Subj : Re: MetroNet
To   : Atreyu
From : deon
Date : Fri May 27 2022 11:47:00

  Re: Re: MetroNet
  By: Atreyu to Deon on Wed May 25 2022 09:49 pm

 > D> Well, as I say, I dont see the challenge you are describing and I have a
 > D> bit more than a dozen. And yes, I could see a single hub sweating if
 > D> it had hundreds of downlinks, especially if using 1990s software.
 >
 > ... but then you cop out with little digs towards software
 > or things here that are "standard". Rob Swindell does the same thing...
 >
 > If you have something to contribute beyond  commentary then please politely
 > show me where I can test it myself.
 >
 > D> Well it doesnt need to be. It's how this ancient hobby used to work and
 > D> most folks are using ancient software - or using modern versions of
 > D> ancient software built with ancient thinking...
 >
 > And again, I can't understand the real meaning of this without you clearly
 > bringing something to the table for comparison. We already have Jas Hud
 > for bravado nonsense.ll bring it back on track.

So Nick, I can see this conversation going in a direction that I'm not interested in - so I'll bring it back to my point.

A comment was made that for a hub to Crash was ridiculous. I didn't agree with that comment, because I actually find crash useful. It enables me to to offload stuff for downlinks fairly quickly keeping my outbound slim, as well has enhancing the response time between messages - which is what I like.

I actually think it is more efficient (when configured appropriately) - in that I think it's pointless polling a system for mail it doesn't have for me, especially when it can deliver it to me when it has some.

In my experience, whether you hub for a dozen downlinks or 100+, I've not experienced any difficulties or problems by having crash as my default configuration, nor have I noticed the impact to my system because it's stuck polling people that have gone awol. I haven't had anybody complain to me either, that getting mail is problematic as a result of this setup.

One of the attributing reasons may be because I do use automation to help - and yes if a node goes awol for more than 30 days, it automatically stop exporting mail to them, and ultimately deletes what's in the outbound for them - especially if I cannot reach them, or do not hear from them. I don't see the point of continuous exporting mail for a node that doesn't collect it.

If/when they come back, a quick rescan gets them everything they missed and the are back on track with (hopefully) a minimum of disruption.

I think this works well on todays modern environment (ie: because of TCP/IP as the transport), because, on the old 1990s technology which was hardware dependant (serial ports) and only one conversation could be had a time per device, and together with the connect/tear down latency (and slower transfer speed) - I couldn't see how a 1990's mailer could cope with "100's" of downlinks using crash - even with a handful of modems hanging of a digiboard, and so in that case, a "call me" approach may be the right process for that setup (which, as I recall, was the common setup in the 1990s).

Which reminds me, I don't recall what happened to my digiboard... hmm


....лоеп
--- SBBSecho 3.15-Linux
 * Origin: I'm playing with ANSI+videotex - wanna play too? (25:25/23)

.