Misleading headline, He's saying to submit before the merge window, not on the last Friday of the merge window.

Linus has been complaining about patches arriving late in the merge window for at least a decade at this point. I still don’t understand why he hasn’t decided that the first week is the dead line for submissions and the second week is only for review and the eventual shenanigans linked to the review. It would definitely solve the problem.

Yeah, that complaint is a bit weird:

> The rule is that things that get sent to me should be ready before* the merge window opens, not be made ready during the merge window.*

That's not what a "window" is. If you want the beginning of the two-weeks period to be a deadline, call it a deadline.

I may have missed it, but I can’t see where he tells kernel devs to “grow up” in the linked page?

He doesn't say this directly, but I suppose his comment here:

"The whole 'do an all-nighter to get the paper in the day before the dealine' is something that should have gone out the window after highschool. Not for kernel development."

is making an explicit comparison to the behaviour of young adults (namely: poor time management, crunching for deadlines by staying up late). I agree with you that it's a poor editorialisation and makes it seem like Linus is being harsher than he really is in his writing.

Unrelated, but in general, a deadline can be used as a tool: avoid unnecessary work, focus. Especially, if it is a soft deadline. It is not just for teenagers, there are whole books on the topic e.g., "The Deadline Effect" by Christopher Cox.

He doesn't, but he does say that adults shouldn't be doing it, which can be interpreted as "grow up/act your age." I wouldn't really interrupt it that way, though.

>The whole "do an all-nighter to get the paper in the day before the [deadline]" is something that should have gone out the window after highschool. Not for kernel development.

>I wouldn't really interrupt it that way, though.

How so? I don't see any other interpretation

And while he might not use the exact words, it's a pretty accurate editorialized summary of what he says here

He's just explaining how kernel development is different from academics.

Telling someone to "grow up" is kinda mean and should be reserved for when people are being extremely immature. Linus didn't say that.

>is kinda mean and should be reserved for when people are being extremely immature. Linus didn't say that.

Isn't being kinda mean Linus trademark?

"Grow up" vs "leave it at high school" is "potato, potatoh" kind of difference...

He did specify high school though, rather than college/university (or a generic like “school” or “academia”).

We’re probably all over-reading this quote but it did come off as a “grow up” comment in my reading as well, intended or not.

There's the thing about how pulling an all-brighter to get something done before the deadline should have gone out the window after high school, which I guess is what's referred to? But this is just Thorvaldsen trying to encourage a healthy and sustainable process for himself and others, the title posted here is ridiculously editorialised click-bait.

Speaking of college and all nighters, I'm reminded of the work patterns of teachers. Particularly primary education teachers.

They put in immense amounts of work during the school year. In exchange for that summer vacation. And pay and job satisfaction, we hope. We can increase pay, and decrease classroom sizes.

But those teachers are still going to take home work and joke with their partners about it at the kitchen table, right?

What's the other solution? Doubling teacher employment and decreasing classroom sizes to half-day?

It's a possibility. I'll call that bluff that it comes without drama.

I'll bookmark this post, you know, because I'm slow to recall.

The pay for teachers is dismally bad. Most that I know purely do it for job satisfaction which I respect immensely.

And, as others have pointed out, some may want that meditative scoring time.

If people continuously fail to follow the process (and we're not talking about general population here) then maybe the process should be reviewed and changed so he does not end up with a two weeks work crammed into 2 days.

Yes, They could perhaps consider telling people to have their work ready before the merge window starts instead of treating the end of the merge window as their deadline. /s

This is something I have been trying to root out at my big company inc. Until I said something deploying late Friday was the norm and all the issues that come with it. Now we deploy on Monday and have a whole week to fix issues and then rest. Rather than cobble shit together by Friday and then throw it over the wall.

Some people produce their best work under pressure, even if that pressure is self-inflicted.

Turn the merge window beginning into a deadline, and only accept follow-on patches during the following 2 weeks review.

I perform my best work under pressure. When everything is on fire, I'm at my most comfortable. Self imposed deadlines don't work at all for me because I know the person who made the deadlines is full of shit.

The person making the deadlines isn't full of shit if they have plans after the deadline. Not making the deadline means rescheduling or canceling plans, which probably will piss other people off. It's especially effective to have your life fully scheduled, as you know that even if you miss the deadline, you can't easily pull an all nighter later because you have other commitments. It's all interdependent.

But sure, if you have only one important task and you have an abundant supply of time (e.g., single, young, no children), then yes, it's much easier to fool yourself into thinking you'll have ample time around the deadline. And maybe you'll have it. But that's not an effective strategy for when life gets hard and people depend on you. It's easier to mold your brain to work consistently early on, instead of trying to change your habits abruptly when life gets hard.