|
sarehu posted:What do you experience then? Non-whiteboard exercises? Conversations about the weather?
|
# ? Dec 24, 2015 21:13 |
|
|
# ? Jun 5, 2024 11:54 |
|
I've never had an interview in 10 years at a place I ended up working at that didn't involve a coding test of some kind. The rationale is that most "programmers" can't actually code their way out of a wet paper bag, which seems to make sense based on what I've seen conducting those interviews. Can you remove a node from a linked list in 45 minutes? Congratulations, you're better than 90% of the candidates I've screened.
|
# ? Dec 24, 2015 22:09 |
|
I interviewed a guy recently who spoke really well about concepts. He seemed to know a lot about iOS and casually dropped some decent trivia in conversation. Within five minutes on the white board we realized he couldn't even do basic control flow in a for loop, even with leading comments like "is there some sort of keyword to help you continue execution at the next iteration or break out of the loop entirely?" People will always find ways to surprise you.
|
# ? Dec 24, 2015 23:49 |
|
Just refactor it into a while loop?code:
|
# ? Dec 25, 2015 00:02 |
|
ultramiraculous posted:I interviewed a guy recently who spoke really well about concepts. He seemed to know a lot about iOS and casually dropped some decent trivia in conversation. Within five minutes on the white board we realized he couldn't even do basic control flow in a for loop, even with leading comments like "is there some sort of keyword to help you continue execution at the next iteration or break out of the loop entirely?" I'm going to be interviewing there in a couple of weeks and now I'm anxious that I'm going to accidentally ruin it with a simple/avoidable fuckup like this. Which type of interview was this?
|
# ? Dec 25, 2015 01:59 |
|
Doctor w-rw-rw- posted:
I'm going to guess a job interview.
|
# ? Dec 25, 2015 04:27 |
|
sarehu posted:Just refactor it into a while loop? Where did you get this one?
|
# ? Dec 25, 2015 04:34 |
|
Just a refactoring of a for loop into a while loop...
|
# ? Dec 25, 2015 05:32 |
|
I'm the strlen e: and the behavior when strlen(s) >= SHRT_MAX Soricidus fucked around with this message at 15:21 on Dec 25, 2015 |
# ? Dec 25, 2015 15:18 |
|
Soricidus posted:I'm the strlen wait, it got called SHRT_MAX rather than SHORT_MAX? In god's name why? Why do programmers have this mania for adopting unnecessary and unpredictable abbreviations?
|
# ? Dec 25, 2015 16:49 |
|
Hammerite posted:wait, it got called SHRT_MAX rather than SHORT_MAX? In god's name why? Why do programmers have this mania for adopting unnecessary and unpredictable abbreviations? Vowels are for Communists, FORTRAN and Visual Basic programmers.
|
# ? Dec 25, 2015 17:00 |
|
Hammerite posted:wait, it got called SHRT_MAX rather than SHORT_MAX? In god's name why? Why do programmers have this mania for adopting unnecessary and unpredictable abbreviations? Something something, eight letter identifiers, SHORT_MA, etcetc.
|
# ? Dec 25, 2015 17:00 |
|
Hammerite posted:wait, it got called SHRT_MAX rather than SHORT_MAX? In god's name why? Why do programmers have this mania for adopting unnecessary and unpredictable abbreviations? There was a time, long ago, when our compilers were limited to about 8 chacters for symbols. Of course, that era was long ago and it is highly unlikely a) they were around to pick up that habit, and b) work on a platform where that's still a thing.
|
# ? Dec 25, 2015 20:53 |
|
Hammerite posted:wait, it got called SHRT_MAX rather than SHORT_MAX? In god's name why? Why do programmers have this mania for adopting unnecessary and unpredictable abbreviations? There's a big correlation between people that do this and people that want 80-character column limits.
|
# ? Dec 25, 2015 20:56 |
|
Pixelboy posted:Of course, that era was long ago and it is highly unlikely a) they were around to pick up that habit, and b) work on a platform where that's still a thing. By "them" you mean the people who wrote the original limits.h?
|
# ? Dec 25, 2015 20:56 |
|
sarehu posted:There's a big correlation between people that do this and people that want 80-character column limits. PEP8, PEP8, PEP8, ONE OF US, ONE OF US. (gently caress the 80 character column limit in PEP8)
|
# ? Dec 25, 2015 20:58 |
|
Subjunctive posted:By "them" you mean the people who wrote the original limits.h? On the platform where that was a thing, no less.
|
# ? Dec 25, 2015 21:04 |
|
ratbert90 posted:PEP8, PEP8, PEP8, ONE OF US, ONE OF US. (gently caress the 80 character column limit in PEP8) PEP8 has an 80 char max? That's atrocious.
|
# ? Dec 25, 2015 22:55 |
|
Pep8 character limit is not that bad. It actually helps when you have multiple files open.
|
# ? Dec 25, 2015 23:02 |
|
necrotic posted:PEP8 has an 80 char max? That's atrocious. Everything in pep8 is recommendations, one of the first is to know when to break the rules, but yes, it has a recommendation for 80 characters, or even 79 to leave space for the +- markers in git diffs.
|
# ? Dec 25, 2015 23:03 |
|
I love pep8, and I love breaking that clause of pep8. Breaking that clause so bad, everyday yessir. I also habitually make multiply nested comprehensions hundreds of chars long with conditionals, ternaries and lambdas out the wazoo just to gently caress with my coworkers, pep8 and my future self hell yes I'm a filthy human being.
|
# ? Dec 25, 2015 23:09 |
|
ratbert90 posted:PEP8, PEP8, PEP8, ONE OF US, ONE OF US. (gently caress the 80 character column limit in PEP8) Isn't PEP8 a 79 column max?
|
# ? Dec 25, 2015 23:36 |
|
I actually do like 80 character columns because I work on a small screen a lot.
|
# ? Dec 26, 2015 01:17 |
|
80 char columns means I can comfortably get 3 panes of code side by side, with line numbers and error markers. Plus it makes things look somewhat consistent. Which is nice. # noqa
|
# ? Dec 26, 2015 01:27 |
|
I have two big screens and I still like 80 character columns because I can have two of them side by side at a readable size while leaving the other screen free for other stuff also I have yet to see a very long line of code that was not improved by breaking it up into several short lines of code. or burning it to the ground, salting the earth it stood on, and calling in a priest to drive the lingering spirit of malice back into the depths of hell from whence it came, in the case of one particularly memorable 800+-character java statement that some fuckwit apparently thought was acceptable because we have word wrap now
|
# ? Dec 26, 2015 01:33 |
|
Soricidus posted:word wrap I'm not one to get upset about lines longer than 80 chars, but word wrapping your code is just wrong.
|
# ? Dec 26, 2015 01:36 |
|
http://chrisseaton.com/rubytruffle/pushing-pixels/ this man is a hero
|
# ? Dec 26, 2015 05:46 |
|
Doctor w-rw-rw- posted:I'm going to be interviewing there in a couple of weeks and now I'm anxious that I'm going to accidentally ruin it with a simple/avoidable fuckup like this. At Facebook or Fitbit? This was an architecture-y interview that usually ends with a little coding if it goes well. It's not meant to be difficult but some people turn it into a train wreck.
|
# ? Dec 26, 2015 06:41 |
|
ultramiraculous posted:At Facebook or Fitbit? This was an architecture-y interview that usually ends with a little coding if it goes well. It's not meant to be difficult but some people turn it into a train wreck. Was Fitbit before or after Facebook for you?
|
# ? Dec 26, 2015 09:06 |
|
fritz posted:Something something, eight letter identifiers, SHORT_MA, etcetc. But there's a USHRT_MAX too
|
# ? Dec 26, 2015 15:24 |
|
That came after C wasn't limited to 8, and was chosen to match SHRT_MAX.
|
# ? Dec 26, 2015 16:43 |
|
Malcolm XML posted:http://chrisseaton.com/rubytruffle/pushing-pixels/ On every pixel, dynamically construct a string, then throw it at the compiler with the expectation that after it fails the local scope lookup, it searches the entire namespace for whatever garbage you managed to spell out. This is idiomatic ruby?
|
# ? Dec 26, 2015 17:44 |
|
JawnV6 posted:On every pixel, dynamically construct a string, then throw it at the compiler with the expectation that after it fails the local scope lookup, it searches the entire namespace for whatever garbage you managed to spell out. This is idiomatic ruby? and somehow graal+truffle made it fast code written to be as slow as loving possible
|
# ? Dec 26, 2015 17:55 |
|
JawnV6 posted:On every pixel, dynamically construct a string, then throw it at the compiler with the expectation that after it fails the local scope lookup, it searches the entire namespace for whatever garbage you managed to spell out. This is idiomatic ruby? Not only that, but the guy who wrote that post apparently thinks that such code is fine and it's the runtimes fault if it's slow?
|
# ? Dec 26, 2015 18:38 |
|
Soricidus posted:Not only that, but the guy who wrote that post apparently thinks that such code is fine and it's the runtimes fault if it's slow? I got out of it more like "this is the poo poo we deal with in the common libraries so we make the optimization so work"
|
# ? Dec 26, 2015 18:43 |
|
Soricidus posted:Not only that, but the guy who wrote that post apparently thinks that such code is fine and it's the runtimes fault if it's slow? That's the perspective you have to take as a developer of the runtime if you want to build something that's actually useful. If he actually thought it was fine he would not have put in so many disclaimers.
|
# ? Dec 26, 2015 18:45 |
|
HardDisk posted:Pep8 character limit is not that bad. It actually helps when you have multiple files open. I will always ignore any such line-length "recommendation". I usually set the right-hand margin line to be 120 (and try to avoid hitting it).
|
# ? Dec 26, 2015 19:01 |
|
Volte posted:It's terrible for a language where whitespace is semantically meaningful and long-line wrapping is discouraged via sheer ugliness. Forcing you to break lines and add whitespace in Python is essentially forcing you to change the structure of your program (rather than just the presentation of it). The PEP8 character limit only superficially affects the physical length of the line, and more accurately promotes a low subexpression depth (since nested subexpressions will rapidly approach 80 characters, and can be easily tamed by extracting them into separate variables). I find this restriction extremely limiting, since I consider a named variable to be a fairly weighty presence within a function (if it has a name, it should be important enough to exist on its own, not just because I need to stick it somewhere while I do some landscaping on the subsequence line of code) and would prefer that temporary values remain anonymous when possible and practical. Having worked at places that strictly enforce PEP8 and at places that don't give a gently caress, I can say the code written to 80 chars max was a lot easier to understand, especially for someone new to the code. The Linux kernel does something similar (8 space tabs to promote low subexpression depth) and it works there too. I don't think Python implementations even have the kind of stack saving optimizations that you'd get with a C compiler, but it's still very much worth doing there. It does take some time to get used to writing though. Pie Colony fucked around with this message at 21:46 on Dec 26, 2015 |
# ? Dec 26, 2015 21:34 |
|
JawnV6 posted:On every pixel, dynamically construct a string, then throw it at the compiler with the expectation that after it fails the local scope lookup, it searches the entire namespace for whatever garbage you managed to spell out. This is idiomatic ruby? I'm the implementation of clamp that allocates an array and sorts it.
|
# ? Dec 26, 2015 22:17 |
|
|
# ? Jun 5, 2024 11:54 |
|
Dessert Rose posted:I'm the implementation of clamp that allocates an array and sorts it. Straight outta hakmem.
|
# ? Dec 26, 2015 22:37 |