Warning: This is a (very) delayed announcement! ;-)
oEmbed is an open standard for embedded content. It allows users to embed some resource, like a picture or a video, in a web page using only the resource URL, without knowing the details of how to embed the resource in a web page.
oEmbed isn't new stuff. It was created around 2008, and despite not being widely supported by content providers, it is supported by some big players, like YouTube, Vimeo, Flickr and Instagram, making its usage highly viable.
To support the oEmbed standard, the content provider just needs to provide a simple API endpoint, that receives an URL and a few other parameters, like the maximum allowed height/width, and returns a JSON or XML object, with ready-to-use embeddable code.
The content provider API endpoint can be previously known by the oEmbed client, or auto-discovered using some meta tags added to the resource's HTML page. This is the point where the standard isn't precise enough: not all of the providers support auto-discovering of the API endpoint, neither all of the providers are properly listed on the oEmbed specification. Proper oEmbed clients should try both approaches, looking for known providers first, falling back to auto-discovered endpoints, if possible.
Each of the Python libraries for oEmbed decided to follow one of the mentioned approaches, without caring about the other one, failing to support relevant providers. And this is the reason why I decided to start writing pyoembed!
- ← Newer
- Older →