markdown guide
 

I write git status at least 4000 times a day, for no reason whatsoever.

 
 

Yep that's me. I even write git status after committing just to make sure ๐Ÿ˜…

 

And also git branch after git checkout just to make sure I'm on the right branchโ€ฆ ๐Ÿ˜‚

My prompt reflects which branch I'm on, super helpful!

 

That has actually helped me once, I'd forgotten to git add some files

 
 

With as much as I check git status, I don't have time for a 3-letter alias! gs is all I can afford ๐Ÿ˜œ

 

Most of the time, git log follows git status in my case. :P

 

tig gives me more information, more concisely. Try it out :)

 

Is it bad ?
Im also guilty of this one + ls and clear

 
 
 

haha me too. So i added a alias as s => git status. Lol lazy me

More shortcuts like this in here github.com/gokulkrishh/dotfiles

 
 
 

Never taking the time to learn hotkeys. Not reaching out to a second set of eyes before spending too much time on something. Not doing enough design up front.

 
 

I use vim for basically everything & I can't remember how to do macros without looking it up. I think just generalize this to "not learning your tools effectively" & I totally agree.

 
 

I rewrite other peopleโ€™s code to my extremely high standard so I can understand it. Such a time waste.

 

I know this far too well. There is this strong feeling of โ€žYou canโ€™t leave it like it is, when you know it.โ€œ but in the end ... it would most likely run for ages without any issues if we wouldnโ€™t rewrite it. And maybe nobody would ever touch it again. :-)

 

Iโ€™m sure this thread will contain much worse habits than that! ๐Ÿ˜‚

 
 
 
 
 

Same! I've never written a test. I know that I should, but... ยฏ_(ใƒ„)_/ยฏ

 
 

Dateline !== No test, or just lrave like this

Agile !== No test

 
 

Stare at my editor -> search for help in duckduckgo -> open social media -> change music on spotify๐Ÿ˜‚

 
 

That's literally me every single day ๐Ÿ˜‚๐Ÿ˜‚

 

Being too lazy to turn off 'helpful' system settings, like the one that autocompletes a closing bracket. I used to end up adding a closing bracket myself and then having too many so had to hit backspace to get rid of one.

Now that that has become muscle memory, whenever I'm not using that tool I hit backspace on autopilot even though there is no 'helpful' closing bracket to get rid of.

 

I do this all the time!
It's all muscle memory now.

The other way the IDE gets in my way is when I want to add a closing bracket or parenthesis but my cursor is already immediately before one and the editor treats it as if they're the same - so I think I've added enough, but oh no.

 

Ugh. Instant rage quit.

There are also some strange moments when the cursor just vanishes completely and I don't know where I am anymore.

 
 

Step 1: Spot a code smell
Step 2: Add a // TODO comment
Step 3: Never get back to it

Repeat.

This is the story of my dev life ๐Ÿ˜”

 

If you're a VS Code user, checkout the Todo Tree extension.

It's a nice alternative to searching in all files for // TODO or // DEBUG before commiting!

 
 
 

This was me until I worked with folks who set great examples. I eventually learned better habits.

 

Not sure I'd agree with this one, depends on what you're trying to accomplish.

Committing directly to trunk has its benefits, everyone gets the changes immediately but you just need to make sure that the changes are self-contained enough to not break anything.

I used to think this was the way to go too until I read Continuous Delivery
all the way through.

 
 
 
 
  • git push --force on a regular basis
  • using VSC to open dot files and not taking the time to learn vim or nano, etc
 

My git alias for that command is git fush ๐Ÿ˜„

 

I need that in my life. I'm an obsessive compulsive interactive rebaser.

 
 

I think all Unix/Linux developers should know vim. Because vi or vim is almost always installed on these variants. Sure for local development a UI is great, a very useful tool! But for editing files on remote machines, the ability to ssh to a remote machine and make a quick change is very important.

 

If you start using โ€”force-with-lease instead the habit becomes quite a bit less bad.

 
 

Sometimes I struggle and go for the longer route instead of just focusing on the quick fix.

I often set a breakpoint and learn how code works from there instead of just looking at the documentation.

A part of me wants to fix everything but I've become much more pragmatic over time ๐Ÿ˜‚

 

I try to do the right thing first.

Often I fail because of missing information, then I do a bunch of quick fixes second.

Then after the fixes built up too much, I do a rewrite as it should be and everything is good.

The second part often happens because I only find the missing information after I tried enough fixes.

Dumb problem and probably unsolvable ๐Ÿ˜’

 

Being too obsessed with code perfection instead of just shipping the feature.

 

Prettier and refmt helped me so much with this

 
 

I still tend to jump in and start coding my first idea. I tell myself it's a proof of concept or that I'm experimenting with different ideas but too often I just end up going with it and not putting the effort in to find a better solution.
I do this even if I have time to plan the project out, because I've worked in tight deadline agencies for so long it's become a habit that's hard to break.

 
 

This. I guess sometimes its also a strength. Atleast I tell myself...

 

Confession:
TFS (Team foundation) lets you commit your code without a message. I sometimes commit without a message...

And I reformat others' code.

The worst offender is always Googling for simple tasks such adding simple HTML elements such as adding CSS link tags and script tags... (I wonder if it's just me not knowing how to add them out of memory ๐Ÿค”)

 

We moved all our code from TFVC to TFS Git which I can highly recommend. You get the best of both worlds (TFS backlog and project control together with Git version control). And you will never commit without a message again ๐Ÿ˜‰

 

Wow ๐Ÿ˜ฎ
Thanks, Jacob.
I wasn't aware it was possible to migrate to TFS Git and keep all the history.

I will check it out~

 

No commit message is such an odd thing to me.

I also always good css link tags ๐Ÿ˜‚

 

git way of forcing a message should be how all VCS should be IMHO.

I also always good css link tags ๐Ÿ˜‚
๐Ÿ˜‚ ๐Ÿ‘Š

 

Not writing (and updating) documentation explaining intent and purpose.

 

I work in a small team (2 front-end devs), and my co-worker is a bit picky about documenting every function, class, method, etc etc.
But sometimes all he documents is something a function name tells it upfront, I really think there's no purpose in doing something like:

/** This function transforms lower case string to uppper case **/
function lowerCaseToUpper(str...);

If you declare functions and variables with intuitive names, sometimes there's no need to explain them :)

 

It is called self-documenting functions for a reason right?

yup... unless that function transforms the string using some alien computation, there's no need for that

 

So many, here are a few that come to mind...

  • git add .
  • git commit -m {same message as the last commit, but I forgot something and donโ€™t feel like writing a new message}
  • git push origin master
 
 

git commit --amend could work, although it would require a sinful git push -f if you've already pushed it.

 

I get debug/Refactor/develop creep. Often I'm working on one thing and I notice something that can be fixed rewritten or moved somewhere so that it can be reused... Or I suddenly decide to stub out something for later later. Sure, it's stuff that had to be done at some point, but hard to explain or justify when you list hours to tasks for a single component and for some reason a component was finished, plus 2 stubbed, plus one refactored, and a new service was created. ๐Ÿคช

 

It's hard for me to work with poorly written legacy code.
From one side I want to rewrite a module I'm working on to meet quality standards. But from other side usually the module is connected with other modules in an unpredicable spaghetti way, so refactoring might be a very deep rabbit hole. And there are always deadlines.
So my bad habit is procrastinating when I face such problems. I need to get a cup of coffee or scroll through twitter for 5 minutes before I can return to the work ๐Ÿ˜‚

 

Writing code that "just works" when I'm feeling tired.

I don't have the effort to think critically about what I'm writing and follow good practices when I'm exhausted or sleepy (generally after work when I'm working on side projects). Its akin to reading through drunk texts when I see what I what I committed the night before!

 

Using mouse when the keyboard can do the job in one third of the time

Stupid names in variables

Not using a git prompt

Windows ๐Ÿ˜Š๐Ÿ˜Š

 

Coming from years of RTS and FPS gaming, often the mouse is simply qick enough ๐Ÿค–

 

Probably overanalysis. I've got an eye for detail thinking about edge cases, but that has a tendency to bog me down before I start writing code for the normal case that I could then iterate on, which I'm trying to work on

 

This is what I came to post. I can spend up to an hour checking if there's something in the code base that can be reused for the task I have at hand, and then thinking about how it can be reused, adapted, or modified for different cases. And sometimes it's just easier to write a simple function or method to deal with the problem.

 

For me, the worst habit comes out most when I'm trying to help others through problems in their code: learning when to step back and look for other solutions.

Just the other night, I got sucked into an 80 reply Slack conversation about a bug in someone's Heroku build. A lot of those messages came from my own misunderstanding in the build steps of their app and how Heroku handles deployment. After reaching a dead end pulling from my own knowledge, I did a quick crawl through GitHub to find similar issues to his own, and we soon found the culprit. Basically, I found that even when I think I understand an issue with a hacky way to solve it, it's always worth an Internet search to find the best solution.

It's really hard to find that balance of when to look for help versus when I should go with my gut, but I'm always working on it!

 

I do a lot of Razor templating (C# with HTML), and when I do something dependent on an external API (e.g., an Instagram widget), I usually wrap it in a try-catch block without handling the exception. I usually just print an HTML comment like, <!โ€” Something went wrong with the Instagram widget. โ€”>.

 
 

Me neither.

The problem is often, that only code fails that couldn't be tested automatically anyway. (In app purchases, etc.)

 

Using Facebook's messenger when I'm coding. I can't let go of this one. I think it's because I isolate myself too much. I don't go out. Not at all. I don't use Facebook (I only use messenger, I don't have a Facebook). I use Twitter to see Feed related to coding as well. So I just can't let go of messenger. I need to talk to people. But I consider it a bad habit because I keep it open while coding. It will probably be healthier to just use Messenger when I'm not coding.

 

I'm feeling this too. I talk to people but I think it's a huge waste of time. Like serious I'm coding 12h a day at a regular base but I'm feeling like I could do the 12h work in 6h if I close social media

 

Finding an exciting library/package/technique and then prowling for problems that my new toy will fix.

Never going back and removing the shell pipes and system calls after the prototype worked.

parsing argv in main.

 

Using "up" arrow key in terminal to find a command instead of typing it.
I know I've used it already! ๐Ÿคฆโ€โ™‚๏ธ

 
 

Use ctrl-r to search for the command in your history instead

 

I know and use that frequently, but man - that up button won't leave me alone ๐Ÿ˜‚

 

Code: When I just want to jump on codeland without think for one sec of what I am doing.

Time: I tend to give estimations too quickly.whrn someone ask me how much time is going to take me... My brain flips and it make me think I am the developer version of flash

Social media

 

porn

however you define porn these days.... it's a distraction. If your porn is hours of netflix, ok. But that is potentially a bad habit. Basically everything that isn't productive is bad -right?

I disagree with this concept. IMHO people need to do various things in order to be productive at 'something'. I think there's some science behind this but meh, look that up on your own time.... I have 1 year to catch up on free BBC Dr. Who episodes via some darkweb torrent somewhere...

 

My editor is VIM, and I use arrow keys. Also I don't use autocomplete, I actually hate it. Not sure why. I like to write words my self. Anyways, most of my time coding is hitting arrow keys, not typing actual code.

 
  • Getting distracted by social media (especially dangerous when my music is on YouTube).
  • Writing tests after I wrote the code, when I should've used TDD and wrote the test first.
  • Not asking in time (although consciously trying to ask more often does help)
  • Being late for the stand-up
 

If I get a bug in X project that I can't figure out for a while I'll jump to another project and work on that one until I get another bug that I can't figure out for a while then I'll jump back to the original project sort that one out, and back and forth

 

Before: relying on IDE. Lately I realised I didn't know any packages/namespaces (java, c# respectively) until I read the name and decided OK this could be the one I need. And I'm talking about main language features not just libraries.

After I've changed that habit I still copy paste stuff for JWT or any kind of filter/middleware then read code and change what I don't like, don't need or need differently implemented. Shame on me.

 

When on a tight deadlines, I don't write any documentation, and sometimes code/component splitting is also left behind.
This leads to some refactoring time after the deadlines are over and boss is not in a hurry to have something ready!

 
 

I have a bad habit of overengineering things, so I never trust the code I wrote previously, even yesterday, or before lunch, or before that last meeting, or.... so have a horrible workflow of git reset โ€”hard then rewriting or restoring from local history all the things I just worked on.

 

I'm running yarn upgrade and bundle update way too often.

 

my co-worker also bumps versions on package.json way too much without any precautions.
sometimes the code breaks and guess what.... NEW VERSION OF SOME PACKAGE is causing it!

 

Obsession with doing everything upfront, instead of making it work and improving it later. I also tend to avoid committing parts of the code related to the feature I haven't finished yet. As a result, I keep the code on my machine for tomorrow and commit when I build the whole feature.

 

I often get overconfident and forgo unit testing until I finish coding a feature. Which often leads to some really fun debugging.

 

Never doing the bandaid solution and always engineering things to a fault. This doesn't sound like too bad of a thing but trust me it's okay to put on the bandaid sometimes. I would have saved many hours if I would have put the bandaid solution on some features that my project owners didn't even want a few weeks / months later.

 

When I need to use search, instead of searching in VIM, I launch sublime and search in there.

 

ripgrep needs to become your friend ๐Ÿ˜‚

 

o.0 this is a thing?! That's awesome! Definitely going to grab that as soon as I can.

It's also the tool that Visual Studio Code uses to search :D

 
 
 
 

I jump around editors so much that quite often I commit files I've been editing in VSC with a :wq at the end

 

My trouble is with esc. Let me click here... open a modal... enter some text... done typing, esc... dammit, where'd it go?

 

Constantly checking social media along with coding.

 
  1. Procrastinating
  2. Not starting my work right away
  3. Being lazy when I should be coding
 

Don't plan early to adapt future changes. Once software getting more and more complex, Even I can not understand what I have written.

 

Opening GitHub for something work related, and get distracted looking deeply on what people I'm following have starred ๐Ÿ˜…

Classic DEV Post from May 10

Discuss: GitHub Special Event

GitHub is holding a "GitHub Special Event" today. Let's discuss what we're thinking leading up to it + while it's happening <3

Ben Halpern profile image
A Canadian software developer who thinks heโ€™s funny.

Sore eyes?

dev.to now has dark mode.

Go to the "misc" section of your settings and select night theme โค๏ธ