|
proper IDEs are great for undoing the work of the loving idiots that made a thing.
|
# ? Jan 3, 2014 01:49 |
|
|
# ? Jun 1, 2024 06:06 |
|
USSMICHELLEBACHMAN posted:i think he just didnt understand how to use git dont use gradle. theres no reason not to use maven for java. its the best built tool for any language anywhere. also pls tell me you weren't checking in the eclipse workspace metadata.
|
# ? Jan 3, 2014 02:14 |
|
Shaggar posted:dont use gradle. theres no reason not to use maven for java. its the best built tool for any language anywhere. also pls tell me you weren't checking in the eclipse workspace metadata. no sorry Shaggar the best tool is a bespoke artisanal huge complex web of Ant build tasks yesssssss yess the xml configuration give it to me
|
# ? Jan 3, 2014 02:39 |
|
ant is so bad. i sometimes have to fix other peoples code from years ago and lots of the projects have ant build tasks in their pom. its awful.
|
# ? Jan 3, 2014 02:40 |
|
Shaggar posted:i sometimes have to fix other peoples code teehee
|
# ? Jan 3, 2014 02:44 |
|
Symbolic Butt posted:I'm really glad academia is moving to python because most of the code I got to work with till now wasn't really mind-numbingly bad. it's mostly silly like "hey what's this about modules? let me put the whole project in this hugeass file" i dont see how python fixes that
|
# ? Jan 3, 2014 04:21 |
|
Shaggar posted:ant is so bad. i sometimes have to fix other peoples code from years ago and lots of the projects have ant build tasks in their pom. its awful. maven is as big a leap forward over ant as ant was over make
|
# ? Jan 3, 2014 04:25 |
|
Bloody posted:i dont see how python fixes that s. butte stares at variables named x,j,x_r_sin_cos_p,var with utterly no type information and then inserts gun into mouth edit: x_r_cos_sin_p will be reused a three dozen times in the hugeass file. 2 of those times it will be a string Brain Candy fucked around with this message at 04:47 on Jan 3, 2014 |
# ? Jan 3, 2014 04:33 |
|
Brain Candy posted:s. butte stares at variables named x,j,x_r_sin_cos_p,var with utterly no type information and then inserts gun into mouth well no you see duck typing therefore
|
# ? Jan 3, 2014 04:36 |
|
Bloody posted:i dont see how python fixes that yeah it doesn't force anyone to break your code in nice chunks or good practices when naming variables but hey it's better than the horror show of deciphering the control flow in fortran projects. I'm not saying it's the best but drat if it ain't better already. and tbh I see a lot of complaints about variable naming in scientific programming "I MEAN WHAT THE gently caress IS THIS DY BULLSHIT"[*] and sometimes it's justifiable but most of the time is just a lack of domain knowledge. scientists (at least the physicists that I know) have their own idioms, you gotta learn them. have those books with the formulas you're debugging near you. [*] this was a thing that some coworker blurted out to me once
|
# ? Jan 3, 2014 05:37 |
|
I like duck typing but it's not really a thing with the code I'm talking about. numpy strips a lot of the dynamic typing aspect from python.
|
# ? Jan 3, 2014 05:40 |
|
Symbolic Butt posted:and tbh I see a lot of complaints about variable naming in scientific programming "I MEAN WHAT THE gently caress IS THIS DY BULLSHIT"[*] and sometimes it's justifiable but most of the time is just a lack of domain knowledge. scientists (at least the physicists that I know) have their own idioms, you gotta learn them. have those books with the formulas you're debugging near you. sure, if the reference algorithm names things x,y,p,q guess what the variable names are going to be? and what you'd write as Δy or turns into crap no matter what how you name it. that isn't the problem with x. the problem is when x is used as the permanent state vector which is filled with fp in some order that isn't documented anywhere. math familiarity tricks people into thinking of programming variables as math variables i.e. short, undescriptive, names because the notation is designed for people writing things on dead trees not that this is unique to scientists, but scientist/engineer code awful because there is utterly no thought to maintenance or readability for other people. "if it works, it is good"
|
# ? Jan 3, 2014 06:07 |
|
oh well at least it's not the kind of code where you get unnecessary abstraction because the cs guy who wrote it was trying to show off how smart he is by creating recursive classes cs people can suck my defaultdict is what I'm saying
|
# ? Jan 3, 2014 06:24 |
|
work experience i have involving python: it's a skin layer over gnuradio for the sole purpose of making gnuradio less usable
|
# ? Jan 3, 2014 06:28 |
|
is there a good beginner maven tutorial out there? it obviously sounds super great, but i'd like to see some examples etc.
|
# ? Jan 3, 2014 18:20 |
|
The Leck posted:is there a good beginner maven tutorial out there? it obviously sounds super great, but i'd like to see some examples etc. http://maven.apache.org/guides/getting-started/ or http://maven.apache.org/guides/getting-started/maven-in-five-minutes.html bam you might not immediately realize how great it can be unless you've spent some quality time trying to get your CLASSPATH set just right. half of the benefit is about classpaths and dependencies; the other half is that maven really wants you to lay out your files and directories in the structure they like. it can be annoying, but being forced into a standard directory structure eliminates a lot of problems prefect fucked around with this message at 18:24 on Jan 3, 2014 |
# ? Jan 3, 2014 18:22 |
|
Shaggar posted:ant is so bad. ant is better than batch files, looks at $30 million software project colossal loving mess, a batch file (not command file) that only works on a desktop, spawns extra prompts that sleep and poll for files to determine when certain stages are complete.
|
# ? Jan 3, 2014 18:39 |
|
there was a bad question here but i answered it myself~
Bloody fucked around with this message at 20:52 on Jan 3, 2014 |
# ? Jan 3, 2014 18:41 |
|
Brain Candy posted:math familiarity tricks people into thinking of programming variables as math variables i.e. short, undescriptive, names because the notation is designed for people writing things on dead trees this is so loving true. if they took the 10s to figure out what x,y,p,q, actually mean even in their general case (i.e., understood the problem) this would be a whole lot easier s.butt is right about domain knowledge but the problem with physics is it's all so loving niche that basically nobody is familiar with your specific dumb problem and specific combinations of retarded letters that differ slightly from the americans/euros workin on the same thing or even the prof down the hall its all hugely incomprehensible and inconsistent and scientists are idiots oh yeah even for basic stuff like integration/differentiation they manage to gently caress it up and use variables that make LITERALLY NO SENSE ever. if they do their math in a slightly different idiomatic way (idk like using xdot instead of d/dx ) then theres infinitely differnet ways they can garble up the math into even more confusing code oh even worse is idiot scientists like this with some CS knowledge trying to show off in C/C++
|
# ? Jan 3, 2014 18:58 |
|
prefect posted:http://maven.apache.org/guides/getting-started/ this man is correct the official docs are pretty good when you need more than a tutorial consider buying the sonatype books
|
# ? Jan 3, 2014 20:52 |
|
Eclipse is crappy and full of bugs. That said it is many times better than using a text editor. The only catch is that you must never use the source control plugins to actually commit code as they do idiot things by default. Use eclipse but do source control from the cli
|
# ? Jan 4, 2014 00:21 |
|
yeah the svn plugin is ok for basic tasks like commits and stuff, but if you want to do anything more advanced its not good
|
# ? Jan 4, 2014 00:29 |
|
bobbilljim posted:Eclipse is crappy and full of bugs. That said it is many times better than using a text editor. The only catch is that you must never use the source control plugins to actually commit code as they do idiot things by default. Use eclipse but do source control from the cli for svn don't use subversive, use subclipse it use to be poo poo, but now branches, merges, reverts, whatever, are v. nice
|
# ? Jan 4, 2014 01:27 |
|
Shaggar posted:also pls tell me you weren't checking in the eclipse workspace metadata. that's exactly what our problem was.
|
# ? Jan 4, 2014 02:47 |
|
one of the benefits of using source control inside of eclipse instead of outside is it automatically ignores eclipse stuff. the root of ur project folder in your repo should look like this: src/ pom.xml
|
# ? Jan 4, 2014 03:21 |
|
yeah that's what we ended up doing. i'm very stubborn and like doing things through the command line, so that's basically what hosed us up. we ended up adding the eclipse metadata to gitignore but without knowing much about what we're doing it's hard to know what to ignore
|
# ? Jan 4, 2014 03:23 |
|
speaking of gitignore, i accidentally just committed 1tb of movie files
|
# ? Jan 4, 2014 03:25 |
|
lmao
|
# ? Jan 4, 2014 03:26 |
|
USSMICHELLEBACHMAN posted:speaking of gitignore, i accidentally just committed 1tb of movie files lol
|
# ? Jan 4, 2014 07:36 |
|
Nice
|
# ? Jan 4, 2014 09:21 |
|
USSMICHELLEBACHMAN posted:speaking of gitignore, i accidentally just committed 1tb of movie files Nice!
|
# ? Jan 4, 2014 09:26 |
Mr SuperAwesome posted:the work of the loving idiots that made a thing. u rang?
|
|
# ? Jan 4, 2014 09:31 |
|
prefect posted:http://maven.apache.org/guides/getting-started/ thanks, i suppose that was the most obvious possible answer to this question. i'm all for a standard structure, so i think this will work out just fine.
|
# ? Jan 5, 2014 00:37 |
|
USSMICHELLEBACHMAN posted:speaking of gitignore, i accidentally just committed 1tb of movie files rebase it to a few hundred commits ago and then force push
|
# ? Jan 5, 2014 04:15 |
|
i just made a thing with spring web flow, and it was really easy and neat. is there something like web flow for other languages?
|
# ? Jan 6, 2014 04:39 |
|
Sharktopus posted:rebase it to a few hundred commits ago and then force push fortunately i wasnt able to push because my attempt to push crashed due to memory issues.
|
# ? Jan 6, 2014 10:25 |
|
prefect posted:site's down
|
# ? Jan 6, 2014 10:41 |
|
i just spent an hour trying to figure out why this windows call that finds spotify's window title for sending to my arduino was loving up half the time i have atmel studio running and it was finding the SpotifyController project instead of the actual spotify window and sending that to the AVR but now i got a sweet program running that shows me the artist/song playing on spotify
|
# ? Jan 7, 2014 05:59 |
|
imo the award for most retarded build system goes to lucene/solr. its not one of those appallingly bad organically grown job security generators or anything, just ant. and maven. there's an ant build, and also a maven build
|
# ? Jan 7, 2014 07:16 |
|
|
# ? Jun 1, 2024 06:06 |
|
Nomnom Cookie posted:imo the award for most retarded build system goes to lucene/solr. its not one of those appallingly bad organically grown job security generators or anything, just ant. and maven. there's an ant build, and also a maven build is it a historical thing, where they're slowly transitioning from one to the other?
|
# ? Jan 7, 2014 12:24 |