Loading packages

Humongous, multi legged software packages, egad!

No problem, repository and local processor agree to use some reserved args. Curl can be made to work with consent, this is entirely reasonable. Once you are linked in the consent network, a write in one mysteriously appears in the other, with full acknowledgment.

There is tons of address space, network it, use it as a session layer protocol. And it makes keys very valuable because they can produce a 64 bit pointer back into args, and they are short, simple and plain text identifiers.

So, yes, I convince myself. I will hunt down or write consent algorithms that can work with curl and manage names segments. Or some kid can beat me to the punch.

The consent/curl interface

I reckon that consent snippet will start a node list for the network, having ultimately to supply the URL to curl.  Everyone already agrees on the names now, and curl keeps url and ip pairs, identifiable by name.  Each url has a consent snippet at that node in the network.  Curl manages addresses, I presume. (I have never used curl, just noted its use).

So, snippets accept list from any node in the consent network, and it will add or subtract and pass the message along, that is the basic messaging protocol.  Of a particular element of the lists are the acknowledgment messages, these can increments an acknowledgment count, and all nodes will be in agreement soon, the count reaches max; or timeout. So consent is built into the messaging protocol. Curl can name the nodes, snippet can name the segments, and data manager can pre-program this operation as needed, by providing consistent name space partition.   Applications, or syntax engines, can import the name space, it is a system wide function, importing.