awful programming: feature flags and suffering

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > YOSPOS > awful programming: feature flags and suffering

susan b buffering: Nov 14, 2016

echinopsis posted:

further to a discuss before about python venv does that make it relatively easy rherefor to make a �portable� program, ie put it all on a usb, open the venv script and then you�re good to go despite python not being on the computer?

i want to lol write a program that imitates keypresses because the software i want to manipulate (i can�t get access to the database) runs on keypresses alone to get the data out of the software

no because it does symlink some of the files

installing python on the usb drive might work or use a tool that creates a self contained executable from your code

# ¿ May 28, 2019 02:41

Adbot: ADBOT LOVES YOU

# ¿ Apr 27, 2024 02:18

susan b buffering: Nov 14, 2016

echinopsis posted:

hmm i�ll see. database is local and that also sounds above my pay grade

have you examined the db file at all? it could be something standard like SQLite or leveldb

# ¿ May 28, 2019 05:16

susan b buffering: Nov 14, 2016

mystes posted:

Use go. It doesn't have stack traces.

lol

# ¿ Oct 24, 2019 05:53

susan b buffering: Nov 14, 2016

eschaton posted:

wait, jira customizations don�t create their own related tables, they just dump everything into one?

JFC

surprised they don�t just dump everything into an EAV table

susan b buffering fucked around with this message at 13:43 on Jun 22, 2020

# ¿ Jun 22, 2020 13:36

susan b buffering: Nov 14, 2016

Carthag Tuek posted:

wait btw, if there are no cached elements, wont that code just create an empty list and throw an excexption when you try to return the first element?

that�s what the conditional is for. empty lists evaluate as false

# ¿ Oct 23, 2020 21:02

susan b buffering: Nov 14, 2016

Phobeste posted:

ah, the holy ternarity

# ¿ Nov 17, 2020 19:48

susan b buffering: Nov 14, 2016

i have an hour and a half long panel interview for an entry-level position at a mid-sized company tomorrow. the email i got about it said i should look into �Franklin magic squares� for the algorithms portion of the interview. i am extremely ready to get it over with.

# ¿ Feb 21, 2021 23:00

susan b buffering: Nov 14, 2016

jesus WEP posted:

our team name is TrentReznor because we�re nine intj males

# ¿ Mar 19, 2021 16:08

susan b buffering: Nov 14, 2016

jesus WEP posted:

doing it in the where clause is the Oracle way which tells you all you need to know

running into (+) in oracle sql for the first time really ruined my day.

# ¿ May 21, 2021 14:13

susan b buffering: Nov 14, 2016

Pile Of Garbage posted:

https://twitter.com/lukaseder/status/1410533024945561601

i�m in this photo and i don�t like it

# ¿ Jul 2, 2021 14:34

susan b buffering: Nov 14, 2016

dont make me tap the old thread title

# ¿ Nov 7, 2021 07:55

susan b buffering: Nov 14, 2016

Shaggar posted:

C# code:

public abstract class Orifice 
{
	public abstract bool AcceptsInput {get;}
	public abstract bool ProducesOutput {get;}
}

i remember this james mickens talk

# ¿ Nov 18, 2021 00:21

susan b buffering: Nov 14, 2016

fritz posted:

fib backoff is when you have to clarify parts of your resume

# ¿ Nov 19, 2021 18:49

susan b buffering: Nov 14, 2016

crazypenguin posted:

also.... "commits" might be misleading to people who haven't tried it. It's really "pull request titles" since most of these projects will also squash and merge

this is how we use conventional commits at work and it�s fine. using them any other way sounds like it�d be pretty annoying.

# ¿ Nov 30, 2021 07:38

susan b buffering: Nov 14, 2016

pokeyman posted:

idk "right-click in your browser, choose inspect element, click on the bottom text field, now you have frontend javascript" is a pretty darn available and quick repl

oops it's minified and sucks to read!

# ¿ Mar 28, 2022 06:38

susan b buffering: Nov 14, 2016

Corla Plankun posted:

Can any of y'all recommend an article on working with contractors? My job now is a tiny startup and they've hired a few to fill some immediate slots as we grow and I feel like I'm not doing a good job with them: namely I'm spending a lot of time treating them as direct reports/teaching them stuff and I am realizing that this is sort of a waste since they don't actually work here. But probably I can't just tell them to gently caress off and figure it out on their own unless i say it really politely and give them clear enough guardrails that they can actually solve the problem and not just spin their wheels.

I wouldn't mind a little bit of wasted time normally but i have a shitload on my plate right now and I really need to trim the low-impact stuff like spontaneous one-on-ones with contractors. Especially at times where i've told the whole dang team I am going to be focused and busy and please leave me alone.

do you have enough people to disperse the responsibility a bit? like "ok james 1099 for this ticket you'll be working on james 1040 will be who you go to for questions/peer-programming."

# ¿ Sep 9, 2022 01:40

susan b buffering: Nov 14, 2016

Plorkyeran posted:

stop doing stupid poo poo with local-name() and just use namespaces correctly: root.xpath("/mmd:metadata/mmd:artist-list/mmd:artist/@type-id", namespaces={'mmd': 'http://musicbrainz.org/ns/mmd-2.0#'})

i just use picard

# ¿ Sep 9, 2022 01:41

susan b buffering: Nov 14, 2016

Corla Plankun posted:

kinda, but so far that hasn't gone great. I think the contractor is just a pretty high maintenance dude whose workstyle is just rushing from perceived blocker to perceived blocker without ever making a coherent plan. It's not great but it is probably good practice to try to figure out what i can do with him.

oh there's a new-ish hire on my team that's like that, kinda sucks to deal with and i don't have great advice. he's slowly gotten more independent but he still pings me like 2-3 times a week despite me not working on that product anymore lol.

# ¿ Sep 9, 2022 01:54

susan b buffering: Nov 14, 2016

Jabor posted:

Yes, but also that you shouldn't be testing behaviours that aren't relevant to the contract of the entry point.

The usual way to mess this up is to have mocks of a dependency, and then have your test involve calling an entry point and then verifying that specific methods on the mock are called. If your implementation changes, now you need to change the test to check for the new sequence of calls on the mock, and your test hasn't actually achieved anything in terms of making sure your new implementation hasn't caused any regressions.

this is a huge problem in the .net project i'm on. like, this is my first time dealing with moq and it was immediately obvious when starting that the original devs were overzealous in its usage. writing / updating tests like that is a huge chore with seemingly no benefit.

at least it doesn't appear we mock anything we don't own, but i think going forward i'm gonna be ripping out the "verify this internal method is called this specific way" poo poo that's peppered throughout our unit tests.

# ¿ Sep 10, 2022 08:13

susan b buffering: Nov 14, 2016

MrMoo posted:

not kidding, and sure, this makes complete sense 🤷‍♀️

# ¿ Sep 16, 2022 04:13

susan b buffering: Nov 14, 2016

i learned about the debugger statement when i had to do some angular development a few months ago and it seemed better than setting breakpoints in devtools at least

# ¿ Sep 21, 2022 23:31

susan b buffering: Nov 14, 2016

we use angular at work and i hate it so much. we're also like 6 major releases out of date which we'll be tackling in the coming months

# ¿ Jan 30, 2023 09:06

susan b buffering: Nov 14, 2016

Powerful Two-Hander posted:

lol is that as much of a nightmare of deprecated/renamed features and functions as I imagine it to be?

from what i've seen it doesn't look it'll be too bad in that regard. i'm sure there will be some headaches but the apis appear to be pretty stable.

# ¿ Jan 30, 2023 16:47

susan b buffering: Nov 14, 2016

Poopernickel posted:

a post from our CI/CD's principal engineer (supporting a dev-team of ~400 people):

this guy sounds like he'd get along with the guy on our team that wants to use puppeteer for api tests

# ¿ Feb 3, 2023 23:59

susan b buffering: Nov 14, 2016

Achmed Jones posted:

i for real don't really even understand how jira could even BE bad

you go there, there's a list of things to do. you either assign one to yourself or already have it assigned, you drag it to in progress, you do thing, you drag it to in review, you drag it to in-ua or whatever, when it gets deployed it gets moved to "deployed"

what are people doing that fucks that up? someone mentioned not being able to drag to a column and that's pretty bad. but it's definitely not something that happens without somebody working to screw ir up

like it can be slow, ok (though it's always been fine for me when not self-hosted), the query language is annoying, whatever. but none of that is really stuff that matters really.

it sucks to use op

# ¿ Mar 1, 2023 02:14

susan b buffering: Nov 14, 2016

tef posted:

they aren't tools agnostic, dogshit processes pick the tools that enable the dick hammering

# ¿ Mar 1, 2023 03:13

susan b buffering: Nov 14, 2016

FlapYoJacks posted:

One of my jobs had a �release process� that involved taking a bunch of random git commits from 9 of the 2,000+ random branches and combining them manually into a tarball with no documentation as to what branches were used or commit hashes.

I left after 6 months of that bullshit.

Edit: the company name is GreenLots. They were acquired by Shell several years ago and provide the backend to Electrify America DC fast chargers.

lol awesome

# ¿ Mar 16, 2023 20:51

susan b buffering: Nov 14, 2016

pokeyman posted:

what's the difference? or is this a language "without a compiler" so you're using a linter

the op is using python op

# ¿ Apr 15, 2023 06:29

susan b buffering: Nov 14, 2016

FlapYoJacks posted:

Is there a better library for AJAX calls? I�ll gladly use it!

just use fetch and queryselector op

# ¿ May 23, 2023 22:41

susan b buffering: Nov 14, 2016

lifg posted:

i like next.js server side rendering. its a return to how websites should be. assemble as much as possible on the server and send over mostly finished html. it�s fast.

i like server side rendering but using js for it is gonna be an lol from me

# ¿ Jun 1, 2023 01:11

susan b buffering: Nov 14, 2016

Kazinsal posted:

I guarantee there's a set of word macros out there that make your document look like it was typeset in latex

i have 100% seen this exact thing somewhere

# ¿ Oct 15, 2023 05:31

susan b buffering: Nov 14, 2016

rider does. topic locked

# ¿ Nov 26, 2023 22:13

susan b buffering: Nov 14, 2016

mystes posted:

and microsoft decided to pointlessly make their own json library because of nih syndrome

it's a better library at this point and the newtonsoft guy literally works at microsoft lol

# ¿ Nov 27, 2023 20:16

susan b buffering: Nov 14, 2016

Shaggar posted:

hes also writing system.text.json.

system.text.json is goofy cause they're trying to use c# language keywords as a replacement for attributes but it doesnt work because the attributes provide way more customization. see: [JsonRequired] vs the required modifier which is also only in c# 11/.net 8.

its dumb. Eventually it will probably be fine tho.

yeah some of the interface bugs me. the last couple versions are where i started feeling good about actively migrating towards it. the httpclient extensions have been ace for that.

i do hate that the default serialization options for asp.net are incompatible with the general default options.

susan b buffering fucked around with this message at 02:48 on Nov 28, 2023

# ¿ Nov 28, 2023 02:39

susan b buffering: Nov 14, 2016

Shaggar posted:

If im in full control of the interface design for a webapi i think system.text.json is fine, but if a web "developer" was involved somewhere then you need the hackyness and workarounds of newtonsoft.

we have to consume some absolutely dogshit apis that are basically some database query results piped through snaplogic as json. the only thing more upsetting than the mess of JArrays and Jtokens we use to read them is the thought of reimplementing them with system.text.json

# ¿ Nov 28, 2023 03:06

susan b buffering: Nov 14, 2016

CPColin posted:

I didn't know figma wireframes had permission to control monitor power

lmao

# ¿ Dec 13, 2023 16:55

Adbot: ADBOT LOVES YOU

# ¿ Apr 27, 2024 02:18

susan b buffering: Nov 14, 2016

redleader posted:

ah ok, after doing a bit of googling it seems that "migrations" specifically refer the scripts generated by some tool (or you write by hand, i guess) then blindly run on any and all databases. the ones that don't care about the actual current schema, just whatever the tooling (or author) thinks the current schema is

idk there are migration tools like dbup that just run user-written scripts in order

# ¿ Jan 17, 2024 06:27

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > YOSPOS > awful programming: feature flags and suffering