Obviously musk doesn't have time to review it all... But saying 'print out a couple of pages of your recent code you are proud of and bring it to a review with your tech lead' is a good way to discuss coding style.
The 'you are proud of' is key to this working though - all your code is visible in the VCS anyway, but asking you to pick a good bit to discuss seems valuable.
From this perspective this seems very reasonable, if I picked some code which I like it will be very hard but I believe it will force me to consider what I don't consider in the day to day.
Am I really proud of what I do? And what do I do really?
Maybe with a bit of back and forth with team mates it might help people like me as well who struggle with imposter syndrome on a daily basis feel like they can be proud of things they do.
It helps to weed out total slackers, of which there could be non-negligible percent. If you have any shred of decent code - you are ok. If you can't produce a page of decent code after working a year - why would anyone keep such programmer.
Why in the Sam Hell would you need printouts. Literally look at git histories and contributions per user. This is bending over backwards to explain what is most likely a simple power play that most probably backfired
If you read the thread, he appears to have asked for the last 30 to 60 days of code, printed. So not code they're proud of, literally everything. And why printed?
And why weren't they already canned? Slacker tls and managers. Next thing you know Elon is looking at himself.
Yeah bring code so we can talk about it is fine at the bottom level. Dumb for owner in first week. The good programmers know nothing will ultimately come of this codiscovery process thus making a bad impression on the good employees
Not a musk fan, but if you purchased any company would you not want to get to know the people that make it work, what they are up to, and what excites them?
Not saying his intentions are noble, but the act of these mettings, in and of themselves, does not signal anything in either direction.
What you say is of dubious value (or even negative value). I imagine Twitter has code reviews like most other places anyways.
And even if were not of zero value, it's not what they were told to do anyways. Printing out your code of the past 30 days is a demeaning power play to test your loyalty. They could just look at your pull requests or commits themselves.
If a tech lead told me to print out some code and bring it to a review, I'd strongly reconsider the team I'm on. Not only is it among the worst ways to review code, but it also includes a lot of irrelevant information.
Why? Printed code is code, NASA/Military does it for formal review. I've worked on avionics systems that are developed in similar fashion. You ought to be able to explain what it does. Most of it is going to be plumbing. Call this, fetch that, join foo, collate bar.
Strong arming Elon's intentions, may be he is used to perhaps how code reviews work in Space industry? Sometimes old school ways are fine, under the hood it is just code.
No, no they do not. I've worked for the government and in that sector. We did not print code to review it. That might've been a thing a decade or two ago, but it is not a thing now and for good reason. And I would strongly question the technical capabilities of a team that was not able to adapt to modern tools even for older codebases. What exactly is your citation that they still do that?
It is simply much quicker and much more efficient to do something like a proper code share in an office with one person driving the code and being able to jump between areas of the code.
To say it bluntly: Anyone that uses printed code for code reviews is wrong. It assumes all functionality is colocated and nearby. It wastes time by forcing people to index through multiple pages of paper. Anyone that demands it should see prompt pushback.
I worked in 2010, so yea it might have moved on. It's probably in powerpoints now ;-) I worked in military (not space) though. I still know people there and I'd guess it is not that modern. We'd also make bluebooks in 2010 with code printed and neatly bound in them.
I worked both in the military and the military industry in the early 2000 and we never printed out any code for review.
We did have very thorough design documents and manuals and all of them have been printed (including documented function headers) as the contract dictated, but we only used the original word documents and the version control system to review our code.
People keep saying printing code is bad for twitter engs.
If I'm forced to print code, I would probably do it. It is essentially a few clicks taking a few minutes of my time.
On the other hand, your tech lead reviewing code on paper is going to be extremely painful. The tech lead who is forced to review code on paper would probably quit first.
What possible value to the company can come from the CEO doing this stunt for coding style? Don't you think there are already company-wide guides for this?
It sounds like something that someone from outside the software engineer world would think it's valuable.
Judging the sizes of the stacked pages from the last 60 days of code would probably be a pretty good indication of productivity. Perfect? No. Good enough? Maybe.
I never thought I would see someone on HN unironically advocate for counting LoC to measure productivity. Unbelievable how this guy's cult of personality poisons brains.
100% right. I just watched an OJ Simpson trial documentary. From that making the small point being famous brings perception then perception management... Which goes fast to nowhere. Fixing Twitter can't start with loc and the new owner reading code. I can guarantee Gerstner never did that at ibm.
I’ve worked in big companies and I’ve worked in small companies. The one constant has been that writing code is required to make product changes. No code likely means no changes. Maybe LOC != productivity makes sense when algorithms are of great importance, like in situations where a genius algorithm can unlock tons of value. Machine learning would be one example. However, most work is feature work, which typically has a straightforward path from start to finish. In product feature work, it’s unclear to me what activity would use time productively that didn’t result in lines of code.
People, even those at other big tech companies, unironically refer to working at Google as "retiring." Not sure 'big tech' is a great example to use when talking about productivity.
As in, the fewer pages, the more concise they are? What about clarity, and which languages are we talking about? Bonus points for Java? I think it’s been fairly well determined that LoC isn’t a good metric for judging productivity.
I think LoC has been determined to be a bad productivity metric because it’s gameable and incentivizes bad behavior. However, that doesn’t apply here as the engineers didn’t know they were going to be evaluated by LoC a priori. I struggle to think of cases were lines-changed wouldn’t be correlated with the productivity of engineers working in a consistent environment if they weren’t trying to game the system.
The problem is some languages are more verbose than others. Considering boilerplate (which is often automated these days) it extra has nothing to do with productivity. More lines doesn't mean more action - I mean, we're on a Lisp-derivative forum.
And sure, if I was getting judged by LoC I'd just paste in a library versus including it, which would generally be worse.
> Amid layoff fears, some Twitter employees were instructed to print out copies of the code they had written recently as possible proof of work, said two people with knowledge of the situation.
People outside SV think its full of smart people, but when you work there, you will see how real the TV show Silicon Valley is. I am convinced if not for the money, a lot of people would leave.
Perhaps Twitter is staffed by middle managers who sell well in interviews but have zero fucking clue.
I’d like to see Twitter fire the cottage industry of professional non-contributors that has seeped into tech for the pay.
Employees like the TikToking “Facebook PM” who showed off 20 mins of emailing for work and the rest of the day brunching/lunching with their crew need to go.
Anxiety and a lot of gossip can make people do dumb things, like make non-technical middle managers issue stupid edicts because they don’t know about git logs.
I’ve worked for orgs where engineer head count alone pushed 250 people. The number of people in power with no technical skills managing technical workers was a huge waste of time.
The article was observed, not the events. I know PR people who plant such articles for profit, all the details are vague enough to seem possible. Even the supposed pictures of eng folk with printed code could be in on the gag. CNBC got trolled by a guy faking being an ex Twitter staffer. Someone from Twitter posted “the algorithm” repo as a troll on Musk.
Why not troll the public and traditional media who are far more obsessed with office life than doing real things for others?
Why bother taking the article seriously? It’s unverifiable. I see no reason to give it a sincere discussion.
It seems like a way to figure out if some people are actually coding. They might be faking it or maybe doing a lot of StackOverflow copy pasta. Print out something you’ve written recently so we can talk about it. Not defending, but that’s the most likely explanation.
In all fairness it's not stupid. Interviewers sometimes ask to show code you have written and did not pay to write it. Here the interview asks to show code you have written and been paid to write.
Copy stackoverflow, understand it, link to the original source. What's wrong with it.
Those lines are normally the hardest. Maybe referencing a weird behaviour, not implemented feature, and stackoverflow is the only place that explains it
OK, so it looks like someone working for Elon, possibly non technical, at first asked employees to print out code to discuss in a review. Possibly resulting from a miscommunication, or to ensure no holdups due to computer crashes or something.
The story I had until I finally got to this article was:
Elon rounded up everybody in the company and personally, possibly and likely with a megaphone, demanded that “whomever is in charge” and/or everybody in general to immediately drop all their work and print off the entire code base so he, Elon, can personally perform a code review with like a highlighter in a meeting room while his 50 Spartan Tesla engineers watching like gremlins so they can cackle at the code quality.
Twitter engineers would then also need to stay at the office so they can be personally called in to be laughed at and fired on the spot. All the other twitter engineers would be lining the halls and be forced to laugh at the fired engineers as they’re escorted out in handcuffs. Failing to laugh and point adequately, or falling asleep, means they would be next.
Listen, it's fine to have outside engineers audit the source code of a newly acquired company, but literally requiring them to print out their code is insane. It is entirely possible that the article was wrong about that point and he just asked for access to their github repos or something, but if it's true it points to incompetent leadership on a truly impressive scale.
Honestly, whiteboard interviews are much better IMO than opening up an online-window-of-sauron that the interviewer can see what I am typing and it makes me super concious. I froze in one of the interviews, couldn't remember syntax about a language I've coded for 8 years. WTF. After the interview was cut short, I had a good 2 hours of self-reflection. Strangest feeling ever when someone misunderstands your strengths from some shallow shit. Have you experienced a time when you can't believe this is how a word is spelled? It doesn't look right, but it is. There is a term for this phenonenon in psychology, I forgot.
Most comments fail miserably to understand the underlying meaning of such move.
Most comments critics the lack of usefullness of such request.
Most comments are funneled in a technological sideration, and fail to consider such request in a context of power take over, social hierarchy and group dynamic.
Please, stop being hypnotized. And think a little out of the box of your modern, sterilized, way of thinking.
Also, it is well known that Musk hiring process emphazises on "doer" profile. Not only the activities of his company are highly seducing to high intellectual profiles, Musk over select them based on the "doer" criteria. People highly intelligent, but emprisonned only in theorical masturbation are not selected.
"Show me your code" is nothing different than "show me the rocket you build in your yard"
Lot of smart people can talk about rocket. Only a few are doing rocket.
Lot of smart people can talk about code, only a few are coding.
So Musk is just trying to detect who are the 20% coders from the dev staff that are doing the 80% of the work. In order to fire the 80% remaining as promised.
This is the first time he bought a big company which was not designed from origin by his paradigm.
Most of big company are kindergarden (to be nice to them). Musk now is turning twitter from kindergarden to efficient, profitable structure, in regards to Tesla or SpaceX.
This is quite a big challenge, as far as Elon's paradigms were not part of the inner culture, staffing policies of the company, since birth.
Let's see how it will go.
ps: If once, you place yourself in the field of complexity, collective intelligence, group dynamic, you first understand that companies structure are the only allowed (accepted) form of experimentation by the capitalism system.
The family structure has been totally destroyed by promoting all kind of law attacking the inner mecanism of family structure viability. Therefore blocking all possibility of collective intelligence from it.
The religious structure has been totally destroyed by instillating all kind of memetic virus such as new-age bullshit to sabotage all the core mechanism of religious structure to offer any fertile environment of collective intelligence.
The only allowed space is the business and company. In such paradigm, every company is an experiment of collective intelligence, group dynamic, and complex system.
Twitter, or any kind of social media have a special place in this experiment. As far as their main business is collective intelligence itself, (introduction of recursion principle).
So for all peoples asking themselves why Musk made this purchase. It's seems pretty logical. As far as the efficiency he demonstrated to design very fertile environment for the emergence of a outclassing collective intelligence, demonstrates that he is well aware of the topic. And purchasing twitter will only (i hope) allow him to enter the metalevel of collective intelligence experiments.
Citation? Feels like he's got an aura of "genius" surrounding him so people think he's super clever. Meanwhile I wonder if he's just a tiny bit more smarter than Trump and he's been lucky in his whole career...
So given you didn’t read it - the engineers were told to print out their last 30 days of changes so that musk could read it.
Then they were told to shred it a few hours later.
This had nothing to do with VCS (which they obviously use), and everything to do with musk’s ego.
The way I view it is that it's more or less a way to harass people. Get them to print out code and then if you do follow-up, you can weed out people that rebuke your authority.
He was never going to read the code (or even really understand it), it's just pettiness from a petty person.
Obviously musk doesn't have time to review it all... But saying 'print out a couple of pages of your recent code you are proud of and bring it to a review with your tech lead' is a good way to discuss coding style.
The 'you are proud of' is key to this working though - all your code is visible in the VCS anyway, but asking you to pick a good bit to discuss seems valuable.