Subj : %HELP ?
To   : Nick Andre
From : mark lewis
Date : Wed Jan 18 2017 08:20 pm


 On 2017 Jan 18 12:45:22, you wrote to Wilfred Van Velzen:

 WV>> Isn't that a very dangerous option? It's only save for nodes with no
 WV>> other links, otherwise a massive dupe dump would be generated.

 NA> Wrong, and not any more dangerous than any other tosser that has Rescan.

agreed...

 NA> A tosser has an area called ABC. There are three downlinks. 229/100,
 NA> 229/200 and 229/300. A new system, 229/400 joins and Areafixes ABC
 NA> with a rescan. They get all the messages but 100, 200 and 300 do not
 NA> because they exist in the Seen-by's in all the messages that exist
 NA> already, and thus 229/400 is added into the Seen-by's/Path going
 NA> forward.

so what happens when 229/400 joins and is also connected to /100, /200 and /300 
when they do the rescan? i thinkg that's the point that wilfred was trying to 
make but he was making it from the POV of a fidoweb system that is connected to 
more than one system for the same area without the entire distribution being 
fully connected to all other systems in the fidoweb...

 NA> The method used is no different than Gecho, TBBS/Flame and a lot of
 NA> tossers out there... pretty sure even Fastecho does it this way but
 NA> perhaps adds some extra kludging to indicate that the message was in
 NA> fact, rescanned.

yup...

 NA> You know very well that there is no fool-proof way for a tosser to
 NA> offer a Rescan feature, because some Sysop, somewhere, somehow, will
 NA> find a way to break it or configure things in a way not intended, and
 NA> we get "dupe-dumps".

exactly... yet we have others working on ""hidden"" agendas trying to change 
the entire methodology of FTN mail distribution to fit their non-standard way 
of interconnections...

)\/(ark

Always Mount a Scratch Monkey
Do you manage your own servers? If you are not running an IDS/IPS yer doin' it 
wrong...
.... Any fool can paint a picture; it takes a wise person to sell it.
---
 * Origin:  (1:3634/12.73)

.