ttfb
ttfb copied to clipboard
phip1611
also add content download time (requires http/1.1 header parsing)
will be more complicated, because it needs HTTP header parsing.. content encoding could be chunked for example..
I need to know the contents length, because the "read_to_end" (EOF) doesn't work well. Some servers even do not send the EOF (program gets stuck)
Hey there!
Firstly, great work on the library. It's been incredibly useful so far. Is there any plans to implement this feature at some point? It would complete the main HTTP timings and it would be incredibly useful :)
Hey. Thanks for the interest in my library.
Generally speaking, I'd love to have full HTTP/1.1, HTTP/2.0, and HTTP/3.0 support. However, for the nature of the library, I need a more low-level access to the corresponding implementations which I'm not sure yet all popular crates for HTTP 1/2/3 offer.
Further, those protocol implementations on top of the TLS stream need to provide me with the following functionality:
- fetch first byte
- fetch all data
I think it might be possible if the corresponding protocol libraries offer the proper abstractions. I try looking into it in the next 14 days.
I'm sorry to tell you that, but I do not have the capacity to work on this, at least in the next two months.
I'm open for (good) pull requests/contributions tho.
I think a way forward is to properly implement a Http/1.1, HTTP/2, and HTTP/3 (see #35) API adapter compatible with the current connector-style code. This adapter must enable it to load the first byte separately from the full resource.