You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Should responseStart measure the time to first by for any response or only measure non-informational HTTP responses (specifically focused on HTTP status code >= 200)?
Currently the spec is ambiguous referring to "first byte of the response" which could be interpreted to include HTTP 103 Early Hints and other informational responses.
responseStart - the time immediately after the user agent receives the first byte of the response from the server
The responseStart is widely used to measure TTFB from the user's perspective. Measuring to the start of an Early Hint response can pollute this colloquial interpretation of this measurement. I would propose that responseStart should ignore informational responses (HTTP Header frames with a 1xx response code) and only relate to the first byte of an authoritative content response (HTTP >= 200)
The text was updated successfully, but these errors were encountered:
Should
responseStart
measure the time to first by for any response or only measure non-informational HTTP responses (specifically focused on HTTP status code >= 200)?Currently the spec is ambiguous referring to "first byte of the response" which could be interpreted to include HTTP 103 Early Hints and other informational responses.
The
responseStart
is widely used to measure TTFB from the user's perspective. Measuring to the start of anEarly Hint
response can pollute this colloquial interpretation of this measurement. I would propose thatresponseStart
should ignore informational responses (HTTP Header frames with a 1xx response code) and only relate to the first byte of an authoritative content response (HTTP >= 200)The text was updated successfully, but these errors were encountered: