Here is a link with info on a 1.3 release of the PragProg Phoenix book: https://forums.pragprog.com/forums/393/topics/14820
Hi @chrismccord thanks for Phoenix! I’m using Phoenix
1.3.0-rc in an umbrella application.
My structure is:
foobar_umbrella - foobar - foobar_web
So within my Phoenix app, naming looks like, for example an API controller:
defmodule FoobarWeb.Web.Api.V1.ArticleController do ... end
Since the MyApp.Web alias has been changed to MyAppWeb, does this mean code will look like?
use FoobarWebWeb, :controller
How do you recommend I approach this update?
Also, I’m using Phoenix from NPM to connect to my Phoenix Channels using websockets, can I update the package from npm to get v2 of the channel protocols?
The module of your web app, would be
FoobarWeb, with calls to
use FoobarWeb, :controller
The package hasn’t been updated yet, but will for the final release. You can specify a path dep in your package.json or vendor it today.
Question about naming tables: How will tables be named when two different contexts has very same schema with different fields but also with clashing ones? Should I avoid situations like this one by name them differently, have one big table or split them somehow?
Rest looks very, very promising. Thanks a lot!
If you have two database-backed schemas in different contexts that are entirely separate, but named the same, then you will need to give one of them a unique name. You can do this by passing the
--table flag to the generator.
We would need to talk about your specific use case to say more, but please check out the new context guide which will give you ideas how to handle this kind of scenario. In the guide we create an
authors table for author specific information for our CMS, rather than polluting our Account’s “users” table with author specific data.
Thanks for Phoenix. Thanks turning all the criticisms about contexts nil. This project is something to be proud of.
Thanks, I passed only an imaginated example here. I’ll read guides carefully. Thanks again for the great work you’ve made with the core team for the Elixir society
Congrats and thanks to @chrismccord and the #phoenixframework team! Started playing with it for couple of weeks now… it looks pretty solid!
Just tried out the new installer, and here’s a quick note about the
test directory structure (for people upgrading from rc.2):
- the module names are now
- the files in
test/support(conn_case, etc.) also have the
Thanks for all the hard work everyone’s put in.
In case it’s useful, you can see the changes to generated files between the two RC’s here: https://www.phoenixdiff.org/?source=1.3.0-rc.2&target=1.3.0-rc.3
Or, if you’re upgrading from 1.2, you can see that at https://www.phoenixdiff.org as well.
I quite liked the table namespacing. Going to miss it as default.
mix phx.gen.html Accounts User users email:string --table accounts_users
Thanks to Chris and the team! been using 1.3.0 for personal projects and loving it so far. It took some time to wrap my head around the new folder structure, but it makes everything more clear
Awesome work, Chris! Kudos to you, and to your team! Thank you!
I’m not quite sure what this comment is aiming for. What kind of feedback are you after around this release?
As promised, 1.3.0 has landed! Check out the officinal announcement here:
Elixir 1.5 released. Phoenix 1.3 released. Now it’s the time to convince my PO that I have to do this upgrade in our app
Thanks Chris and to the Core team!