The older I get the more I value boring, predictable tooling
Posted by minimal-salt@reddit | ExperiencedDevs | View on Reddit | 42 comments
12 yoe dev here and I genuinely cannot tell you the last time I got excited about a new framework dropping
Early career I was always pushing to try the new thing. Looking back some of those calls aged badly (it is part of the process, but still.). A fancy build setup that the next dev had no idea how to touch. An abstraction that felt clean until something broke in production and tracing it was a nightmare
Now when I'm evaluating something the first thing I ask is how it behaves when things go wrong. Not how clever it is, not how clean the docs look on the homepage
Dull and predictable is genuinely underrated. I'll take a boring tool I understand over an elegant one I have to fight every few months. I miss those times
Jeep_finance@reddit
Yes. Or the older I get the more value I proven thing. ie - are we going to have to reinvent the wheel 6 months from now when new feature request means we need X, and new flashy thing doesn’t work with the normal way of doing X
mechkbfan@reddit
Pretty much why I went to Angular over React
I swear it was every month I got told my current design was out of date. Flux, redux, thunk, etc. I can't even remember
I feel we were mostly past the build tools but seems Vite is hot new thing?
I'll wait a year or two thanks
JohnWangDoe@reddit
React won tho
Datamance@reddit
The biggest lie you’ve ever been told is that you can’t use vanilla JS, HTML, and CSS to build a website that works on every browser worth supporting. The browser IS the framework.
JohnWangDoe@reddit
yea but vanilla is and html is hard to read imo, especially after 3 month? I can refactor or extend things with react so much faster.
Apterygiformes@reddit
i don't think there's any winners really
Sunstorm84@reddit
I feel like a winner whenever I don’t have to work with angular, though.
TenYearsOfLurking@reddit
ppl are rewriting their angular apps from observables to signals as we speak. oh and reactive forms are not cool anymore you need to use signal forms...
oh and you gotta refactor for zoneless, or you are not cool.
let's stop pretending that churn doesn't happen in angular world
UnderstandingDry1256@reddit
React context and you’re set for 8 years to date haha
witchcapture@reddit
Didn't Angular have the Angularjs to Angular 2 transition? I never used Angular myself but I heard it was very painful.
Vite I wouldn't even describe as the hot new thing really, at this point it is just stable and boring (positive). The default bundler unless you need something special.
super_powered@reddit
To this day I’m still shocked angular is still around after the angularity 1 to angular 2 upgrade. Had not been backed by google it would have been dead in the water after that imo.
Hovi_Bryant@reddit
Technologies != software design. If we're not asking ourselves why such technologies exist within the codebase, and what problem they're aiming to solve, then Angular, React, Vue, Svelte, I'd argue the application is already directionless.
To OP's point, stability matters. But that should be after mapping the technology to the core domain issues the tech aims to solve.
iegdev@reddit
Precisely why I stick to backend development with C#/.Net
Stable enough to be boring, but with enough new stuff added each year to keep it interesting
barockok@reddit
100% this. After years chasing the shiny thing, I found the real cost isn't the learning curve — it's the 3am debugging when the shiny thing breaks in production and nobody on the team understands why.
The framework you can reason about when half-asleep is worth more than the one that makes you feel clever during standup.
Currently building a voice AI system and deliberately chose boring, well-understood tools over the hot new vector DB everyone is raving about. No regrets.
observability_geek@reddit
Same, I also don’t try to reinvent the wheel or impress the team. I think it happens in other aspects of life as well. I’m over 40, and I think it tends to happen around that age. I also try to minimize my use of AI tooling.
YetMoreSpaceDust@reddit
I'm constantly asked "why do you want to do x yourself?". My answer is always the same - because when something goes wrong, I'll know where to look to fix it.
phantomplan@reddit
100%. I swear that as time has went on, I hear more devs on stand-ups spending way more time updating packages, frameworks, and tooling without actually building features and fixing bugs. And some even seem to just enjoy that shit too.
If I were going to spend 3/4 of my time just installing packages, upgrading packages and frameworks, and having meetings about installing+upgrading shit then I would have just sold my soul and went into IT systems maintenance a long time ago (no shade to any IT guys reading this, it just absolutely doesn't interest me)
loretta7inker7640@reddit
it's like the excitement fades with experience
SplendidPunkinButter@reddit
Devs: Don’t reinvent the wheel!
Also devs: Wow, check out this new framework/programming language!
WildWinkWeb@reddit
Something lost on everyone in SaaS, tools and devops is that 90% of people HATE learning new things.
remy_porter@reddit
I like learning new things if they’re actually interesting to learn! That’s never true for frameworks, or almost never. Once in a blue moon somebody drops a framework that is legitimately exciting, that takes a complex problem and lets you describe solutions with elegance and clarity. These are rarer than unobtainium.
Mostly frameworks solve the specific problem the originators had pretty well, and then codifies their opinions about software as “best practice” and forces everyone to do things their way, even if it’s not a good fit.
PlanOdd3177@reddit
I'm still early in my career but I agree with you. The tools are just an obstacle to solve the real business problems. I just want something that's been tried and tested and stays out of my way as much as possible.
engineered_academic@reddit
Hiring for tool X is something that really needs to be considered at and beyond the startup level. It doesn't matter how well put together your code is if you can't find additional developers to help you scale and maintain the application.
Sweet_Witch@reddit
So you made stuff in some fancy new frameworks and other developers were later unhappy to maintain some obscure and after some time not so new framework?
SureConsiderMyDick@reddit
lol, no what. Read better
Sweet_Witch@reddit
Read better what?
Op "Early career I was pushing to try new things"
Op clearly states here he was pushing these things.
AralSeaMariner@reddit
What's your point though? OP is saying that was a mistake in retrospect. You're just here to rub his nose in it?
JuiceChance@reddit
It is called experience not boredom.
DixGee@reddit
The variety of frameworks and tools also makes it hard to prep for interviews. People expect you know every new thing that comes up in the market.
baezizbae@reddit
And have more years of experience than the thing has existed for in those really special cases
Working_on_Writing@reddit
This is one of the many reasons I moved into management. I just got tired of fighting with whatever framework is so hot right now and unstable, and having to listen to the latest hype about whatever and pretend I care. For me a framework or a library or an OS is a means to an end, and that end is solving a problem for somebody.
When I'm trying to solve problems I want tools which just work and feel good to use.
I think that's actually what most people want, for example, Stripe won in payment processing not be being the hot new thing but by doing what needed to be done in a straightforward way which just works.
As with everything there are loud voices at the edges making things seem more fraught than they are.
UXyes@reddit
I’m of the exact same mindset.
JohnWangDoe@reddit
favorite tools?
throwaway0134hdj@reddit
I have significantly less year experience but that’s also been what I have been trend towards. Tried and true tooling vs new gimmicks mix. A ton of things can be solved with just simple SQL and Shell commands.
amendCommit@reddit
My docker compose scaffolding has barely changed in nine years, yet my bosses still whine about "introducing new tooling" and eject code from containers before complaining it's broken whenever they get the occasion. Boring and predictable is subjective. To me shipping containers is boring and predictable, to some others it feels like a new and brittle workflow.
Ok-Pace-8772@reddit
If containers is brittle for someone they have skill issues. Simple as that.
jacobpackert@reddit
100% this. Boring is great. Instead of trying to be clever, it’s often worth it to just say “what is everybody else doing” and then do that as well.
papawish@reddit
Unfortunately that is why many companies do ageism.
While a company needs a strong backbone of reliable services, and you could be part of that. What's really valued by the bureaucracy is the will and ability to lean into each new CEOs brainfart.
They still need you, but that might impact your career negatively.
Expert-Reaction-7472@reddit
CEOs dont care what front end framework is FOTM
papawish@reddit
Recently proved false by forced adoption of MCP (which is garbage) and OpenAI buying Astral.
CEOs think they know it all and have no problem forcing tech on employees because someone at an after-party in Vegas told them it was the future.
tcpukl@reddit
I think this boredom is one of the reasons I like game Dev.
Throughout my career we've had new challenging hardware coming out every few years.
caprisunkraftfoods@reddit
Trueeeeee. Been feeling this vindicated again with Astral's buyout from OpenAI. uv has looked neat from the first moment I saw it but I simply didn't want to jump ship to yet another python environment manager that will be abandoned. I was just starting to look at making the switch on some of my active projects and then the buyout happened.