4 ms·
If you're building a basic webapp, you shouldn't be using a front-end framework at all! React has one of the gentler learning curves of the front end framework
by xiaoma 10y ago
If you're building a basic webapp, you shouldn't be using a front-end framework at all!
React has one of the gentler learning curves of the front end frameworks, but it's still a lot more work than just rendering everything server-side and adding whatever minimal effects you need with jQuery.
I second sandGorgon's recommendation to use Rails. PHP is another good option. It's absolutely amazing how many successful new software companies are being built off of WordPress and/or Laravel.
I say this as someone who was using React in February 2014 and who loves to geek out over newer more powerful technologies. The thing is, your users don't care. Groupon was launched as a crappy WordPress site + a mailing list, built way past its breaking point, migrated to Rails, broke again and was a $5B company in 5 years. Many, many startups with harder-core technologists started years earlier still haven't reached where they are now even with a perpetual string of problems and failures from their IPO until now. Teespring started on Rails long after that was cool and they're a juggernaut.
Use something productive and ship it.
- needcaffeine 10y ago> Use something productive and ship it. This! Can't agree with it more. Your clients don't care if your backend is written in PHP, Rails, or Go. Use whatever is your best tool in order to ship it quickly. It's "acceptable" to have a reasonable amount of fun, but don't waste time learning a whole new stack at the risk of validated learning. Unless it's supposed to be a fun project to begin with.
- thenomad 10y agoI couldn't agree more with this. However, I'm curious: at what point do you think the flavour-of-the-month JS frameworks actually do make sense for a development team / company? This might just be me being old and set in my webdev ways - I remember when PHP4 arrived on the scene, for goodness' sake - but I do struggle to see the business case for a lot of the current JS-everywhere, a-million-tools-for-everything environment, and I suspect I'm missing something key.