This repository was archived by the owner on Nov 19, 2024. It is now read-only.

Description
Sorry to create one mixed issue, just wanted to give some feedback as someone who used heroku for a long time and saw your "render VS Heroku" article and thought it would be worth giving it a try.
Sorry if I missed something, I just spent an hour or two deploying a test deploy on render.
Compared to "heroku" I noticed a few things:
-
you can't tail a service without knowing its "id", that's very counterproductive as the name of the deployment is very easy to remember, but the id is impossible. A few other commands I tried on the command line ended up with me having to look for a deploy "id" as opposed to just typing the deploy name which is easy to memorize.
-
there is no "render service logs" to pull like the last 100 lines of a log or so, the tail function also doesn't show logs before I asked to tail, so it's hard to check what just happened if you had the terminal closed when it happened.
-
render asked me to get an API KEY on the dashboard instead of following the regular flow that command line apps like "heroku login" do
-
render using colours on the command line seems a bit OTT ( personal opinion of course )
-
every deploy on Heroku will have it's own IP address, at least for all I know when creating multiple deploys they never go out with the same outbound IP which can be very beneficial in some applications
Compared to vercel:
- One of the things I love the most on vercel is the "preview deployment", so as soon as you create a PR they spin a temporary environment and provide a ( login protected ) preview environment, so other people can try it out while reviewing the PRs