I was thinking the "30-60" day time window may be a somewhat subtle way of catching out engineers who haven't written a thing in that amount of time. that would be one way to identify low-performers.
not unproblematic by a long shot, but quick, and probably somewhat effective. and it wouldn't require any deep knowledge of twitter's codebase on the part of the reviewers.
Any slightly competent software leader would know that you can track every character of code written by any dev using version control.
Printing it out as a mark of honor/shame is something a petty, arrogant tyrant would do. Musk has been around people who are afraid to say no to him for too long.
I've been in a few situations where we printed it out. It's just a lot faster than granting someone access to internal systems. It's also a firewall of access if you need that legally. Also you can sort paper into "to look at later" and "seems fine" (like huge chunks of json), it's annoying but not terrible for diligence.
> a lot faster than granting someone access to internal systems
Why? Musk could order access for Tesla engineers in minutes. Large companies have automated on-boarding and access control.
> a firewall of access if you need that legally
There is zero chance Musk would care about this. He breaks laws publicly and exuberantly. He has never shown a respect for red tape.
> Also you can sort paper into "to look at later" and "seems fine" (like huge chunks of json), it's annoying but not terrible for diligence.
I just can't imagine paper being useful to understand even a small project with 40-50 files in it. I'm not really buying there being any reason for it except a legal one, and this situation at Twitter wasn't a legal issue.
I wasn't defending their supposed use of paper, just pointing out reasons why you might find yourself doing so.
I've also been at quite large companies that need 3-5 days notice to onboard contractors, background checks, etc. Besides, he wants to fire 70% of them... you think people are rushing to help him fire everyone themselves included?
He fully understands VCS and is a pretty capable software engineer.
My guess is he is planning to fire ~50%+ of the engineering team and this is just his way of letting each person get their chance to show their worth in a quick interview.
The article says only some engineers received the invite to print code, and Washington Post's sourcing (some engineers at Twitter) is the same as yours.
You want me to post an evidence that Elon didn't ask for a print out. Like asking my friend to show all thousands of emails in his inbox and say "see? Elon didn't email me at all"?
You said it yourself that only some engineers were asked that. So, by your logic, most weren't even asked anything. Twitter has 5000 engineers. My friends are definitely likely within the thousands of engineers who aren't asked for anything. I don't think I need to dox my friends for this.
You or casey should be the one who provides a more reliable evidence, especially when you make a crazily ridiculous accusation that is borderline comical.
Casey said he had screenshots but not disclose them. Well, okay. We'll take his words for it? He could have easily redacted names and etc. But nah.
Well, he did say "subscribe to read" in his tweet. I'll call it. He will never disclose the screenshots or more detailed evidence. He got retweets and some subscribers out of it. Then, we will have fun with this story, forget, and move on
The fact that you even ask me to prove a negative means you don't have a better evidence and that you like this fake news so much that you don't want it to be false.
I wasn't asking you to 'prove a negative', I was specifically asking you to prove a positive. Which is that you have friends that work at Twitter that say otherwise.
Moreover; if you DID have friends that work at Twitter this thing would be very easy to disprove because they could very easily go through the corporate slack or whatever they use, see if people were asked or not and done. Mission accomplished. That doesn't even require doxxing your friends specifically.
But you're not going to do this. Just like Casey, you can easily take screenshots, redact names and done. I don't think you actually have friends at Twitter to be honest nor do I think you were ever willing to engage in good faith.
> I wasn't asking you to 'prove a negative', I was specifically asking you to prove a positive.
Yeah, you did ask for the negative. My friends weren't asked for anything at all. Not even code review.
This aligns with what you said earlier about "only some engineers were asked", but now you purposefully ignored what you yourself said.
How would they even prove that Elon didn't ask for a print out if they didn't get any message from Elon?
> But you're not going to do this
You are damn right I'm not.
1. My friends are likely in the thousands of engineers who aren't asked for this. You said it yourself some engineers are asked for this, and you think Elon asked for this kind of sensitive thing in a public slack channel that everyone could see? That is ridiculous. (Hint: nothing in slack either. My friends did search for it)
2. I'm not the one who makes this
ridiculously comical accusation. The more ridiculous accusation, the stronger evidence should be required. The onus is on you and Casey who insist that the story is true.
I know you like the story but damn please have some standard for verifying whether the news is fake.
> you think Elon asked for this in a public slack channel that anyone can see? That is ridiculous.
No, this is you misunderstanding Slack. We know they use Slack for internal communications and we know during times like these people are going to discuss anything controversial going on at the time. If people were told to do something ridiculous, there is a guarantee that they would also discuss this on the internal slack in some general channel. Either the news story itself or the fact that people were asked.
> I'm not the one who makes this ridiculously comical accusation. The more ridiculous accusation, the stronger evidence should be required. The onus is on you and Casey who insist that the story is true.
No, the onus is on you. You said you had friends at Twitter that disagree. There are multiple other points of evidence pointing in the direction of engineers being asked to print out code. They've substantiated their side of the conversation, you are refusing to do the same despite making a clear claim that can very easily be shown to be true. If you aren't going to throw down your own claimed evidence then there's no reason to believe anything you type. And now you're trying to avoid this by changing what I've said.
> We know they use Slack for internal communications
Yeah for non-sensitive things. Nobody discusses highly sensitive in a slack channel that can be seen by thousands of employees, especially something that is applicable to only some engineers.
> They've substantiated their side of the conversation
They have not. There is the obscure tweet from Leah Culver. Then, there is Casey saying he has screenshots with the ending of "subscribe to read more".
> You said you had friends at Twitter that disagree
Having friends at twitter is a common thing. There are 5000 engs there.
You said only some engineers are asked. Twitter has 5000 engs. It is not unbelievable that my friends are not asked anything.
Meanwhile reviewing code on paper is such a ridiculous thing to do. It is more painful for the reviewers themselves, meanwhile it takes an eng 1 minute to click print all. Why would they even ask for this? It doesn't even make sense.
Meanwhile you or Casey provides such a filmsy evidence and insist that this ridiculous story is absolutely true.
For one, I never said 'some engineers are asked', you made that up. The closest thing I said was 'multiple engineers were asked and shared that they were told to do so'. Which means we have anywhere from M confirmed to N unconfirmed.
Second, this is not a sensitive topic, given that we've seen someone an employee (according to you) make light of it. So either it's something that's fine to joke about (and therefore should be easy to disprove), or it's a 'sensitive matter' and Leah's tweet should probably be taken seriously.
And in either case, employees can and do talk about things like this in said Slack channels. All the time. Presumably it would be easy for your friends to ask their coworkers and their coworkers coworkers on Slack for confirmation. And for you to provide this evidence. That is my standard for verifying news, which you are not hitting.
> Second, this is not a sensitive topic, given that we've seen someone an employee (according to you) make light of it.
It is a sensitive topic.
Leah tweeted it obscurely that we can't tell whether it is a joke or the truth.
Leah Culver bought a house for 3m before the pandemic. The house is the iconic pink painted lady as her second house. She has a second house in SF. She doesn't need a job. She is rich as fuck. She wouldn't have endured this kind of brainless activity.
> Presumably it would be easy for your friends to ask their coworkers and their coworkers coworkers on Slack for confirmation
They did ask and didn't hear anything about it.
The caveat is that they aren't gonna ping execs or ask it in a public channel with thousands of employees in it. They gossip but in a tighter nit group.
The onus is on people who make ridiculous accusations. Having friends there is not ridiculous. My friends not hearing anything is also not ridiculous since twitter has 5000 engs. Reviewing code on paper is ridiculous. It is even bad for the reviewer
I only see Leah Culver tweeting obscurely. In the next tweet, she responded to Musk's tweet about "comedy is allowed on twitter" that she "just followed order".
Leah Culver is wealthy that she doesn't need a job. She bought the painted lady a few years back for $3m. This could easily be a gag from her.
Are we back to measuring performance in klocs? I can get you a bunch of incompetent devs that change hundreds ofines of code a day. All you need to do is change indentation and eols or just rewrite basic stuff too lazy to understand.
According to the engineers in the undercover interviews, a large part of the staff only work "when they feel like it", which would result in some taking months of paid time off or just hanging out in the office all day.
kloc is not great as a general metric, but in this case it might make it obvious where you can trim a lot of fat..
I see and the issue lies with the engineers? If thats the company’s policy the whats the problem? He should change the policy and only then review performance in that context.
right, they'll fire the gurus who know how everything works and barely write a line of code anymore, and things will be pretty hilarious when they have their next outage.
not unproblematic by a long shot, but quick, and probably somewhat effective. and it wouldn't require any deep knowledge of twitter's codebase on the part of the reviewers.