Microsoft Edge handles <object> Oddly

Ignore the bottom, original report. The user was returning a 204 response, which understandably prevented a follow-up GET.

That being said, Edge still makes two trips, for each object[data], unlike Chrome and Firefox. Edge does a HEAD, then GET. Chrome and Firefox do a simple GET.

---

When Microsoft Edge encounters <object data='/foo'></object>, it fires off an initial HEAD request to gather meta-data about the source. Following this, a GET request is made for the content of the source.

Because the initial request is a HEAD request, and thus explicitly forbids body content, it seems reasonable to respond to this request with a 204 No Content message. Oddly, Microsoft Edge treats the 204 as reason to terminate the process entirely, thus making no follow-up GET request for content.

Repro Steps:

1. Download Object Data Test.zip
2. npm install
3. node index.js
4. Navigate to http://localhost:3000/
5. Observe network traffic

Internet Explorer, and Microsoft Edge take the two-step approach.
Google Chrome, Mozilla Firefox, Opera, and Brave make a single GET request.

Originally reported on Stack Overflow.