due to the fact that both queries were initiated through the same connection
This is where your assumption falls short
HTTP1.1 prioritizes socket usage. Every HTTP request establishes a new socket. According to protocol guidelines, the closure of the socket signifies the end of the HTTP request.
Content-length
An HTTP request can be marked as complete before the TCP socket closes. When the response includes the Content-length header, the size of the HTTP packet matches the specified byte count in the header. However, it's important to note that the Content-length header is not mandatory for all HTTP1.1 responses. This header typically accompanies static data transfers like images or videos whose sizes are known by the server.
In instances when the server cannot determine content length, such as with script outputs, the response is streamed from the script's output to the browser via TCP/IP without prior knowledge of the response size at the time the header is sent.
Multipart response
The server has the capability to deliver multipart responses over the same socket using a technique similar to multipart requests (commonly used for file uploads). By designating a specific string in the Content-Type header, distinct documents within the TCP/IP stream are separated.
However, there is no established protocol for converting multiple HTTP requests into a multipart response, unlike multipart requests which transmit form data as individual documents.
In what scenario would a multipart response be advantageous you may ask? Multipart responses are utilized for streaming videos, such as enabling mjpeg streams. Each "video" essentially consists of numerous JPEG files delivered in sequence rather than a single JPEG file.
Pipelining
HTTP1.1 supports the transmission of multiple requests over a single socket through a feature known as pipelining. Yet, due to past issues with early web server implementations, most browsers tend to steer clear of utilizing HTTP1.1 pipelining. In certain versions of Firefox, activation of this feature may necessitate manual intervention via the about:config page. Many contemporary web frameworks do not support pipelining.
Distinguishing between which response corresponds to which request in HTTP1.1 pipelining is simplistic—responses align with their respective requests based on order.
The first response consistently pertains to the first request, and likewise for subsequent pairs. A deviation from this sequence indicates an error, prompting cautiousness from modern web browsers regarding pipelining usage.
HTTP/2
Reinventing the communication protocol of HTTP, HTTP/2 introduces a multiplexed stream layer beneath the conventional HTTP framework. This binary-based stream protocol assigns each stream a unique identifier akin to sockets. Consider it a proprietary "socket" protocol transmitted across standard sockets, where requests and responses are identified by their designated stream id.