Hacker News Reader

Comment by kweingar | Hacker News Reader

The most bizarre aspect of this was when Twitter engineers were told to print out their code on paper.

There is literally no good reason to do that. At best it is staggering incompetence. At worst it is an intentionally pointless and expensive exercise that serves purely as a loyalty test.

View on HN

The text messages that were published as part of the suit to close the acquisition showed Musk discussing "return to office" as a way to naturally trim some of the workforce by having the dissatisfied remote workers quit on their own.

My take on the paper printing is that it is a similar strategy to get people to quit by themselves, because they think it's stupid.

The two aren't similar at all.

Commute to work = hours lost every day with drastic lifestyle change. Some people even move further away from the office during pandemic.

Printing code on paper = 1 minute of work. Paper and ink are paid by company. It is not really that much of a problem.

Actually, this is a punishment on the reviewer. No jump to definition and etc.

The printing story is fake. I call it now.

Imagine if you printed off your code and ended up with only a few pages of paper you’d fee pretty silly and may quit from embarrassment

Just a thought nothing to back up this claim

Do you need to print code for that? You can just look at your PR history instead.

Also, if you print code, it will contain other people's code because it is likely not only you who code on those files...

This story is really too stupid for everyone involved.

It is not even good nor efficient for the reviewers. It is like the reviewers are also punished.

Too stupid to be true.

>Imagine if you printed off your code and ended up with only a few pages of paper you’d fee pretty silly

I would copy/paste some hundreds pages from GitHub so Musk would give me a raise. :)

Can someone from Twitter do this and let us know how it goes

Why would I quit from embarrassment? This doesn’t make sense.

Have you written more than a few pages of code in the past 30 days?

I’m not paid by lines of code.

Good lord no, that’s the worst metric to measure someone by. The diffs one produces on a mature codebase are usually tiny but with a small novel in the ticket about what the change will affect.

I could pump those numbers up easily but my coworkers would hate me for it.

Yup, Jason Calacanis "back of napkin"ing a 30% reduction in workforce based on nothing more than some texts with Musk and an hour or two of bloviating.

Do you think the trimming will apply to all levels of performance? Will it push out the poor performers, the high performers, or just take out a random cross section?

Is there a correlation between employees who like to avoid the office and employees you don’t really want on the team any more?

> At worst it is an intentionally pointless and expensive exercise that serves purely as a loyalty test.

The engineers would have used company's resources. I wouldn't be bothered if I am forced to do this. It is a 1 minute of work to print something.

I don't think this part is true. It is likely fake. The source for the print out story is so shaky. Leah Culver tweeted obscurely which is hard to tell whether it is a joke. Casey claims to have screenshots but refuse to show the redacted version of them and end with "subscribe to read"....

The code review part might be true but it is not a spicy story.

> It's one minute of work to print something

There's a bunch of comments echoing this sentiment, and I'm wondering if something's wrong with me. I'm 27 and haven't printed anything in an office... maybe ever? I've only peripherally noticed HR and office managers constantly struggle with printers. At this point, I'd probably be the one asking my parents how to print something :')

At the office like twitter, printers are already set up on your work laptop for you.

You basically chooses a printer (there are 10s of them) and click print.

Literally a minute of work.

Also, you are lucky. Visa employees need to print stuff from time to time for h1b, green card, and citizenship. And you bet it. They print it at work because nobody has printer at home.

I am pretty confident that most developers would take significantly longer to gather and organize all of their code, make sure that everything is in the correct order for presentation, and print it out on paper.

It would probably take me several minutes (if not longer) just to figure out how to display all of my recent changes across several repos, let alone print it without long lines of code getting cut off at the margin.

There are several other considerations you’d need to think of as well. Am I just printing diffs? If so, how much context do I need? If not, do I print out whole source files, and do I need to tag the parts that were part of my commit? Should I print 100-page generated files that I technically checked in to source control?

I would be very interested in a workflow where all of this can be accomplished by a typical engineer in a minute.

> Am I just printing diffs? If so, how much context do I need? If not, do I print out whole source files, and do I need to tag the parts that were part of my commit? Should I print 100-page generated files that I technically checked in to source control?

These should be answered by the order from Musk.

I assume the news just reported on the high level details and skip the low level details.

The other theory is that this is straight up fake news. That is why there is no more details. I actually think it is fake news.

Now let's pretend it is real. Whatever the answers to these questions are. They don't sound difficult to do by any mean.

If they want the diff, we print pull requests.

If they want the whole file, we jump to the files we touch and print the files.

If they want 100 pages, we go through each file and print until we hit 100 pages.

If they don't want generated files, then we avoid them.

What is going on here? It is not that hard by any mean. I'm not sure why you act like you can't do this thing easily.

I'm not quite sure what your argument is.

I suppose if you click slowly, it may take you 5 minutes? Okay? Is that exponentially harder or what? Would you get insanely frustrated because the task requires 5 minutes of your time?

It's easier than giving source code access to all the outside staff that are coming in to actually read it. But it doesn't really sound like it's a meaningful review, since "last N days" of work isn't useful without context, and only people on the team can give them that.

It's truly bizarre.

Were they? The article said they were told to "stop printing". As in "stop the presses", perhaps. Still sounds weird, though - maybe the author of the invite is not a native English speaker?

Maybe this is a way of asking engineers who have zero pages to print to walk out?

Or maybe this was to set the expectation that only those who care enough to bend to the whimsies of the new management should stay?

Or both?

Nah paper is still more efficient than computers in quick meetings. You can imagine many seconds being wasted if they were on a laptop and the engineer needs to find his files.

Paper the UX is perfect and instant.

Paper is instant until you say “can you show me that part of the code that handles X” or “where is this function defined” and then it’s multiple orders of magnitudes slower.

(Editing to say, you could also just tell the engineers “hey make sure your code is pulled up and ready to go before the meeting starts”)

"Please send us links to your changes before the meeting begins." And display them on a large screen.

I don't think it's really relevant, though. The entire situation is cruel and silly.

Elon fan brain is a debilitating condition

I very recently reacted the same copy/pasted code that was in in 7 files, each one had 6 touch points of 5 lines each, and the touch points were 50-300 lines as part. Would be a lot of paper for what had been a 30 minute exercise. I cantquite how to picture going over that diff on paper. The time wasted doing so seems sad, as do the natural resource waste (which tends to make me think the guy got into electric cars because it was cool and had an open market opportunity rather than ecological reasons - but that is a very random aside)

How do you search, tag, jump to code definitions, look up libraries referenced if it is in paper?

Unclear to me why this is being downvoted. I always read code on paper, it’s easier to annotate.

If jumping around a codebase is hard on paper, it’s a good indication that the code does not consist of cohesive modules.

Sadly these days most people no longer adhere to 80 column limits, so it has to be printed in landscape. In the worst cases, landscape on legal paper…

>There is literally no good reason to do that

Or simply Elon is more unstable than Dogecoin and wanted to show new employees who is the boss, it was a really pathetic power move.

Reply on news.ycombinator.com