Activity
Posted in Scraping Web Pages with Ruby Discussion
I haven't seen the next episodes yet but up until this point I'm wondering if a Scraper class is really a recommended way to do this? It seems like you are just replacing the Nokogiri API with your own API without a clear benefit. Are added layers of abstraction helpful later on? Maybe for testing? Or by reducing the API surface area maybe?
What's great about the puma plugin is the cost savings when running on something like Heroku. Only one dyno needed for small apps including a Queue. Something that was quite expensive before because of the extra worker and extra Redis.
Since it adds so many tables I was wondering if it works with a separate database. Like a second SQLite. Maybe even a SQLite just for the queue while the actual application runs with PostgreSQL. Could be useful in situation with through-away jobs.
This is a great series. Seeing how you approach open source and pull requests. Thanks!
What happens when passing stored: false?
Posted in How to upgrade to Rails 7.1 Discussion
I'm sorry but this is a bit bare bones. Not the depth I came to expect from GoRails. This should at least cover the config/new_framework_defaults_X files for a complete upgrade flow. Maybe some tipps and good practices about upgrading 3rd party gems along with rails. Upgrading rails can be quite tricky for production apps and this video covers only the best case scenario. Someone may even be on 6.x and may wonder how to upgrade to 7.1.
However, I can see how this video is still from value for someone who just wants to tinker with 7.1 beta quickly. But then I can just follow the official (edge) guides. What I'd like to see from a GoRails video are are insights from long-term users/professionals who provide information that goes way further than the official guides. I'm thankful for what you guys provide so I hope this doesn't come across to harsh.
Adding to this, it would be nice if the initial rails app is a first commit and the episode specific changes a separate one.
This is a great summary and takes some of the scare away I had regarding Net::HTTP. However one of the reasons I always went with Faraday was the error handling. Net::HTTP seems to have all kinds of different exceptions I do not really care about. Maybe you can include the error handling into one of the upcoming episodes.