On the front end: Choo is really solid and very little to learn.
Ask HN: Fastest/easiest framework to build a web application in 2017?
21–30 of 40 posts
Re: Ask HN: Fastest/easiest framework to build a web application in 2017?
#22Next.js/react/postgraphql/postgres/beanstalk.
My stack of choice as well. The graphql part definitely adds to the learning curve for new devs but I think it’s the future. Checkout github.com/apex/up. Now you won’t even need the beanstalk part. All you need are RDS instances.
Re: Ask HN: Fastest/easiest framework to build a web application in 2017?
#23One thing to consider - which one has a superb documentation?
Re: Ask HN: Fastest/easiest framework to build a web application in 2017?
#24If your goal is fastest to get a production ready MVP built, go with Rails or another similar mature stack in a different language. Chances are the majority of problems you encounter will have been solved by someone.
If you want to learn something new or want to be on the hype curve, don't expect to have something production ready fast. Expect to have challenges learning how things work beyond a blog/to-do list tutorial for new-to-you tech.
Re: Ask HN: Fastest/easiest framework to build a web application in 2017?
#25I'd be interested in views on fastest to develop with - and not relative to what you are most experienced at
I think your expierence in a system/framework or programming language is a key factor in "fast" development and to be honest, the question for the "easiest and fastest" framework is totally wrong and you won't get nothing out of this question. All answers will come from people with different expierence levels in programming, frameworks, projects and many other things - all answers are biased towards the expierence of the person who answers your question.
Re: Ask HN: Fastest/easiest framework to build a web application in 2017?
#26I would still recommend Rails or Laravel(PHP).
Both are similar so choose the one you know/fits you best.
nothing to something:
Laravel you can go from zero to auth + dashboard in like 5 minutes. There are lots of packages to speed things along.
Some of the items I'm recommending are paid, but they are well worth the value they provide and support the creator of Laravel and has enabled him to work on Laravel full time and hire another full time developer to work on Laravel and these products.
There is also Laravel Spark that's basically SaaS in a box, it's $99/site but worth it for a head start.
Laravel has a really great community and great tools that make it super nice/fun to use.
Laravel Valet, quick and easy local dev env. for OSX.
Laravel Forge, it's a SaaS that will spin up servers on (Digital Ocean, AWS, +others) and allow quick deployments plus easy SSL setup through LetsEncrypt. (There is also envoyer.io for zero downtime deployments.
Laracasts.com for learning.
Re: Ask HN: Fastest/easiest framework to build a web application in 2017?
#27Flask (python) very minimalistic but quick and nice documentation .
Re: Ask HN: Fastest/easiest framework to build a web application in 2017?
#28Since you are asking about "rapidly building a web application" I think you also need to consider your hosting platform. If you use a framework that has full support on one of the big PaaS services, you can save a lot of time configuring and maintaining your production and development environments. For my most recent web app, I chose Python/Django on Google's App Engine. I spent less than 15 minutes configuring the d…
That is, IMO, one of the big areas that Phoenix falls flat on. Sure there are buildpacks that you can use for Heroku deploys but AFAIK there is only one hosting service that delivers a truly easy PaaS deploy process for Phoenix apps and that is Gigalixir (and from what I hear they aren't cheap).
Re: Ask HN: Fastest/easiest framework to build a web application in 2017?
#29Rails, without a doubt. Don’t waste your time on Django and all the Node stuff is just a mountain of distractions. (I’ve been doing this for over a decade. At one point I was a Django zealot and avoided Rails like the plague. Boy was I wrong)
Can you explain why a python developer should make the switch to rails for web apps? The difference between Django and rails is pretty small from my oberservation. They have similar ease of use and performance characteristics.
Been in the same boat. Started with Python but transitioned to Ruby / Rails for web dev.
Django is a good framework. Ultimately why people choose it or something else is about your own taste. The reason I didn't choose it is that Django give users more freedom i.e. configuration over convention which for me leads to https://en.wikipedia.org/wiki/The_Paradox_of_Choice (Maybe things have changed?) It's strange because Python is the opposite in terms of philosophy. Rails goes the other direction and is very opinionated which is also strange considering Ruby's philosophy is the opposite i.e. 10 million ways to do the same thing.
The next issue is community consensus. Python doesn't seem to have it. (Again things may have changed.) You can see this with packaging and the 2x vs 3x war. Not sure if Python has finally standardized packaging yet but it's hard not having a good package system everyone uses once you get exposed to ruby gems, npm modules, and even maven.
For the record, I'm not saying Python is bad. Just giving reasons for why I personally do not use it unless I'm forced.
Re: Ask HN: Fastest/easiest framework to build a web application in 2017?
#30There's no real answer to this question. Use a stack a few times in a row from scratch and it's now the most prudent stack for you to use on your next project hands down.
Also, focusing on "rapidly building" possibly over-fixates on Day 1 when it's Day 60+ where your stack choice is really going to matter once you need to respond to changes in business requirements.
However, there are a few high level decisions I'd consider pretty uncontroversial:
1. Relational databases like Postgres tend to be ideal for fast development since you can respond to requirement changes by atomically migrating your entire data layer to a different schema. This is pretty hard in NoSQL databases for example. You can also encode constraints in the data layer and then the application layer can just depend on them always being true instead of trying to encode them into the application layer.
2. Keep the application server as stateless as possible. For example, use something like S3 instead of going the tempting route of saving to the filesystem. A filesystem dependency is convenient on Day 1 but now you cannot spin up more than one server nor launch an application off of that machine. For similar reasons, it's also nice to host your database in the cloud like on RDS.