Ask HN: What's your stack for your side projects? | Hacker News Reader

Ask HN: What's your stack for your side projects?

14 pointsuptownfunk33 comments

Looking to build some simple SaaS ideas out, trying to figure out what tools are currently in use and how I can get started without wasting time on learning things I don't need or that are obsolete or that I won't use. Tired of being that guy who has ideas but can't make them a reality.

View on HN

Been using same stack for 10 years; php, mysql and javascript. A friend calls it ghetto coding however this unruly stack could have paid for my house.

Most side projects are about helping people who think pancakes when you talk stack. This is my advice for what it’s worth;

Solve a real problem. Don’t create something you can’t support. Be nice to people, they are the ones who will pay you. Recieve input but ignore people who tell you work is rubbish (we’ve all been there).

Good luck.

I have to say, no matter how hate php gets. it's a great environment to get the shit done.

Couldn't have said it better myself. You get it. :)

You should check out Laravel. Laracasts.com is a great place to start.

This.

Your users don't know or care about what stack you use as long as it solves their problem.

My most profitable side projects also happened to contain the worst code.

I think developers can be categorized into 2 types - engineers and entrepreneurs.

The first type is obsessed with the programming and technology aspect of things (won't ship until the code is perfect, clean and optimized).

The second type just wants to ship the damn thing, close the deal, and secure the money.

Don't worry about what's "obsolete". Worry about what you can use to achieve your goals. If you want to build webapps on an AS/400 using RPG/IV, then go for it. This industry is already way too "fad driven" where everybody is walking around thinking that only the latest, shiniest, newest, cutting-edge, "whatever" matters.

That's not to say that some tech doesn't have more merits than others, but you don't necessarily have to worry about always using "the cool thing of the day".

All of that said, I primarily use Groovy + Grails, Postgresql, Bootstrap, and jQuery for building webapps. For backend microservices I mostly use Java with Spring Boot.

Yes, there are "sexier" things out there, but this stack lets me accomplish my goals, and I know it well and am very productive with it. And for me, that's what matters most.

Tired of being that guy who has ideas but can't make them a reality.

If there's a stack you already know well, I'd say stick with it. If not, I'd suggest picking basically anything reasonably modern/mainstream for webapps. So, probably not actually RPG/IV, or assembly, etc., but anything that's used in this domain, and where there is plenty of documentation, community, etc. to support you. Use PHP, Python/Flask, Java/Play, Scala/Play, Node.js, Ruby/Rails, Groovy/Grails, Drupal, Joomla, etc.

You must really like Grails. If you do a Search All by Date on "Grails" at the bottom of the page, 4 out of the last 5 results are your comments promoting Grails as a solution to various Ask HN queries.

  Ask HN: What's your stack for your side projects?
  mindcrime a day ago comments 

  Ask HN: Rails or Flask for new API project in 2019?
  mindcrime 6 days ago comments 

  Ask HN: What is your favorite tool to build web startup?
  mindcrime 7 days ago comments 

  Ask HN: How many languages/frameworks do you regularly use?
  mindcrime 17 days ago comments 

  Ask HN: Go-to web stack today?
  mindcrime 25 days ago 

  Ask HN: What´s your favorite JVM Language?
  mindcrime a month ago 

  Ask HN: I'm a front end dev looking for a back end language. Which one? (No JS)
  mindcrime 3 months ago 

  Ask HN: What's your goto for creating a REST API?
  mindcrime 3 months ago 

  Ask HN: A tool that tracks various websites for keywords and give you updates?
  mindcrime 4 months ago

And your point is?

I'm not "promoting Grails as ..." in most cases. Notice that most of those questions are phrased in terms of "What do you use for X?".

People ask questions, I answer honestly, just like I expect everyone else to. Do you find something exceptional in that?

"Things I don't need or that are obsolete or that I won't use"

That's a silly way of thinking.

* Define the problem to solve

* Look at potential solutions

* Pick one that fits

* Make stuff

Before you pick a solution (Language, framework, tool) you need to define the problem, because all solutions are better at and for some things than others.

So, what problem are you trying to solve?

I'm just trying to get an idea of what's being used out there to build a rapid prototype of some simple ideas. Trying to avoid something exotic, just trying to see what will get me the fastest from idea to prototype. I understand that this will be different for many people, but I imagine tools like assembler/C++ wouldn't be many people's first choice for building a SaaS prototype.

Sure, I get that, but there are as many opinions as there are tools, and you'll get lots of people fighting each other over what's best.

Assembler, maybe not, but C++ makes perfect sense, in some cases.

And that's the problem.

You haven't specified what kind of problem you're solving, so it's almost impossible to suggest ideas.

It's kind of like saying "I want to build something, what tools should I buy?" (In regards to wood working).

I get tons of mileage out of ColdFusion, JavaScript, and Object Pascal. I make things for web, desktop, servers, GUI and cli, but you could do the same with PHP, TypeScript, and Java.

Or Nodejs, Dart, and Go.

I've used all of those in various combinations, and I gotta say that CFML + JS + Object Pascal is the most productive combo I've come across, but that's what's worked great for ME, and I only picked that stack after working with a whole bunch, and most of them professionally at some point.

Figure out what you want to build, then figure out what tools to use to make it.

The most important thing though, is that you start making stuff.

You won't "waste" time by learning Java/Nim/JS/Go/Pascal/TS/whatever, even if you end up switching languages later.

Programming isn't about memorizing syntax, it's about solving problems, creating solutions, and there's an awful lot that applies to just about any language once you get down to actually coding stuff beyond the basics.

Incredibly helpful remarks, thanks!

Glad it helped!

Now go make stuff :)

Exactly, use tools or language to solve problem

>Tired of being that guy who has ideas but can't make them a reality.

Then you need to build something. Doesn't matter what you use if you never build.

Personally I just use whatever I feel like learning with. My current fun project is using golang, and react.

Golang + react + Postgres and it’s open source You can find it here https://github.com/godiscourse/godiscourse

Rails, devise, Vue.js, bulma. What I like with rails&devise is the possibility to have a base app running with authentication, db migrations, a nice standard folder structure and efficient pages, all this in an afternoon.

My speed stack: Front end: jQuery and Bootstrap Back end: Parse Server DB: MongoDB (mLab) Hosted on Heroku.

Very little learning needed. Parse Server will abstract out all the DB stuff too. Things like Angular make it easier in the long run, but you can always refactor.

IMO the trick is just to set a minimum target and aim to scale only 10x that.

What are you aiming for to "get this off the ground"? Side income of $1000/month? Charge them $5/month? That means you're aiming for 200 users/month. Build something that supports 2000 users/month. Upgrade it as problems appear.

React statically hosted on Github pages. Analytics on Google Analytics or Amplitude.

Whenever I can, no backend. If I need, either Google spreadsheet, Firebase, App Engine... let's say I choose the one with less maintenance, depending on what I have to do.

Usually a new stack if I am learning something.

If it's not, I pick an old project of mine, rename the project settings/README, remove the contents of the Src/ directory and start repurposing it.

Node, React, and Go.

Backend Go, frontend React.

I want to check our Erlang/Elixir, and maybe Ember for the frontend. Trying to figure out the most efficient way to build things!

Pretty much the same here. I have tried Elixir/Phoenix and found it worthwhile. Just need to keep using it to be more proficient with it.

For the front-end, I've pretty much switched over to Vue either using only local state or vue-stash. It's much lighter weight for smaller projects and I like the single file components with default separation of layout and code.

Can you share your project address here? So we can do code review together

Rails, and postgres. Elasticsearch if there is a search layer to it. For UI: JQuery.

Laravel.

Slowly learning Vuejs for frontend.

Same, I still prefer javascript/jQuery but am learning Vue.

Backend: Nodejs, postgresql.

Frontend: Not worked on it yet but will use react.

Go, Python, C, Postgres, Redis

Could you elaborate when you choose Python and when you choose Go? Does these two languages complement themselves or they are substitutes?

ruby bash terraform, but i learn new things all the time

someday, golang will replace ruby

I’m using golang now, Golang template is terrible, rails is better than it. But golang is good for back-end

golang templating is god awful, but i make heavy use of gomplate which leverages golang templates internally

Reply on news.ycombinator.com