Julia's Downloads standard library uses libcurl. Supply the proxy environment before starting the process that performs the download:

export http_proxy='http://proxy.example:8080'
export https_proxy='http://proxy.example:8080'
export no_proxy='localhost,127.0.0.1'
julia fetch.jl

An example fetch.jl can use the ordinary Downloads API:

using Downloads
Downloads.download(
    "https://downloads.example/data.csv",
    "data.csv";
    timeout = 30,
)

The https_proxy variable selects the route for an HTTPS destination; the http:// value specifies an ordinary HTTP proxy. For downloads permitted to use public endpoints, the HTTP proxy list provides candidates for that value. Do not change that value to https:// unless the proxy listener itself supports TLS.

libcurl accepts http_proxy only in lowercase. Other protocol proxy variables also have uppercase forms. Its bypass convention is no_proxy or NO_PROXY, with comma-separated entries; http_no_proxy is not the documented general bypass variable.

A Julia process launched by an IDE or background service inherits that launcher's environment. A variable exported later in another terminal does not reconfigure it.

For diagnosis, Downloads.download accepts verbose=true, and its debug callback can receive detailed transport events. Review that output locally and remove secrets before sharing it.

A package using another HTTP implementation needs its own configuration review. A successful Downloads call establishes the behavior of that request, not every network operation written in Julia.

ХӀокхунах тера йаззамаш