Hi Jakob, Jakob Eriksson <ja...@aurorasystems.eu> writes: > On September 16, 2013 at 9:30 AM Tomas Hlavaty <t...@logand.com> wrote: >> >> 3) I guess most of the overhead of the http request is probably >> establishing the connection. My bett is that it doesn't really >> matter if you send 1kB or 5kB of data. For example, if I run this > > It might matter, the only way to make sure is to measure. For instance > you can fit an answer inside an IP packet completely, it may make a > difference. (There are also many other things to consider, for instance > if there are CSS or Javascript resources which are blocking start of > browser rendering, until those are loaded.)
yeah, but this is getting too theoretical. The IPv4 packet length field is 16 bits, for a maximum size of 65535 bytes. IPv6 - The extended length option provides for a 32 bit length field, supporting packet length's up to 4294967295 bytes. but those are further fragmented maximum length Ethernet II frames (1518 bytes, with a payload of 1500 bytes) Alex' example might fit in one Ethernet frame. My too, if you remove the additional feature (clickable graph and popup bubbles) and gzip the page source. But then there is sync, and all the communication slows down on ack... Anyway, it's fast enough, and the and it doesn't make any difference if it loads in 3ms (locally) or 60ms (from logand.com). Btw, both times are somewhere around the frame rate of TV (40ms) and human eye needs about 100ms to not notice :-D The example doesn't contain any CSS or Javascript, intentionally. Cheers, Tomas -- UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
