|
Man I make so much money writing JavaScript and Ruby and this thread has me convinced I'm a simpleton. I've been doing this job for 15 years and there's no way I could ever pass a whiteboard interview.
|
# ? Jan 28, 2020 06:17 |
|
|
# ? May 10, 2024 20:09 |
|
https://twitter.com/ryxcommar/status/1221891444627181568
|
# ? Jan 28, 2020 06:20 |
|
He lost me at the emphasized best solution to Fizzbuzz.
|
# ? Jan 28, 2020 11:31 |
|
prom candy posted:Man I make so much money writing JavaScript and Ruby and this thread has me convinced I'm a simpleton. I've been doing this job for 15 years and there's no way I could ever pass a whiteboard interview. i was hoping i wouldnt feel this way after 15 years. dangit.
|
# ? Jan 28, 2020 13:12 |
|
Perhaps it's an indication that whiteboard tests aren't really useful for determining whether someone's going to be good at their job. "Ah, you know how to build a red/black tree from scratch! Now, please configure webpack for us."
|
# ? Jan 28, 2020 13:54 |
|
Sagacity posted:Perhaps it's an indication that whiteboard tests aren't really useful for determining whether someone's going to be good at their job. They are useful for determining whether someone is going to be good at their job. They aren't very useful for telling if someone won't be good at their job. They emphasize false negatives over false positives.
|
# ? Jan 28, 2020 14:06 |
|
Yeah, configuring webpack is easily twice as difficult as implementing red/black trees!
|
# ? Jan 28, 2020 14:06 |
|
Jose Valasquez posted:They are useful for determining whether someone is going to be good at their job. They aren't very useful for telling if someone won't be good at their job. They emphasize false negatives over false positives. With the corollary they they're really only effective if you tailor the test to the type of work you do rather than a pet algorithm problem. Guess which one most companies do...
|
# ? Jan 28, 2020 14:40 |
|
Blinkz0rz posted:With the corollary they they're really only effective if you tailor the test to the type of work you do rather than a pet algorithm problem. Guess which one most companies do... you mean you don't use the liu hui pi algorithm daily when writing css?
|
# ? Jan 28, 2020 14:44 |
|
Further, there's a deeper issue with the way many pet algo problems are delivered and judged, namely that they select for folks who either memorize specific implementations or spend time preparing for the interview by cramming Cracking the Coding Interview. Interviewing well requires a lot of preparation that most folks who get pulled into the process don't have time or interest in doing.
|
# ? Jan 28, 2020 14:53 |
|
Blinkz0rz posted:Interviewing well requires a lot of preparation that most folks who get pulled into the process don't have time or interest in doing.
|
# ? Jan 28, 2020 16:05 |
|
I think it's time this series was reposted: https://aphyr.com/posts/340-reversing-the-technical-interview part 2 is my favorite: https://aphyr.com/posts/341-hexing-the-technical-interview part 3: https://aphyr.com/posts/342-typing-the-technical-interview
|
# ? Jan 28, 2020 18:09 |
|
The Fool posted:I think it's time this series was reposted: https://aphyr.com/posts/340-reversing-the-technical-interview Highly pro-click.
|
# ? Jan 29, 2020 00:50 |
|
prom candy posted:Man I make so much money writing JavaScript and Ruby and this thread has me convinced I'm a simpleton. I've been doing this job for 15 years and there's no way I could ever pass a whiteboard interview. I feel the same, and it doesn't help my impostor syndrome at all. But my current boss says he's happy with me and I'm happy where I'm at compensation wise, so no reason to change anything up if I can avoid it.
|
# ? Jan 29, 2020 00:59 |
|
The Fool posted:I think it's time this series was reposted: https://aphyr.com/posts/340-reversing-the-technical-interview I both love this and don't understand it very well. Knowing Haskell might help.
|
# ? Jan 29, 2020 12:31 |
|
Can confirm they do get funnier the more you understand them.
|
# ? Jan 29, 2020 15:40 |
|
Good time to remind people that his twitter account is NSFW, but is also pro-click if you’re into that.
|
# ? Jan 29, 2020 16:07 |
|
lifg posted:Good time to remind people that his twitter account is NSFW, but is also pro-click if you’re into that. I kind of appreciate the danger of scrolling my timeline with a 5% chance of suddenly, buck naked and extremely jacked dude tied up with a ball gag. Keeps things interesting.
|
# ? Jan 29, 2020 16:21 |
|
Naar posted:Can confirm they do get funnier the more you understand them. I spent too much time last night learning clojure just to understand it better, it does indeed.
|
# ? Jan 29, 2020 16:48 |
|
Today I hit my peak developer. I merged a PR, it got a weird error which I fixed on master and pushed straight to the acceptance environment where it destroyed the database. Then I noticed it was 5PM and I went home. I'll solve it on Monday.
|
# ? Jan 31, 2020 20:11 |
|
Seems weird for a place to be in a position of having required PRs but then also still have the ability to push to master and deploy without oversight.
|
# ? Jan 31, 2020 20:22 |
|
PRs are not required but agreed upon, the push was with another developer present but all that makes not for a nice narrative.
|
# ? Jan 31, 2020 20:41 |
|
Xik posted:Seems weird for a place to be in a position of having required PRs but then also still have the ability to push to master and deploy without oversight.
|
# ? Jan 31, 2020 20:41 |
|
Just lol if you don't practice the ancient technique of "push and pray".
|
# ? Jan 31, 2020 21:18 |
|
Yeet to master.
|
# ? Jan 31, 2020 21:35 |
|
Cuntpunch posted:He lost me at the emphasized best solution to Fizzbuzz. Yeah WTF. A coworker and I will sometimes poo poo out code like that and then delete because we know nobody else can maintain it... and we barely could either.
|
# ? Jan 31, 2020 21:40 |
|
What's the point of code:
|
# ? Jan 31, 2020 21:46 |
|
Pile Of Garbage posted:Yeet to master. I made this my push message just this morning. This is some psychic poo poo.
|
# ? Jan 31, 2020 22:24 |
|
Pile Of Garbage posted:Yeet to master. A good phrase.
|
# ? Feb 1, 2020 04:24 |
|
A coworker submitted a pull-request that adds a new file to our project. This file is structured similarly to one or more other files in the project. This is C code posted:typedef u8 (*ThisIsAFunctionPointer)(u8 InParamLen, u8* const InParams, u8 OutBuffCapacity, u8 * const OutBuff); I commented that he should make his functions const-correct (don't take in-parameters by non-const reference, don't have declarations with top-level const). My coworker replied that this change should be deferred to some other time, since making it const-correct would make this new file inconsistent with the existing files of the project. Odds are that when this gets merged, it's going to stay this way forever because no one is going to have time to fix stuff like this. I feel upset over this. I feel ok with letting the old code remain as-is and preventing new code from repeating the code-smells that the old code has. Got any tips for understanding when consistency does or does not matter in projects?
|
# ? Feb 3, 2020 05:30 |
|
Tell them they need to fix all the other ones too. Anyway, assuming that isn't impossible and isn't a multi-day job, it's what I'd do. The old ones weren't going to be touched until the someday that is today.
|
# ? Feb 3, 2020 05:36 |
|
taqueso posted:Tell them they need to fix all the other ones too. It wouldn't take a lot of time. One day or less. But this PR was submitted so late in the sprint, that I wouldn't feel good about increasing the scope of the ticket.
|
# ? Feb 3, 2020 06:15 |
|
Faith For Two posted:I feel ok with letting the old code remain as-is and preventing new code from repeating the code-smells that the old code has. Got any tips for understanding when consistency does or does not matter in projects? "Consistency" is a pretty bad reason to refrain from making improvements, imo. Standards evolve and improvements shouldn't be avoided simply because everything can't be improved all at once. You could always file a follow-up ticket to bring the rest of the code up to the new standard. Of course, if you all never get enough time to take on work like that, you have additional problems here.
|
# ? Feb 3, 2020 06:46 |
|
"I sort of agree that it's usually better to go with consistently, but this is some technical debt that's been in the firing line for a while. It would be good to start committing what we want it to look like, at least then whenever someone touches the old files we can point to this one and say "if you have to touch them make it look like this instead"."
|
# ? Feb 3, 2020 06:54 |
|
Faith For Two posted:the sprint
|
# ? Feb 3, 2020 06:57 |
|
Const-correctness is actually quite frustrating if a large amount of established code is doing it wrong: you either find yourself editing a massive amount of code or const_casting all over the place. The deep problem is that C gets the default wrong.
|
# ? Feb 3, 2020 07:22 |
|
Faith For Two posted:A coworker submitted a pull-request that adds a new file to our project. This file is structured similarly to one or more other files in the project. quote:I feel upset over this. I feel ok with letting the old code remain as-is and preventing new code from repeating the code-smells that the old code has. Got any tips for understanding when consistency does or does not matter in projects? Faith For Two posted:It wouldn't take a lot of time. One day or less. But this PR was submitted so late in the sprint, that I wouldn't feel good about increasing the scope of the ticket. I like to have an explicit reference to these code quality attributes https://iso25000.com/index.php/en/iso-25000-standards/iso-25010 in my team's Definition of Done, and to explain to the company why these are important (less maintenance needed, fewer production incidents and bugs, and so on). In my experience, usually companies are willing to accept something like this, at least if the explanation comes from an experienced developer who's been in the company for multiple years. In any case, the balance between good code and perfectionism is a difficult one, and so is the balance between writing quality code and delivering fast. The DoD and a list of quality attributes aren't going to fix that entirely, since those will always be subjective measurements. I found I got a feeling for what's right after a couple years of doing this job. Everyone in the team will have a slightly different idea, but together you can find a compromise.
|
# ? Feb 3, 2020 08:12 |
|
Faith For Two posted:It wouldn't take a lot of time. One day or less. But this PR was submitted so late in the sprint, that I wouldn't feel good about increasing the scope of the ticket. It's just process busywork, but if it works for you then
|
# ? Feb 3, 2020 09:03 |
|
My coworker did only a couple of changes I recommended. The rest, he put in a new ticket. I went ahead and approved the PR because I don’t want to argue about stuff.
|
# ? Feb 3, 2020 15:49 |
|
|
# ? May 10, 2024 20:09 |
|
taqueso posted:Tell them they need to fix all the other ones too.
|
# ? Feb 3, 2020 18:32 |