> i dunno... i realized some time back that this echo was being gated
> and that's why they were coming back... i get the same thing with 
> messages that get posted to newsgroups... this is how folks can make 
> sure their post actually made it thru the gate...

 N> I realize now that I should have done proper zone gating between
 N> fidonet  and scinet, problem is that I came up with that theory
 N> after creating  different echo's for each network.

i can understand that... and actually, different message areas for each
network, even if those message bases carry the same message is proper... the
trick, as it were, is to blackhole gate them between the two network...
blackhole was actually the name of the FTN based software to handle this ;) if
the area is gated to more than one FTN domain, then there's more than one toss
needed... for example, the first to toss to fidonet and the point or node of
the gate... then blackhole to run the traffic thru the gate and then the other
side's tosser to toss to that side's message bases... if there's another
domain in the mix, then there's at least one or two more tosser runs... but
this depends on where the gate is to be done... either way, blackhole was
written to toss as a gate between FTN networks and then the other side has to
run their tosser to place the messages into their bases... i hope this makes
sense... it has been a long time since i've done this and i know that
blackhole is way ancient and possibly not available any more... i /think/ i
still have a copy in my files areas but this depends on if i shut off allfix's
management and purging of "old" files before it threw it away like it did so
many others... the OS/2 fernwood area being one of those which i rebuilt from
what i could find out there on the interwwebs thingee...

> but there is a big difference between the gates and their methods,
> though... if i try to respond via netmail to a message that has come 
> thru your gate, it gets addressed to the user at your system... even 
> if the user doesn't have an account on your system... and so the 
> private netmail reply doesn't make it back thru the gate and on to 
> the destination...

 N> That's because this is gated via NNTP.  

to a point, i can understand that but there is a way to handle this... GIGO
does it for my system all the time... i believe that FSC-0035 is the proposal
used... it adds a REPLYADDR to the control lines and maybe one other control
line... these tell the message reader/editor where to address the message to
and who to address it to... TimED, for one, reads these and silently addresses
the messages to UUCP at the gate's address... this for private netmails, of
course... without this, the messages won't get thru the gate and they may not
even make it back to the gating system depending on if you are getting the
area from the gating system or downstream a hop or more in an echomail
stream...

 N> I do see the proper formatting  for fidonet's fqdn but i believe 
 N> that internet to fidonet gates died a  very long time agi.

yes it did... in today's world, and actually way back when, too, the gating
system was the one responsible UNLESS they fit into the requirements for the
original fidonet.org domain... one of those being to have a MX of a specific
format... but when the original gateway, run by burt juda, shut down, then it
was every gateway for itself... TBH, every node in fidonet has a specific
address thru my gateway... but none of them are in the fidonet.org or any
similar domain... technically speaking, every fidonet system (that i allow to
use my gateway) is a subdomain under my own domain and those addresses follow
a specific format (the f.n.z.domain) format where "domain" is what i have in
place for them...

> in other words, if i try to netmail myself from one of my returned messages
> thru your gate, my netmail would go to your machine, get gated to
> that other network and then regated back to fidonet addressed to me 
> at my fidonet address... my gating software is GIGO and GIGO calls 
> this fastreturn and forces it for fidonet traffic... GIGO forces it 
> to eliminates the double gating unless you are in more than one 
> domain like rbbsnet and you need to send mail from rbbsnet.net to 
> fidonet.org or whatever... i've included at the bottom of this 
> message the comments from the GIGO configs concerning this... i know
> that many things have changed since then... especially the fidonet 
> domain as well as loosing other domains...
>
> from what others and myself have been seeing over the years and as more and
> more synchronet boards come into fidonet and start doing more and 
> more gating this is a problem in synchronet but i don't know why or 
> how it can or should be handled... sadly, i don't remember how stuff 
> should be addressed to pass thru a FTN node to a QWK fed node and 
> that adds yet another layer of complication to the whole shooting 
> match :?

 N> Is GIGO open source?  

it is now but i don't know what has been done to it in recent years... just
like TimED, i still run the original last release from the author along with
my paid for registration keys ;)

 N> Is it available for Linux?  

no clue... bu i have, i think still, the original sources as released by the
original author, jason fesler (who now works as one of the, if not the, top
network engineers for Yahoo!)... but i know that it was released and someone
else made some modifications to fix some things and alter others... but what,
specifically, i cannot say... where is it hosted? i really don't recall at
this time :(

 N> if so i'd like to give it a whirl.

it is very old school... for instance, my binaries are dated 1996 for several
support programs but the actual GIGO binaries for DOS and OS/2 have dates of
14 Jan 2000 on them... plus i know that i've written my own replacements for
some of the tools...

 N> Funny thing about the dupes when I gated this echo a while back,
 N> they  were not being created by Synchronet as all synchronet did was
 N> gate this  echo to the bbs-scene.org NNTP server but the dupes were
 N> being created  by IREX which is what I used to gate it between
 N> scinet and fidonet. 

right and that is understandable due to the differences in the MSGIDs going
thru and coming back... after i had thought about it for a bit, it made sense
considering the gating that i do here... on the one hand, that is... in a
perfect world, there wouldn't be any dupes... not even if gated out to xyz on
your system and gated back in from xyz to the same area by any other system...
there was some work done in this area of gating tech way back when but i don't
know anything more than that... too much time has passed and the old 72pin
5volt biomechanical memory is getting rather fragile these days as the peek of
the hill grows closer -=B-/

)\/(ark

 * Origin:  (1:3634/12)