Discussion Topic: Ansible Models (non 3I) Cian Witherspoon (03 Oct 2019 01:45 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Thomas RUX (03 Oct 2019 03:36 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Cian Witherspoon (03 Oct 2019 03:41 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Thomas RUX (03 Oct 2019 11:38 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) shadow@xxxxxx (04 Oct 2019 14:15 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Rupert Boleyn (04 Oct 2019 15:23 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) shadow@xxxxxx (04 Oct 2019 21:20 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Cian Witherspoon (04 Oct 2019 23:53 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Rupert Boleyn (05 Oct 2019 03:56 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Jeff Zeitlin (05 Oct 2019 04:06 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Cian Witherspoon (05 Oct 2019 15:42 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) shadow@xxxxxx (05 Oct 2019 20:28 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Kelly St. Clair (05 Oct 2019 20:57 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) shadow@xxxxxx (06 Oct 2019 05:13 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Jeff Zeitlin (06 Oct 2019 19:10 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) shadow@xxxxxx (04 Oct 2019 14:15 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Thomas RUX (04 Oct 2019 20:49 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Catherine Berry (04 Oct 2019 21:04 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Jeff Zeitlin (03 Oct 2019 23:33 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Thomas RUX (04 Oct 2019 02:07 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Cian Witherspoon (04 Oct 2019 03:50 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) shadow@xxxxxx (04 Oct 2019 14:15 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Jeff Zeitlin (05 Oct 2019 00:23 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Jeff Zeitlin (05 Oct 2019 00:13 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Jeffrey Schwartz (04 Oct 2019 13:20 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) shadow@xxxxxx (04 Oct 2019 14:15 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) shadow@xxxxxx (04 Oct 2019 14:15 UTC)
Re: [TML] Discussion Topic: Ansible Models (non 3I) Cian Witherspoon (04 Oct 2019 17:55 UTC)

Re: [TML] Discussion Topic: Ansible Models (non 3I) shadow@xxxxxx 04 Oct 2019 14:14 UTC

On 3 Oct 2019 at 20:50, Cian Witherspoon wrote:

> So it's actually going against those assumptions that I'm looking
> for in this discussion. What if it isn't any of those? What if it's
> more like the telegraph?

well, the telegraph led to the fax (yes, really) and later they came
up with the teletype.

Everything since is pretty much elaboration.

> What happens when you start putting interesting limitations on the
> tech and its use? Everyone, even many modern novelists, tend to just
> keep it as a generic internet, or phone system (which allows you to
> run a basic internet), just faster than the fastest ships available.

Well, even then there are limitations. Even if range is unlimited,
you *can't* connect to everybody. So mail will get forwarded. and
quite likely via something like the early internet/usenet "bang
paths" (basically the "address" line is a list of nodes to route
thru) This isn't as nice as domain based addressing, but it allowed
you to *deliberately* route around nodes that had problems (or that
you had reasons to not want your message to pass thru)

"News" (ie messages aimed at anybody who is interested) works
differently. You'll want to organize it by topic/subtopic etc (like
usenet news, and fidonet groups).

In the OTU, neither is a big problem. In variant TUs with lots of
alien starfaring civilizations, it gets a lot weireder until you have
Vinge's "net of a million lies" from "A Fire Upon the Deep". among
other lines you *need* stuff like his "Language-path" header as
messages get translated and re-translated along the way.

> What if it's the high tech equivalent of messenger pigeons? (Charles
> Stross had this, with entangled pairs of quantum bits that got used
> up and had to be shipped by STL ships).

https://en.wikipedia.org/wiki/IP_over_Avian_Carriers

RFC 1149
https://tools.ietf.org/html/rfc1149

Also RFC 2549
https://tools.ietf.org/html/rfc2549

> One of the more interesting models I saw was EVE Online - lorewise,
> the ansible was made of a paired array of entangled hydrogen atoms
> that transmitted electrical signals between them. This forces a model
> based around a client array onboard, with the service array being
> connected in a server rack to a switchboard. Network topography
> becomes an important part of warfare. The limitations placed on the
> model drives stories based around dealing with them.

Sounds like the linked "wormholes" Laurence Dahners uses in his Ell
Donsai stories. They use the paired "chips" *mostly* to connect
devices to switching units (much like a telephone exchange) which
then connect it to whatever other device you want to contact.

Yeah, that does have the problem of somebody blowing up an
exchange/server and rendered all the devices that connect to it
useless.

But smart folks will have connections tpo multiple servers *or* a few
*direct* connections (think leased line phone service) that don't
*use* the servers.

Also, since both the "wormholes" and quantum entanglement are *by
definition* undtectable/untraceable you only have to worry about
destruction of the "server" if the links are server to server, rather
than device to server.

For small "link" equipment, you'd go with device to server and only
the folks providing the service know where the servers are.

For "large" link equpment, you'll use normal methods (radio, cables,
and the like) to connect to the servers and only use the ansible
links to connect *between* servers.

That *does* make the servers a point of failure. Which is why you
take a tip from the way the Injternet is set up. You have all the
servers connect to multiple servers. So taking one out means you've
only isolated the folks connected to it.

Everybody else just continues on because they can route thru a
different server.

That sort of thing was one of the design criteria for the ARPAnet and
for the Internet that grew from it.

--
Leonard Erickson (aka shadow)
shadow at shadowgard dot com