4 ms·

Next wave web development: Elixir+Phoenix vs. Meteor+Angular2+TypeScript

17 points by blunte 10y ago

My Background: Java Servlets/JSP, then Rails, and now Clojure fan. Elixir+Phoenix is very attractive, but Meteor+(some DOM system)+(some ES6-ish language) seems very nice. The ultimate question, as a polyglot and a "get it done easier and not regret it too much later" kind of person, is which is worth the plunge? I have a project right now to do, and I must commit.

zer00eyz 10y ago
My question is "does it matter". In 24 months is everything your suggesting going to be replaced with something newer? The answer is maybe, and any guess is as good as any other. Pick what you like, or think is going to make you happy, validate your choice and leave your self notes as to why. Come back to it later and see if you rational of today holds true in the future!
dimitri-gnidash 10y ago
Is that a personal project? I would be careful using any of the cutting edge technologies on the commercial projects. Imagine the maintenance issues the developers that follow you will have if you do not pick the stack correctly. For commercial projects, I would urge you to use something that is well-adopted React or Angular 2. If the project is personal, then why not just have two smaller projects to utilize both tech stacks or even combine them. Nobody depends on you, and you have all the flexibility to play around and write the software as you please. If you do end up playing with both, share your experience, and tell us which technology you prefer.
anilgulecha 10y ago
For most web-app use cases your prior Rails knowledge would be a good starting point for building out your backend/api. A simple popular and well-supported framework (Angular/React) would make for a very effective frontend.
nerdywordy 10y ago
I can only speak for myself, but I find Elixir+Phoenix an awesome dev environment. It's a fantastic first functional language to get started with.
sergiotapia 10y ago
Meteor is dead, it really pains me but it's done. 1. Stagnated feature set playing catch up to other boilerplates. 2. You will run into performance issues at 100 users. 3. Still no real SQL support. Instead they're going all in on Apollo (graphql). 4. Brain drain. Too many influential package maintainers have left Meteor entirely, including Arunoda. Go with elixir and Phoenix - you won't think about performance until you get acquired lol
devdad 10y ago
Angular 2 with TS for front-end is an OK choice and works well for us. Choose another backend though, Meteor has seen it's prime. Chances are that a good old Express server will be more than enough.
karmajunkie 10y ago
These are not mutually exclusive options. You can use the heir to meteor (Apollo-client[1]) with elixir and phoenix quite easily. Absinthe [2] is shaping up to be a very complete implementation of the graphql spec, with proof-of-concept of non-spec features like subscriptions. I've been very happy with this combination, and it allows you to swap your front-end easily for another implementation (e.g. go with ember or react instead of angular). [1] https://github.com/apollostack/apollo-client https://github.com/apollostack/apollo-client [2] http://absinthe-graphql.org/ http://absinthe-graphql.org/
bhdzllr 10y ago
Controversial opinion about Meteor by CEO Geoff Schmidt in another "Ask HN" 21 days ago: > Meteor isn't going to take over all JavaScript development the way Rails took over Ruby development, at least not anytime soon. That's just not how the JavaScript ecosystem works. However, I think it will be the #1 full stack JavaScript framework for a long time to come and will continue to be a really great option for teams that want to build JavaScript apps quickly, especially apps that have a realtime or collaborative element. [1] [1] https://news.ycombinator.com/item?id=13303345 https://news.ycombinator.com/item?id=13303345
pmarreck 10y ago
If you're optimizing for "fun and functional", go Elixir/Phoenix and don't look back. You'd still need some frontend tech, but there's a good amount of integration tools out there already for things like Elm, React, probably Meteor, Angular etc.

17 points by blunte 10y ago

zer00eyz 10y ago

My question is "does it matter". In 24 months is everything your suggesting going to be replaced with something newer? The answer is maybe, and any guess is as good as any other. Pick what you like, or think is going to make you happy, validate your choice and leave your self notes as to why. Come back to it later and see if you rational of today holds true in the future!

dimitri-gnidash 10y ago

Is that a personal project? I would be careful using any of the cutting edge technologies on the commercial projects. Imagine the maintenance issues the developers that follow you will have if you do not pick the stack correctly. For commercial projects, I would urge you to use something that is well-adopted React or Angular 2. If the project is personal, then why not just have two smaller projects to utilize both tech stacks or even combine them. Nobody depends on you, and you have all the flexibility to play around and write the software as you please. If you do end up playing with both, share your experience, and tell us which technology you prefer.

anilgulecha 10y ago

For most web-app use cases your prior Rails knowledge would be a good starting point for building out your backend/api. A simple popular and well-supported framework (Angular/React) would make for a very effective frontend.

nerdywordy 10y ago

I can only speak for myself, but I find Elixir+Phoenix an awesome dev environment. It's a fantastic first functional language to get started with.

sergiotapia 10y ago

Meteor is dead, it really pains me but it's done. 1. Stagnated feature set playing catch up to other boilerplates. 2. You will run into performance issues at 100 users. 3. Still no real SQL support. Instead they're going all in on Apollo (graphql). 4. Brain drain. Too many influential package maintainers have left Meteor entirely, including Arunoda. Go with elixir and Phoenix - you won't think about performance until you get acquired lol

devdad 10y ago

Angular 2 with TS for front-end is an OK choice and works well for us. Choose another backend though, Meteor has seen it's prime. Chances are that a good old Express server will be more than enough.

karmajunkie 10y ago

These are not mutually exclusive options. You can use the heir to meteor (Apollo-client[1]) with elixir and phoenix quite easily. Absinthe [2] is shaping up to be a very complete implementation of the graphql spec, with proof-of-concept of non-spec features like subscriptions. I've been very happy with this combination, and it allows you to swap your front-end easily for another implementation (e.g. go with ember or react instead of angular). [1] https://github.com/apollostack/apollo-client https://github.com/apollostack/apollo-client [2] http://absinthe-graphql.org/ http://absinthe-graphql.org/

bhdzllr 10y ago

Controversial opinion about Meteor by CEO Geoff Schmidt in another "Ask HN" 21 days ago: > Meteor isn't going to take over all JavaScript development the way Rails took over Ruby development, at least not anytime soon. That's just not how the JavaScript ecosystem works. However, I think it will be the #1 full stack JavaScript framework for a long time to come and will continue to be a really great option for teams that want to build JavaScript apps quickly, especially apps that have a realtime or collaborative element. [1] [1] https://news.ycombinator.com/item?id=13303345 https://news.ycombinator.com/item?id=13303345

pmarreck 10y ago

If you're optimizing for "fun and functional", go Elixir/Phoenix and don't look back. You'd still need some frontend tech, but there's a good amount of integration tools out there already for things like Elm, React, probably Meteor, Angular etc.