Join devRant
Do all the things like
++ or -- rants, post your own rants, comment on others' rants and build your customized dev avatar
Sign Up
Pipeless API
From the creators of devRant, Pipeless lets you power real-time personalized recommendations and activity feeds using a simple API
Learn More
Search - "code smells"
-
Was lead developer at a small startup, I was hiring and had a budget to add 3 new people to my team to develop a new product for the company.
Some context first and then the rant!
Candidate 1 - Amazing, a dev I worked with before who was under utilized at the previous company. Still a junior, but, she was a quick learner and eager to expand her knowledge, never an issue.
Candidate 2 - Kickass dev with back end skills and extras, he was always eager to work a bit more than what was expected. I use to send him home early to annoy him. haha!
Candidate 3 - Lets call him P.
In the interview he answers every question perfectly, he asks all the right questions and suggests some things I havent even thought of. CTO goes ahead and says we should skip the technical test and just hire the guy, his smart and knows what his talking about, I agree and we hire him. (We where a bit desperate at this stage as well.)
He comes in a week early to pick up his work laptop to get setup before he starts the next week, awesome! This guy is going to be an asset to the company, cant wait to have him join the team - The CTO at this stage is getting ready to leave the company and I will be taking over the division and need someone to take over lead position, he seems like the guys to do it.
The guys starts the next week, he comes in and the laptop we gave him is now a local server for testing and he will be working off his own laptop, no issue, we are small so needed a testing stack, but wasnt really needed since we had procedures in place for this already.
Here is where everything goes wrong!!! First day goes great... Next day he gets in early 6:30am (Nice! NO!), he absolutely smells, no stinks, of weed, not a light smell, the entire fucking office smells of weed! (I have no problem with weed, just dont make it my problem to deal with). I get called by boss and told to sort this out people are complaining! I drive to office and have a meeting with him, he says its all good he understands. (This was Friday).
Monday comes around - Get a call from Boss at 7:30am. Whole office smells like weed, please talk to P again, this cannot happen again. I drive to office again, and he again says it wont happen again, he has some issues with back pain and the weed helps.
Tuesday - Same fucking thing! And now he doesnt want to sign for the laptop("server") that was given to him, and has moved to code in the boardroom, WHERE OUR FUCKING CLIENTS WILL BE VIEWING A DEMO THAT DAY OF THE PRODUCT!! Now that whole room smells like weed, FML!
Wednesday - We send P a formal letter that he is under probation, P calls me to have a meeting. In the meeting he blames me for not understanding "new age" medicine, I ask for his doctors prescription and ask why he didnt tell me this in the interview so I could make arrangements, we dont care if you are stoned, just do good work and be considerate to your co-workers. P cant provide these and keeps ranting, I suggest he takes pain killers, he has none of it only "new age" medicine for him.
Thursday - I ask him to rather "work" from home till we can get this sorted, he comes in for code reviews for 2 weeks. I can clearly see he has no idea how the system works but is trying, I thought I will dive deeper and look at all of his code. Its a mess, nothing makes sense and 50% of it is hard coded (We are building a decentralized API for huge data sets so this makes no sense).
Friday - In code review I confront him about this, he has excuses for everything, I start asking him harder questions about the project and to explain what we are building - he goes quiet and quits on the spot with a shitty apology.
From what I could make out he was really smart when it came to theory but interpreting the theory to actual practice wasnt possible for him, probably would have been easier if he wasnt high all the time.
I hate interview code tests, but learned a valuable lesson that day! Always test for some code knowledge as well even if you hate doing it, ask the right questions and be careful who you hire! You can only bullshit for so long in coding before someone figures out that you are a fraud.16 -
My boss isn't really a developer. He isn't part of the development team and doesn't know any technical details about the product. He doesn't want to code, "too much effort", he just wants to boss. But he wrote some php in the early 2000's and is really, really proud of his codecademy html/css badge...
And that makes him dangerous.
Today I hear him talk from behind his laptop: "Right, we have this page for creating management groups, but we can't edit them yet. I can fix that!"
This task is literally on the current sprint, but he doesn't know that because he doesn't attend scrum meetings and ignores everything people say to him.
Me: This smells like probable cause, let's look with suspicion over his shoulder.
Boss:
"OK, right-click create.blade.php -> copy.
then right-click directory -> paste.
now just rename file to edit.blade.php!"
I start walking to the office kitchen.
Boss mumbling in the background:
"Now all I need to do is just copy the whole method in the controller, change the post url in the form, and modify the <h4> at the top, so it says edit instead of create."
Boss, looking at me now:
"This is so easy... creating and editing is almost the same thing, you can just copy paste all the code from one template to the other! I don't understand what you developers are always complaining about!"
Me: *Hands him a roll of paper towels*
Boss: "What is that for?"
Me: *points at code*6 -
!!fml
"Root, go fix this bug. It'll take you two days."
The "bug" is a feature that was never implemented for one particular payment type.
The code in question is two years old, full of typos, smells, junior-isms, and is convoluted AF. The feature's commit touched 190 files and implemented many other features as well. Thus far, I have been unable to narrow down where this particular feature's code lives for the other payment types, nor which code or payment paths lead to it. Burned out, I can barely focus on the screen, let alone follow its many twisting and dynamically-inferred paths. I hint as to the ticket's scavenger hunt nature during standup.
"But I wrote comments on the ticket telling you exactly where to look to fix it," Thundercunt admonishes in front of the team.
"Sure, you did," Root replies. "You reworded what the original dev had said in the comments 20 minutes prior, and agreed with him. His comments were helpful, but it doesn't tell me how any of it works," she continues.
TC scoffs and closes the meeting.
Root stares blankly, seeing neither code nor screen, questions her life decisions, and recalls the previous tickets she has worked on: nearly every one of them busywork, fixing other people's bugs. Bugs she never could have gotten away with if she tried.
"Why do I put up with this?" She asks. "They don't care, and it's killing me."
But the bills remain, and so must she.
"Fuck my life" she finally decides.20 -
Hello there, just couple of words about PHP. I've been develop on PHP more than 10 years, I've seen it all 3,4,5,{6},7. Yes PHP was not good in terms of engineering and patterns, but it was simple, it was the most simple language for web to start those days. It was simple as you put code into file, upload it via FTP and it works. No java servlets, no unix consoles, no nothing, just shared hosting account was enough to host site, or even application with database. As database everybody used to have mysql, again because its simple to start and easy to maintain. So PHP+MySQL became industry standard on Web during 00-2012, and continues in some way.
You can write HTML and logic inside single file, within php code, even more single file may content few pages, or even kind of framework. That simplicity and agility sticks everybody who wants to develop sites with PHP.
This is pretty much about why it is so popular.
Each good or wannabe PHP developer in an early days write its own framework or library (like in javascript this days because of nodejs)
Imagine that PHP has hadn't have package manager, developers used to have host packages on their own sites, then various packages catalog sites created, and then finally composer. A gazillions of php code had spread over internet, without any kind of dependency control. To include libraries to your projects you have to just write include, or require. Some developers do it better than others.
So what we have ? A lots of code, no repositories, zip archives with libraries, no dependency control.
Project that uses that kind of code are still alive even today, they are solid hose of cards, and unmaintainable of course.
And main question that I'm trying to answer is Why PHP is not good ?
- First is amount of legacy code which people copy and pasted into their project, spread it even more like a virus.
- Lack of industry standards at the beginning lead to a lots of bad practices among developers. PHP code usually smells.
open source php projects in early days was developed in same conditions so even in phpbb, phpnuke, wordpress, drupal used to have a lot of bad practices in their codebase. So php developers usually not study by another library, instead they write their own frameworks/libraries.
- "It works", - there are no strong business demands, on web development, again because lack of standards, and concerns.
This three things are basically same, they linked to each other and summarize of answer of why PHP have strong smells and everybody yelling against it.
Whats is with PHP nowadays ? Of course PHP today is more influenced by good practice of webdev. Composer, Zend, Laravel, Yii, Symphony and language it self became more adult so to say, but developers...
People who never tried anything except PHP are usually weaker in programming and ecosystem knowledge than people who tried something else, python, perl, ruby, c for instance.
Summary
PHP as any other programming language is a tool. Each tool has its own task. Consider this and your task requirements and PHP can be just good enough solution.
"PHP is shit" - usually you heard that from people who never write strong applications on PHP and haven't used any good tools like Symphony or Laravel.
Cheap developers, - the bigger community, the more chance to hire cheap developers, and more chance to get bad code. That can be applied on any other language.
PHP has professionals developers, usually they have not only php on scope.
That's all folks, this is very brief, I am not covering php usage early days in details, but this is good enough to understand the point.
Enjoy.8 -
I worked for over 13 hours yesterday on super-urgent projects. I got so much done it's insane.
Projects:
1) the printer auto-configuration script.
2) changing Stripe from test mode to live mode in production
3) website responsiveness
I finished two within five minutes and pushed to both QA and Production. actually urgent, actually necessary. Easy change.
The printer auto-configure script was honestly fun to write, if very involved. However, the APIs I needed to call to fetch data, create a printer client, etc... none of them were tested, and they were _all_ broken in at least two ways. The CTO (api guy in my previous rant) was slow at fixing them, so getting the APIs working took literally four hours. One of them (test print) still doesn't work.
Responsiveness... this was my first time making a website responsive. Ever. Also, one of the pages I needed to style was very complicated (nested fixed-aspect-ratio + flexbox); I ended up duplicating the markup and hacking the styling together just to make it work. The code is horrible. But! "Friday's the day! it's going live and we're pushing traffic to it!" So, I invested a lot of time and energy into making it ready and as pretty as I could, and finally got it working. That page alone took me two hours.
The site and the printer script (and obv the Stripe change as well) absolutely needed to be done by this morning. Super important.
well.
1) Auto-configure script. Ostensibly we would have an intern come in and configure the printers. However, we have no printers that need configuring, so she did marketing instead. :/ Also, the docs Epson sent us only work for the T88V printer (we have exactly one, which we happened to set up and connect to). They do not work for the T88VI printers, which is what we ordered. and all we'll ever be ordering. So. :/ I'll need to rewrite a large chunk of my code to make this work. Joy :/
2) Stripe Live mode. Nobody even seemed to notice that we were collecting info in Test mode, or that I fixed it. so. um. :/
3) Responsiveness.
Well. That deadline is actually next Wednesday. The marketing won't even start until then, and I haven't even been given the final changes yet (like come on). Also! I asked for a QA review last night before I'd push it to production. One person glanced at it. Nobody else cared. Nobody else cared enough to look in the morning, either, so it's still on QA. Super-important deadline indeed. :/
Honestly?
I feel like Alice (from Dilbert) after she worked frantically on urgent projects that ended up just being cancelled. (That one where Wally smells that lovely buttery-popcorn scent of unnecessary work.)
I worked 13 hours yesterday.
for nothing.
fucking. hell.undefined fuck off we urgently don't need this yet! unnecessary work unsung heroine i'm starting to feel like dark terra.7 -
If programmers became musicians we would see
- Wake me up when my build ends, 21 cores, Boulevard of broken CI pipelines by Blue Screen Day
- Smoke from my cabinet by Deep For-Loop
- This is how you debug me by Loopback
- Post-release rhapsody by debug queen
- Another bug in the code by Programmer Floyd
- Smells like bad code by Coders from Botswana
- A place for my code, Cure for the bug by Likin to code at dark
etc etc..5 -
Last week, the team lead told me that he can't merge because my code has code smells and going forward, can't have that. We use Sonar and well the way to "fix it" according to him is to mark the line using //NOSONAR.
Most of the issues are minor like Unused imports and for me incomplete TODOs.
And before the "verbal" rule was only need to fix Major + issues. And well the reason I use TODOs is to mark code that probably needs changing in the future. I know there's going to be some feature that these lines have to be changed. But the requirements are fully defined yet from business.
But I sort of blew up on him. YOU WANT TO ENFORCE ZERO CODE SMELLS NOW?!?!?! AND THESE MINOR ISSUES? MARK THEM WITH NOSONAR?
HERE'S WHAT I THINK FOR THE LAST X YEARS... THE CODE DESIGN IS SHIT, MINOR CODE SMELLS AND MANUALLY MARKING THE ONES U NEED TO KEEP... ARE THE LEAST OF OUR PROBLEMS...
THE OTHER PROBLEMS I'VE MENTIONED BEFORE EVER. MOS YEAR BUT YOU DIMWITS NEVER LISTEN.
YOU THINK MY TODOS ARE BAD... 90% OF THE CODE AND FEATURES (THE ONES NOT DONE BY ME) LOOK AND SMELL LIKE MONKEY SHIT. UNDOCUMENTED, MESSY, FULL OF BUGS.
AND GUESS WHAT? NEW FEATURE, SOME DEV FORGETS TO CHANGE SOME COMPONENT THAT DEPENDS ON IT. WOULDN'T IT BE GREATE IF THERE WERE BOOKMARKS... O WAIT...
i just was catching up on comics again and saw this one... with triggered my memory and this rant... My first thought was to forward it to him...11 -
Hey y'all! It's been a while since I logged in.
But I'm back with a dev jingle, just like last year:
Jingle bells,
Batman smells,
Look how my code compiles,
Oh what fun, it is to see,
No reds in my console!16 -
Folks...
I think I need to get away from web development...
Honestly, no grudge held against web/mobile development itsef... But the projects, the teams, the workflows... It's always shitty af.
I'm fed up with the bad architecture, poor management decisions, unmaintained legacy code, broken windows, arrogant juniors, arrogant seniors, code smells left to rot, the freaking red door... Hell! The fucking "we don't have time for that" answer to testing... Damn!
Been there done that.
Feels like it's always the same crap and unfortunately, it's rare to start a professional project from scratch.
Fucking angular, broken piece of shit.
Fucking react (& RN) community modules, broken pieces of shit.
Fucking lazy-ass node developers.
Fucking ES and fucking garbage proposals submitted to the TC39.
I wish I could do Haskell / Rust / Clojure professionally... I could even enjoy Go with a good team... Anything but that huge pile of dogshit JS and its community of brainfucked so-called developers.10 -
So I'm a perfectionist, especially when with code, smells, solid, design patterns, naming conventions, etc and I be have this co-worker that blackmails me every time he doesn't want to do something saying "I don't know it so my code is gonna be ugly".6
-
Best tool:
Your hands!
- incredibly flexible
- express a lot of commands trough very little code (just raise the middle finger and tell me if you are not expressing something VERY strong with VERY little complexity)
- reusable
- interfaces
- smells of good soap
Worst tool:
Your brain
- highly power consuming
- wrinkly, ehw!
- overthinks a lot
- imposter syndrome
- hooked on sugar like it was cocaine
- hooked on cocaine like it was sugar
- refuses to comprehend chthulu5 -
So before the Age of JavaScript, when programming was trying to be an engineering discipline, I felt like we were getting close to figuring out what worked and what didn't. We had rules of thumb (more general than Patterns) and code smells.
Then JavaScript came in and no one had time to think about "engineering" anymore. I'm fine with MVP and small iterations, but the disdain I see for making code clean and extendable and improvable is baffling (and annoying). First-time coders might never have had to fix someone else's code, but two weeks in a chair should have fixed that.
It's not that understanding code is so hard (although it can be); understanding the _intent_ is hard. This MVP is great, but when no one had time to document what is actually supposed to happen, programmers have to reverse-engineer the *design*.4 -
Ordered a takeout wok/poke this evening from a new place. The only thought after tasting - what kind of junior made this....? I hope I write code better than he cooks food.
It looks like what it's supposed to look
it smells as it should
it has all the ingredients
but the taste is just... Not right.
I takes special skill to make it so wrong!3 -
Let's try this.
In the project I'm working there is an strict rule : NO COMMENTS!!!
I mean wth, the times I've spend hours trying to understand the crappy legacy code in VB.Net that has been there almost decades, that wouldn't happen with comments, I know i know there are some supernatural developers that think in binary and their eyes work as compilers, but I'm not like that, so seriously go to hell.
P.S. Of course I follow that rule, after all, my code is so damn perfect that even a baby can understand it.
jkundefined devops etiquette stupidest pichardo for president stupid stupider stupid stuff jk rant code smells comments3 -
Lead developer wants to put SQL statements in to records of the database for the code to execute (in Rails). This smells really bad to me, am I over reacting?8
-
A (work-)project i spent a year on will finally be released soon. That's the perfect opportunity to vent out all the rage i built up during dealing with what is the javascript version of a zodiac letter.
Everything went wrong with the beginning. 3 people were assigned to rewrite an old flash-application. Me, A and B. B suggested a javascript framework, even though me and A never worked with more than jquery. In the end we chose react/redux with rest on the server, a classic.
After some time i got the hang of time, around that time B left and a new guy, C, was hired soon after that. He didn't know about react/redux either. The perfect start off to a burning pile of smelly code.
Today this burning pile turned into a wasteland of code quality, a house of cards with a storm approaching, a rocket with leaks ready to launch, you get the idea.
We got 2 dozen files with 200-500 loc, each in the same directory and each with the same 2 word prefix which makes finding the right one a nightmare on its on. We have an i18n-library used only for ~10 textfields, copy-pasted code you never know if it's used or not, fetch-calls with no error-handling, and many other code smells that turn this fire into a garbage fire. An eternal fire. 3 months ago i reduced the linter-warnings on this project to 1, now i can't keep count anymore.
We use the reactabular-module which gives us headaches because IT DOESN'T DO WHAT IT'S SUPPOSED TO DO AND WE CANT USE IT WELL EITHER. All because the client cant be bothered to have the table header scroll along with the body. We have methods which do two things because passing another callback somehow crashed in the browser. And the only thing about indentation is that it exists. Copy pasting from websites, other files and indentation wars give the files the unique look that make you wonder if some of the devs hides his whitespace code in the files.
All of this is the result of missing time, results over quality and the worst approach of all, used by A: if A wants an ui-component similar to an existing one, he copies the original and edits he copy until it does what he wants. A knows about classes, modules, components, etc. Still, he can't bring himself to spend his time on creating superclasses... his approach gives results much faster
Things got worse when A tried redux, luckily A prefers the components local state. WHICH IS ANOTHER PROBLEM. He doesn't understand redux and loads all of the data directly from the server and puts it into the local state. The point of redux is that you don't have to do this. But there are only 1 or 2 examples of how this practice hurt us yet, so i'm gonna have to let this slide. IF HE AT LEAST WOULD UPDATE THE DATA PROPERLY. Changes are just sent to the server and then all of the data is re-fetched. I programmed the rest-endpoints to return the updated objects for a very reason. But no, fuck me.
I've heard A decided (A is the teamleader) to use less redux on the next project and use a dedicated rest-endpoints for every little comoutation you COULD DO WITH REDUX INSTEAD. My will is broken and just don't want to work with this anymore.
There are still various subpages that cant f5 because the components cant handle an empty redux state in the beginning, but to be honest i don't care anymore. Lets hope the client will never find out, along with the "on error nothing happens"-bugs. The product should've been shipped last week, but thanks to mandatory bugfixes the release was postponed to next week. Then the next project starts...
Please give me some tips to keep up code quality over time, i cant take this once more.
I'm also aware that i could've done more, talking A and C about code style, prettifying the code, etc. Etc. But i was busy putting out my out fires, i couldn't kill much of the other fires which in the end became a burning building (a perfect metaphor for this software)4 -
Is your code green?
I've been thinking a lot about this for the past year. There was recently an article on this on slashdot.
I like optimising things to a reasonable degree and avoid bloat. What are some signs of code that isn't green?
* Use of technology that says its fast without real expert review and measurement. Lots of tech out their claims to be fast but actually isn't or is doing so by saturation resources while being inefficient.
* It uses caching. Many might find that counter intuitive. In technology it is surprisingly common to see people scale or cache rather than directly fixing the thing that's watt expensive which is compounded when the cache has weak coverage.
* It uses scaling. Originally scaling was a last resort. The reason is simple, it introduces excessive complexity. Today it's common to see people scale things rather than make them efficient. You end up needing ten instances when a bit of skill could bring you down to one which could scale as well but likely wont need to.
* It uses a non-trivial framework. Frameworks are rarely fast. Most will fall in the range of ten to a thousand times slower in terms of CPU usage. Memory bloat may also force the need for more instances. Frameworks written on already slow high level languages may be especially bad.
* Lacks optimisations for obvious bottlenecks.
* It runs slowly.
* It lacks even basic resource usage measurement.
Unfortunately smells are not enough on their own but are a start. Real measurement and expert review is always the only way to get an idea of if your code is reasonably green.
I find it not uncommon to see things require tens to hundreds to thousands of resources than needed if not more.
In terms of cycles that can be the difference between needing a single core and a thousand cores.
This is common in the industry but it's not because people didn't write everything in assembly. It's usually leaning toward the extreme opposite.
Optimisations are often easy and don't require writing code in binary. In fact the resulting code is often simpler. Excess complexity and inefficient code tend to go hand in hand. Sometimes a code cleaning service is all you need to enhance your green.
I once rewrote a data parsing library that had to parse a hundred MB and was a performance hotspot into C from an interpreted language. I measured it and the results were good. It had been optimised as much as possible in the interpreted version but way still 50 times faster minimum in C.
I recently stumbled upon someone's attempt to do the same and I was able to optimise the interpreted version in five minutes to be twice as fast as the C++ version.
I see opportunity to optimise everywhere in software. A billion KG CO2 could be saved easy if a few green code shops popped up. It's also often a net win. Faster software, lower costs, lower management burden... I'm thinking of starting a consultancy.
The problem is after witnessing the likes of Greta Thunberg then if that's what the next generation has in store then as far as I'm concerned the world can fucking burn and her generation along with it.6 -
So I got this new job as Java developer, the people are really great but is the kind of companies that only takes care for fast results and not for code quality.
Because this I have to deal with libraries updated 4 years ago, classes with 8000 lines, methods with 500 lines, a WHOLE lot of work arounds because there is no time to really fix the issue unless it affects directly the customer (something not working or being really slow) aaand we use fucking svn.
Some of this practice's they know and encourage it (+1000 lines classes for example) and every time I try to talk about good practices in the code everyone seems so interested but there is always no time.
Sooo I will stay here for at least two years, I hope I can make a change for good in their code smells.3 -
Storytime!
(I just posted this in a shorter form as a comment but wanted to write it as a post too)
TL;DR, smarts are important, but so is how you work.
My first 'real' job was a lucky break in the .com era working tech support. This was pretty high end / professional / well respected and really well paid work.
I've never been a super fast learner, I was HORRIBLE in school. I was not a good student until I was ~40 (and then I loved it, but no longer have the time :( )
At work I really felt like so many folks around me did a better job / knew more than me. And straight up I know that was true. I was competent, but I was not the best by far.
However .... when things got ugly, I got assigned to the big cases. Particularly when I transferred to a group that dealt with some fancy smancy networking equipment.
The reason I was assigned? Engineering (another department) asked I be assigned. Even when it would take me a while to pickup the case and catch up on what was going on, they wanted the super smart tech support guys off the case, and me on it.
At first this was a bit perplexing as this engineering team were some ultra smart guys, custom chip designers, great education, and guys you could almost see were running a mental simulation of the chip as you described what you observed on the network...
What was also amusing was how ego-less these guys seemed to be (I don't pretend to know if they really were). I knew for a fact that recruiting teams tried to recruit some of these guys for years from other companies before they'd jump ship from one company to the next ... and yet when I met them in person it was like some random meeting on the street (there's a whole other story there that I wish I understood more about Indian Americans (many of them) and American engineers treat status / behave).
I eventually figured out that the reason I was assigned / requested was simple:
1. Support management couldn't refuse, in fact several valley managers very much didn't like me / did not want to give me those cases .... but nobody could refuse the almighty ASIC engineers. No joke, ASIC engineers requests were all but handed down on stone tablets and smote any idols you might have.
2. The engineers trusted me. It was that simple.
They liked to read my notes before going into a meeting / high pressure conference call. I could tell from talking to them on the phone (I was remote) if their mental model was seizing up, or if they just wanted more data, and we could have quick and effective conversations before meetings ;)
I always qualified my answers. If I didn't know I said so (this was HUGE) and I would go find out. In fact my notes often included a list of unknowns (I knew they'd ask), and a list of questions I had sent to / pending for the customer.
The super smart tech support guys, they had egos, didn't want to say they didn't know, and they'd send eng down the rabbit hole. Truth be told most of what the smarter than me tech support guy's knew was memorization. I don't want to sound like I'm knocking that because for the most part memorization would quickly solve a good chunk of tech support calls for sure... no question those guys solved problems. I wish I was able to memorize like those guys.
But memorization did NOT help anyone solve off the wall bugs, sort of emergent behavior, recognize patterns (network traffic and bugs all have patterns / smells). Memorization also wouldn't lead you to the right path to finding ANYTHING new / new methods to find things that you don't anticipate.
In fact relying on memorization like some support folks did meant that they often assumed that if bit 1 was on... they couldn't imagine what would happen if that didn't work, even if they saw a problem where ... bro obviously bit 1 is on but that thing ain't happening, that means A, B, C.
Being careful, asking questions, making lists of what you know / don't know, iterating LOGICALLY (for the love of god change one thing at a time). That's how you solved big problems I found.
Sometimes your skills aren't super smarts, super flashy code, sometimes, knowing every method off the top of your head, sometimes you can excel just being more careful, thinking different.4 -
I swear sometimes debugging a new piece of code feels like waddling through a sewer, trying to find a clog. You are unsure where you are, everything smells, there are decades of old shit piled up by the sides and once you start tugging it turns out the entire sewer settled onto it and it's as much of a clog as it is a structural element now.3
-
We rewrote the whole thing, except for iFraming some old pages in. We had to, the system was fucking awful and couldn't cope with any of the new mission critical requirements.
Client didn't understand the scope. Our project leader somehow snuck it in and we worked on it for months. We were sure we'd be kicked off the whole project... Somehow things didn't crash and burn. How it didn't blow up defies rational thought and the laws of physics. The new system worked, the client was happy, and boss made a lot of money.
Lead dev worked weekends for what feels like an eternity, it really was his baby and no one else on our company could have done it. It's where I finally learned how to do things the proper way; DDD, unit testing and TDD, architecture, building strong components in front-end, you name it. Before that I had a great nose for code smells and how not to do stuff, but now I got to see a proper system for the first time. It was glorious.
Then lead dev left and the system degraded quite a bit because new team didn't keep to the architectural patterns or general best practices. But we had a good run.1 -
This code smells, its like a puzzle that only will get more complicated the more you try to solve it.
- me after plowing my keyboard for a week writing subpar ad hoc code -
I am just student looking for job, and got this pre interview test:
Develop an Android or iOS app with login and password input field, download button, place for image we prvided.
... reading further:
What we are looking for in the code ?
internal quality:
-consistent formatting of the source code
-clean, robust code without smells
-consistent abstractions and logical overall structure
-no cyclic dependencies
-code organized in meaningful layers
-low coupling and high cohesion
-descriptive and intention-revealing names of packages, classes, methods etc.
-single small functions that do one thing
-truly object-oriented design with proper encapsulation, sticking to DRY and SOLID principles, without procedural anti-patterns
-lots of bonus points for advanced techniques like design patterns, dependency injection, design by contract and especially unit (or even functional or integration) tests
external quality:
-the app should be fully functional, with every state, user input, boundary condition etc. taken care of (although this app is indeed very small, treat it as a part of big production-ready project)
-the app should correctly handle screen orientation changes, device resources and permissions, incoming calls, network connection issues, being pushed to the background, signing deal with the devil :D and other platform intricacies and should recover from these events gracefully
-lowest API level is not defined - use what you think is reasonable in these days
-bonus points if the app interacts with the user in an informative and helpful way
-bonus points for nice looks - use a clean, simple yet effective layout and design
... I mean really ? and they give me like 2 days ?4 -
ScalaJs React compiles Scala to React.js.
There's some cool typing involved but I haven't done web front-end since nested tables were meta, so there's lots to learn.
There's exactly one senior dev at my company who is fluent in this ScalaReact, so I tag him in the PR for my project. Every day at 10:00 am, slack publicly posts a reminder with @mention that he hasn't reviewed my PR.
Three days later I haven't heard anything so I send a DM over slack asking for feedback... No response.
Four days after the PR I beg for 10 minutes of pairing time, because something in my component hierarchy smells funny. He doesn't have time for me until 5:00 .
I've now built almost a weeks worth of work on the original PR and the feedback I get is 'this works, is performant, and has no obvious bugs, but you can't merge it until you restructure the underlying component hierarchy'
It takes me and another senior dev an entire day of pairing to implement the changes without breaking anything. But, I asked for the feedback because I wanted to learn and write good clean code so I'm irritated but willing to move on.
Yesterday I posted in slack that I was having a hard time following my callback chains to find where the color was assigned to a <td (because I had to add a coloring rule). I wanted to know if I could change the type signature of a component from Tagmod (one or more HTML tags) to VdomTagOf[TableCell] so that it would be clear where the color was assigned.
Instead of just telling me 'no' and giving some context, the react dev gives me:
"Why would a dev need to know about the type unless they’re actually trying to use the thing ? Those are all great questions, but id suggest trying not to prematurely optimize for those until they actually come up"
I flipped my shit. After you couldn't make time for me for a WEEK I had to justify to the CEO why I was spending a day on PURE refactors to accommodate your PREFERENCES. Meanwhile when I'm being VULNERABLE and exposing that I am confused and struggling to complete my task you DISMISS my concerns and attack my motivations.
Unfortunately, this is all happening in the public slack channels and I start defending readability and my premise while triggered. Now I'm riding the shame train for fighting in public slack and trying to pretend none of this ever happened.1 -
Debugging is fine, totally part of the job!
Constantly fixing sh1t and new reports of another pile of sh1t coming every day like somebody is throwing them with shovels at us just to open the codebase that is written by the folks who aren't here anymore with some list of obscure libraries that is last maintained about 5 y ago is not ok.
It is not buggy codebase it is actually coddy bugbase!
I tried to be vocal several times to change technology to more suitable one, to make some improvements and to remove code smells(there is a ton of it, smells like organic garbage dumpster with rotten eggs) but "everything works" and there is no real "value for the customers" in that(fixing, refactoring etc.)!!!
Yea it works with sh1t ton of bugs reported every week. Nobody gives a shit, just contempt with their mediocre lives solving bug at the time while i feel like I'm wasting my time and talent on wrong people and fixing other's shit.
That is what happens when prototype becomes product and ships to production because numbers, money and sh1t!
this is why we who care about our career can't have nice things! I am not god damn pest control, I am f*ckin developer.1 -
Rails views are not meant to have a ton of logic, local variables and 3 or 4 levels of if/else nesting. That's what presenters, view models and assorted other patterns are for. Or helpers, if you really have to.
Yes, this codebase is so packed with legacy it still runs Rails 2.3, and there's no plans to upgrade it, but that's no excuse to keep writing code like it's 2008. MVC does not mean all code must fit in a model, a view or a controller, ffs.1 -
What's the general Software Engineering rule of thumb again for frontend templating code?
If I look at certain websites, I notice some code smells in PHP such as:
$.modal = <?php echo $(base)["username"] != 'me' ?' ': echo 'style="display=none"' ?>
or just in general places in the code where PHP gets used as a templating engine for gluing together pieces of HTML code based on conditionals spread out over the codebase and the database itself too. To make things worse, this carries over to JavaScript ajax functions. As a developer, this to me just seems like spaghetticode.
On the other hand, many popular frameworks properly do templating, such as EJS, containing templating in one place and not mixing it with logic too much but just having simple output like <%= %>.
I know I've seen frameworks like Angular 1 contain pieces of HTML into directives, but maybe that's something different, more 'OO'-simulating or cleaner.3