|
i am going to learn go
|
# ? Aug 4, 2016 20:22 |
|
|
# ? May 24, 2024 15:07 |
|
YeOldeButchere posted:congrats on experiencing the rarest event in the history of the universe i guess thanks! database complained about a duplicate primary key being inserted by a stored proc that adds new records using NEWID()
|
# ? Aug 4, 2016 20:24 |
|
HoboMan posted:also i'm pretty sure i just had a guid collision i would put money on this being caused by you generating guids incorrectly statistically, that's many orders of magnitude more likely than you doing it correctly and getting unlucky
|
# ? Aug 4, 2016 20:24 |
|
hobbesmaster posted:the problem is that programmers that appear to have never heard of composition ruin oop Exactly. I think OOP is fine as long as you "favor composition over inheritance" as they say.
|
# ? Aug 4, 2016 20:24 |
|
YeOldeButchere posted:congrats on experiencing the rarest event in the history of the universe i guess my team when i win at dota
|
# ? Aug 4, 2016 20:25 |
|
HoboMan posted:thanks! Is the NEWID() being copied into a variable or temp table or something that persists outside the scope of the proc execution? That seems a heckuva lot more likely.
|
# ? Aug 4, 2016 20:31 |
|
HoboMan posted:is oop actually bad? i personally hate it but i'm a terrible programmer oop is not innately bad, it's just like one of several tools you may consider using to organize and design things with your brain bad programmers seizing on oop as the be all end all of software design is pretty bad tho -- and then doubling down by solving any problems they encounter with heavier and heavier oop abstractions
|
# ? Aug 4, 2016 20:34 |
|
testing neual networks on an atom processor is kind of lol
|
# ? Aug 4, 2016 20:35 |
|
also yes i am aware i am probably missing something but i want to belive
|
# ? Aug 4, 2016 20:38 |
HoboMan posted:also i'm pretty sure i just had a guid collision Well are you generating 1.25 * 2^64 GUIDs?
|
|
# ? Aug 4, 2016 20:41 |
|
HoboMan posted:is oop actually bad? i personally hate it but i'm a terrible programmer "Object oriented programming is a term that encompasses many ideas. Half of them are obvious. The other half are mistakes." I think that was from Alan Kay?
|
# ? Aug 4, 2016 20:45 |
|
i think the thing people really like about object oriented programming is using the dot operator
|
# ? Aug 4, 2016 20:47 |
|
I can't lie, I really loving love using the dot operator. That's why I prefer the function chain LINQ to the lovely query notation.
|
# ? Aug 4, 2016 20:50 |
|
Bloody posted:testing neual networks on an atom processor is kind of lol why? training them on one would be a waste of time for anything bigger than a toy model, but i can see one that's already been trained being used there, so you'd want to see how it performs
|
# ? Aug 4, 2016 20:54 |
|
MALE SHOEGAZE posted:i think the thing people really like about object oriented programming is using the dot operator i'd like to just dot operator chain every step in my entire program please
|
# ? Aug 4, 2016 21:00 |
|
i'm sure boost spirit can make that happen
|
# ? Aug 4, 2016 21:18 |
|
Finster Dexter posted:I can't lie, I really loving love using the dot operator. That's why I prefer the function chain LINQ to the lovely query notation. if you do query notation right and squint, it's just like F#
|
# ? Aug 4, 2016 21:29 |
|
Finster Dexter posted:I can't lie, I really loving love using the dot operator. That's why I prefer the function chain LINQ to the lovely query notation. there's a function in my codebase that is entirely comprised of a 30 line linq statement using the query syntax that i just refuse to even try to read
|
# ? Aug 4, 2016 22:17 |
|
that reminds me of the notorious 12 lines statement composed entirely of ternary operators (multiple per line) in a 16,000 lines .cpp file that was computing the value of a global state enum used in hundreds if not thousands of places i've seen in a past project. you'd be debugging something and anxiety would slowly build up as you moved up the callstack and got closer and closer to login.cpp because you knew that if you tracked down the issue to that file, your options were to gently caress with it and break a million things or apply some band-aid solution somewhere below in the callstack anyone with half a brain chose the latter option because loving with that kind of poo poo is the first step to getting blamed when things go south
|
# ? Aug 4, 2016 22:32 |
|
linq query notation is bad because it just looks like a sql markov chain. embrace the dot
|
# ? Aug 4, 2016 22:35 |
|
hobbesmaster posted:i'm sure boost spirit can make that happen can't overload the dot operator in c++. stroustrup wants to make it happen but like with unified call syntax, even he can't come up with an unambiguous specification
|
# ? Aug 4, 2016 22:40 |
|
i've always figured that the problem with overloading the . operator in c++ is that sooner or later you're gonna have to mess with whatever member variables you have in your class. and that's going to be done through either this->_butt or (*this)._butt, but either way you'll end up with some infinite recursion somewhere
|
# ? Aug 4, 2016 22:55 |
|
AWWNAW posted:three times in as many months an internal recruiter has invited me to interview and then backed off when figuring out i dropped out of college for a job i'm kinda worried about this too. i dropped out of my final year of my cs program to get a job in france, and though i have a masters degree now i never actually got a bachelors
|
# ? Aug 4, 2016 23:02 |
|
Shaman Linavi posted:my algorithms class was taught by a dude that used kinesthetic learning our networking professor taught protocol by passing out sheets of paper to each student that said (a) you are number N, (b) you have messages A, B, C, and (c) you must deliver these messages to X, Y, Z. then after everyone read their sheet, he said "go!"
|
# ? Aug 4, 2016 23:08 |
|
YeOldeButchere posted:i've always figured that the problem with overloading the . operator in c++ is that sooner or later you're gonna have to mess with whatever member variables you have in your class. and that's going to be done through either this->_butt or (*this)._butt, but either way you'll end up with some infinite recursion somewhere no that's not a problem, this is always a raw pointer and this->butt_ (underscore in front is resevred, ugh) never calls any user defined code
|
# ? Aug 4, 2016 23:14 |
|
Condiv posted:i'm kinda worried about this too. i dropped out of my final year of my cs program to get a job in france, and though i have a masters degree now i never actually got a bachelors if a background check came back confirming a master's but not confirming the bachelor's HR would probably blame the background check company
|
# ? Aug 4, 2016 23:59 |
|
so loving future posted:oop is not innately bad, it's just like one of several tools you may consider using to organize and design things with your brain mainstream OOP languages heavily discourage you from structuring your programs like the above so OOP might be fundamentally bad
|
# ? Aug 5, 2016 00:00 |
|
Mr Dog posted:This also doesn't make much sense to me, I'm afraid. linked lists are ridiculously bad for performance due to poor use of the CPU cache your default container should be a vector
|
# ? Aug 5, 2016 00:02 |
|
YeOldeButchere posted:why? training them on one would be a waste of time for anything bigger than a toy model, but i can see one that's already been trained being used there, so you'd want to see how it performs it is a toy model
|
# ? Aug 5, 2016 00:04 |
|
i have been going through a learn about neural networks phase. it is cool. my computer can identify mnist handwriting digits p deecently
|
# ? Aug 5, 2016 00:07 |
|
here's a good article about why you should heavily prefer vectors over linked lists https://kjellkod.wordpress.com/2012/02/25/why-you-should-never-ever-ever-use-linked-list-in-your-code-again/
|
# ? Aug 5, 2016 00:09 |
|
that's a lot of words for "hardware engineers are lazy idiot fuckers that prefer to copy and paste existing designs and add layers of cache instead of giving us processors that actually fit the ram+register model described in reference material" why haven't you morons cranked up cpu frequency to 100GHz yet????
|
# ? Aug 5, 2016 00:15 |
|
sometimes it fucks up but sometimes it works well tbf i think that 9 looks like a 4 so i will take this subpar performance and call it good and continue iterating on new and/or different poo poo
|
# ? Aug 5, 2016 00:17 |
|
also math.net numerics is good and performant
|
# ? Aug 5, 2016 00:18 |
|
Bloody posted:i have been going through a learn about neural networks phase. it is cool. my computer can identify mnist handwriting digits p deecently what's that neural network tutorial page that starts with that?
|
# ? Aug 5, 2016 00:19 |
|
i would tell you but its at best not very good so
|
# ? Aug 5, 2016 00:22 |
|
Bloody posted:i would tell you but its at best not very good so oh word, got any better ones?
|
# ? Aug 5, 2016 00:25 |
|
mnist is like the hello world of nn these days. i'm pretty sure this is what the tensorflow tutorial does, for example, but it's been a while since i've looked at that
|
# ? Aug 5, 2016 00:27 |
|
fritz posted:ive barely touched a maven but given a choice between xml and makefile syntax xml ftw
|
# ? Aug 5, 2016 00:36 |
|
|
# ? May 24, 2024 15:07 |
hackbunny posted:can't overload the dot operator in c++. stroustrup wants to make it happen but like with unified call syntax, even he can't come up with an unambiguous specification This is probably going to be fairly strong evidence that I belong in this thread, but I came up with a use for operator.() the other day that wasn't a smart reference or whatever. I was thinking it would be neat to make a Cached<T> template that would overload operator.() to access a cache of values for that function. So like you could do this: C++ code:
|
|
# ? Aug 5, 2016 00:41 |