Text 240, 107 rader
Skriven 2006-04-18 18:12:34 av mark lewis (1:3634/12.0)
    Kommentar till text 182 av Peter Knapper (3:772/1.0)
Ärende: none
============
 PK> The issue ONLY appears for INCORRECTLY CONFIGURED mailers, and it
 PK> is only the NEW SYSOP within Fidonet Aus that is likely to miss the
 PK> point and can be stung by this issue, and New Sysops are exactly
 PK> the ones that NEED protecting, not the experienced Sysop.

 ml> errrmmm... this same problem exists for every entry in 
 ml> the nodelist, 

 PK> No it does not, and don't try and sweep it under the rug that way.

i'm not trying to trivialize this particular situation... please don't take me
that way...

 PK> If a new sysop in North America tries to dial a Node in India and
 PK> ends up dialing the PSTN no. 911 repeatedly, then I am pretty sure
 PK> he may be in a spot of bother over this error, so YES a different
 PK> version of the SAME issue can happen there was well. 

exactly... and the india situation was the FIRST one back in the day...

 PK> The DIFFERENCE between these 2 scenarios that I am trying to point
 PK> out is that the North American issue is as a result of PUBLISHED
 PK> and KNOWN standards, whereas the Aussi issue only happens because
 PK> of the stuffing of an INVALID country dial-code into the Nodelist. 

remember, the nodelist is not governed by telcos and their numbering plans...
the nodelist is governed by fidonet and fidonet standards...

 ml> then... not the specific problem of dialing emergency services but 
 ml> there is still the same basic problem if one does not have their 
 ml> dialing tables set up properly...

 PK> So you agree that we should do all things possible to minimise the
 PK> impact of incorrectly configred nodes? 

i agree that all nodes need to be properly taught and trained in the
configuration of their mailer software (at least), yes... i do _not_ believe,
however, that we should remove /possible/ obstacles... no one removed the 91-1
stuff for any Z1 nodes... no one did anything about the first ISDN ONLY nodes
that were listed that a POTS system could not connect with... however, once
their specifics were known, then folk were taught about them and life went
on... before then, some folk did get bit and a few rather hard but no one
helped them repair that wound they suffered...

 PK> Great, get rid of 000 in the Nodelist. 

sorry, no... that is the =one= main thing that my mailer's POTS side uses to
determine if it should hold mail for a system or not... my mailer's IP side
uses 000- and the ITN and IVM flags to determine if the IP side can connect to
a remote system or not...

 ml> so.... what it _really_ comes down to is that it is /not/ a 
 ml> *technical* problem but a social problem where one hasn't been 
 ml> taught or learned to completely configure their software before 
 ml> using it as well as a second social problem of letting the machine 
 ml> have its head without sufficient monitoring...

 PK> Absolutely, that has never been denied. 

some would have others believe differently... some are sticklers for technical
situations and are backing this to the hilt for various reasons... none of
those reasons are technical, though... but they'd like "you" to believe it...

 PK> So rather than force the problem back on the new Sysop, 

no one is forcing the problem on anyone... as long as they enter the arena with
full knowledge, there is nothing being forced... it is no different than what
you and i went thru when we joined the network, is it?

 PK> Fidonet should do all it can to minimise the potential for 
 PK> error, the best way to do that is to remove all INVALID 
 PK> ISTD codes in the Nodelist. 

hahaha... if that "minimizing potential for error" is the main basis, then
practically /ALL/ software in fidonet should be banned for cause... i'm not
even going to try to start naming any particular packages but everyone knows
about the message base regurgitations that occured regularly when new versions
of certain packages were released... everyone already knows about the problems
with messages larger than 8k... no, umm... 16k... no... hummm... 32k... wait a
minnit! what is the largest message allowed by the standards?? [/me scratchs
his head] should i even broach the incessant netmail routing problems that have
plagued this network for almost as long as it has existed??

 PK> I am glad you recognise that this non-technical "problem" can be
 PK> minimised by some logcal thinking.

i'm glad you put the word "problem" in its respective place... between quotes
to show that it is not truely a problem ;)

now, tell me something... my mailer doesn't have to use 000- as a country code
indicator for IP nodes... it is a fully configurable option... so that i and
others who run mailers of similar setup to mine, do not have to kludge things
together, what would you propose we, fidonet, use in the phone number field,
since that is the *ONLY* place that my system can get the POTS number, the IP
number or the IP domain name from, to indicate that the system is POTS or IP
based???

and before ANYone spouts up with "use a shim to translate your nodelist," i
want to fully remind them that fidonet was not designed for any system to have
to use a shim or anything else to edit their local nodelist so that they could
contact any node in the network that they have a compatible method for... that
is pure fecal material and wasn't ever suggested back when ISDN nodes started
appearing (thanks to the then Z2C) so why should it start now??

)\/(ark

 * Origin:  (1:3634/12)