00:12 zmike[d]: I'm cooking.
00:12 zmike[d]: https://www.youtube.com/watch?v=QrGrOK8oZG8&
03:41 cubanismo[d]: mhenning[d]: In my youth I used to loudly proclaim that I never write bugs. Recently, I was recalling how I eventually went to see how many bugs I actually wrote, and I'd just finished reading The Mythical Man Month or something, and was feeling pretty provoked by how it boiled engineers down to lines of code/month numbers and all that, and my bug rate per lines of code was almost exactly the
03:41 cubanismo[d]: industry average it listed.
03:41 cubanismo[d]: It was pretty annoying.
03:42 cubanismo[d]: All part of my process of journey through not-selling-out-buying-in.
03:57 sonicadvance1[d]: Just make sure to write as many asserts as bugs, that way you can say that the assert build would have caught it.
03:59 sonicadvance1[d]: `Syscall: modify_ldt stub!`
03:59 sonicadvance1[d]: All the homies love asserts.
04:28 narantakoki: Well you are bugs and toxic snakes anyways, a snake with one brain cell, which would not be a problem, you terror others on their own territory and in their own hotels, and are searched for by bounty head hunters too. I understand you want to assault me same as Muammar Gaddafi by your team of scammers, why else you'd tell my fans that interracial anal bums like angry fuck crocodiles
04:28 narantakoki: harassing me and extorting with their crankbuffalos are my girlfriends, that is otherwise suicidal statement, cause the bums were treated with force already same with their crankbuffalos, they broke their bones and cracked their teeth, it is called a rabbitlip that marks out scammers, extorters , liars , anal courtesans, and other type of penniless thieves. Which was all performed correctly
04:28 narantakoki: and is performed in more brutal contents and degree if same things happen again that those sick people command others and their resources.
04:32 HdkR: https://old.reddit.com/r/OneOrangeBraincell/ is personally my favourite subreddit.
04:50 gfxstrand[d]: cubanismo[d]: Yup. We all write bugs. We all write bad code that has to be replaced later. One of the hardest things sometimes for a junior engineer is when someone comes along and wholesale replaces something they spent months on because they found a better way. I sometimes try to encourage them by telling them just how much of my own code I've deleted because it was trash. IDK if they find that
04:50 gfxstrand[d]: encouraging or terrifying, though. 🤣
04:52 gfxstrand[d]: What makes a good engineer isn't not writing bugs. It's taking the time and care to fix them. And not only fix them but take the time to understand the code (and the hardware, if applicable) and the bug and fix it properly instead of just applying the first change that makes the test pass.
04:57 gfxstrand[d]: And (pardon the soapbox, please) this is why I care so much about commit message and MR descriptions. Just because someone wrote "no functional change" in the commit message doesn't mean they didn't introduce a bug. We're humans. We make mistakes. We miss things. Especially ADHD dyslexic humans like me. When those bugs get discovered, `git bisect` is the best tool we have for figuring out what
04:57 gfxstrand[d]: went wrong. When re-reviewing a patch from a year or even just a month ago and trying to figure out where it went wrong, having the context of what's changed, why it changed, and what the author intended to do is hugely helpful. Otherwise you're just guessing and getting frustrated.
04:59 gfxstrand[d]: We don't just do it for today's reviewer. For them, you can just explain it on IRC or discord. We do it for our future selves. One of the best gifts you can give your future self or the future other person who has to fix your bug is a good commit message.
04:59 gfxstrand[d]: /soapbox
04:59 chikuwad[d]: :birdnotes:
05:43 mhenning[d]: yeah, I also care about commit messages
05:43 mhenning[d]: ideally chesterton's fence isn't an issue because whoever put up the fence wrote down their thought process
05:46 chikuwad[d]: yeah, I've always been very particular with commit messages and having One Commit do One Thing only, much to the annoyance of a friend who sometimes contributed to my project
07:29 asdqueerfromeu[d]: zmike[d]: Doesn't Zink have no bugs though?
10:53 zmike[d]: more important than all the soapboxing above: never drink your own koolaid
10:55 kayliemoony[d]: cubanismo[d]: at some point, you realize all you write is full of bugs
10:55 kayliemoony[d]: and one should pivot from "consciously trying not to write them" to "making it *impossible* to end up writing them"
10:55 kayliemoony[d]: at least imo
10:56 kayliemoony[d]: as i've gotten older and spent more time around large projects i've become more and more inclined to express invariants at compile time, heavily test, etc
12:26 ahuillet[d]: gfxstrand[d]: Ugh, thanks.
15:06 cubanismo[d]: kayliemoony[d]: Yeah, I think somewhere in that process is where you go from "coding" to the beginnings of "software engineering."
15:07 cubanismo[d]: Just coding seems like a lot more fun... until you have to start fixing your own bugs.
15:08 cubanismo[d]: However, today I just wish all that institutionally-prescribed software engineering process would get out of the way and let me finally send out these trivial drm_fourcc.h changes to add the blackwell modifiers.
15:41 gfxstrand[d]: cubanismo[d]: Mood...