behind them. I didn't really understand that until I'd worked for one.
People make mistakes and have personality differences, so one person's yellow flag is another's business-as-usual. IMO it's only fair to set aside some of those lesser warning signs, but when someone continuously accumulates a significant number of them, my experience is that it's equivalent to a giant red flag.
That having been said, a new person-in-charge jumping in immediately and ordering a bunch of significant changes without taking the time to understand the environment is generally a giant red flag on its own. There are a few exceptions, of course, like if lives are in danger.
I find it hard to find much redeeming of how he's handled this entire takeover of Twitter: from the attempts to weasel out of buying it due to his own lack of due diligence to the impromptu layoff of half of the workforce, his behavior here stinks to high heaven.
Nitpicking is fine and encouraged, but the engineer himself started the thread by saying, "I have spent ~6yrs working on Twitter for Android and can say this is wrong."
I agree with what you're saying. It was the way Elon made the statement that activated my spidey senses that he was talking out of his ass. Twitter has (had) some of the best engineers in the world.
They may have had some of the best engineers in the world but that is decidedly less important when it's combined with some of the most mediocre mgt in the world.
Are there any objective employee sentiment surveys about what they think? Even after all the bad press, I can picture that many are still cautiously optimistic (or just interested to be along for the ride), especially if they previously weren't happy with Twitter's direction. There is a lot of internal rot in tech and a lot of people putting up with it because of their pay. I've worked in jobs where a lot of people would have been happy to see someone come in and shake things up, even if it was disruptive and risky.
From the outside it looks like complete and utter madness.
The thing is, I don’t care if he runs Twitter into the ground, it sure is a good show while it lasts. I’m also pretty certain the engineers will have no trouble finding a new job.
But what concerns me is that radioactive aspect of it.
If Elon has lost his mind, what does that mean for his other ventures?
I dunno, an ambitious young eng with no other responsibilities could stay just to see how much they're about to run basically by themselves. what other ways could you get that kind of experience? if the site goes down well ah well that was a difficult task. if they manage to keep the site going though, juggling keeping the shitton of microservices up with a megalomaniac at the helm trying to kill them off. that experience is probably worth a few hundred k in TC at whatever place lets them tell the story and gives them an offer
They could leave, but I think unionizing is a better solution if they can pull it off, and it would set a better precedent for programmers affected by similar things in other companies.
I don't want to defend Elon – far from it – but I think the author doesn't have their ducks quite in a row here.