Open Kenttleton opened 6 years ago
Every error sends @Kavinci a:
Sounds perfect lol
I vote for delivery by pigeon foot
What do you guys think about leveraging our GraphQL server by sending caught clientside errors to it and then redirecting to a specified error-purpose channel within our slack community?
That sounds noisy. Ideally you want to catch as much info around exceptions as possible to help identify what conditions/contexts caused them. As well as the frequency of the occurrence, associated user comments (if feedback was provided), and that exception's history.
Honestly, dropping in the free version of something like Sentry
would probably make the most sense. If you want to expose the data, you can share individual exceptions, or get creative with its API and share features.
Maybe...It very well could be. It also could be non-noisy and pretty helpful too without much investment, we just have to get a little creative :)
We already center around Slack, so why not make it convenient for us as a group? Leveraging our GraphQL server, which is compounded with Apollo Client and does a great job handling errors out of the box, and having it ping the channel using a small bit of logging logic serverside would make a lot of sense wouldn't it?
We are a really small group of people who like to build stuff. We won't learn as well if we only just consume, right?
Public note to self: Talked with sales team over at TrackJS and they agreed to give us full access to their error logging platform since we are open source. All they asked was to include a small note in the footer of the site saying like "tracking errors thanks to TrackJS".
As we are building integrations with various services we should think about how and where to log errors since we are using a hosting service without a file storage for a hard copy of the log. This may be a good use for a database table of errors that are accessible from the admin/user section of the website.
Thoughts, comments, concerns?