|
|
|
|
|
by simoelalj
2066 days ago
|
|
We combined different approaches to make it work: First, inspecting all network requests that are happening on the web client. That helped to inspect all requests made to read a page and edit content (a lot of data is transiting, maybe that's why it's slow on the startup...). And then I realized there were few existing open-source projects that already done that (https://github.com/kjk/notionapi and https://github.com/jamalex/notion-py), but there were lacking an HTTP API endpoint so that we can easily plug them that to Zapier/Integromat or a Webhook. These two repo didn't had a way to convert a Markdown to a Notion content - which would make it complicated to know every type of blocks that makes a Notion page. Finally you needed to be tech-savvy to be able to deploy the open-source projects and there wasn't any simple integration SaaS out there – so it was limiting a large section of people who use Notion extensively, and are familiar with automation tools. |
|