Hi Winston,

     About "Kerm_RLF.INI File" of May 7 (addendum):

WS> A load from a RAMDisk into memory should be silent...
MS> A ~RAMDisk~ can make the difference...  ...may save the day...  

     I forgot, there's at least one more good reason to load a ~RAMDisk~
than just speed or quietness:  it allows for *SAFER* testing as well!  I
tried mr. Santos's 3rd-party archaic pre-1985 implementation of the real
`Kermit' file transfer protocol (once again) and the 669 Kb .ZIP archive
against which is tested it never survived the trip - not even the system
or connection until it came back to me that the `WC!' BBSes simply never
supported "Block-Check-Type 3"!  What i called `WC-Kermit' and which was
bundled in the `WildCat!' package when Santronics acquired it is USELESS
FOR ALL PRACTICAL PURPOSES;  the least a reasonable Sysop should do with
it is to disable it, unless, for some obscure reason, one must create or
sustain presistant preconceptions in order to make `ZMoDem' sound better
than a twenty-two years old (mature) piece of work like `MS-Kermit'.  No
`WC-Kermit'/`MS-Kermit' tests made the PC reboot or hang with a critical
"Exception 6" error message after i switched to
"Block-Check-Type 2" but
i felt that having `ScanDisk' run from `AutoExec' was'nt excessive after
that!  P->  Even after i had "Block-Check-Type" set right i noticed that
my "Percent transferred" line was gone and the "%"
progress-bar with it;
apparently, `WC-Kermit' is even more BROKEN than i thought:  no transfer
seemed to be complete past the total file-size - IT KEPT GOING, forever!

                                  %-b,

     It's not impossible one will be luckier with small files so i guess
you can try "Set Block-Check-Type 2"...  Honestly, i didn't have time to
waste myself so i stopped trying from there because i doubt it would do.

                                   :(

WS> Am I forgetting anything?  
MS> ...he removed the `MS-Kermit' option before i had a chance to make
MS> him prove its real value.  Too bad!  Sysops are like that...

     In every single test i made "File type" displayed as
"BINARY" but i
can testify that the file i received had nothing to do with the original
binary past its seventh character...  Translation in "Binary" mode?  Oh,
well, i give up!  If Craig Healy wants his BBSers to think that `Kermit'
and that thing over there are equivalent, euh...  any help is pointless!

                                   :(

WS> ...96-byte packets to match the "ancient" 128-byte modem buffer size
WS> assumed in the 1985 WildCat! version of the Kermit protocol?

     By this "pre-1985" token of mine i mean `MS-Kermit' as it was prior
to the 1985 release where its ninety-six bytes limit was raised...  It's
not so clear why `WC!' makes that eighty bytes instead but i wouldn't be
surprized to learn that it came from a much earlier version, actually!!!

                                  8-o

     I didn't know about a 128 bytes buffer in `WC!' but if it came from
the `Kermit' people i suppose it should be possible to find traces of it
in the source-code or perhaps the author, Joe R. Doupnik, would be happy
to comment that.  I'll let others check this:  mr. Doupnik wasn't really
having a pleasant time as i joined the NewsGroup and my name alone could
make him worry - like that of a former "friend" (C.A.), i presume!  %-b,

                                         Salutations,  

                                         Michel Samson
                                         a/s Bicephale
                                         http://www.iquebec.com/dos-inet
                                         (UserName="U" and
PassWord="P")


... Make BBSing UNIVERSAL, support DOS+TCP/IP+TelNet+ZMoDem/Kermit+.QWK!
--- MultiMail/XT v0.42 - Numbers might make BBSing UNIVERSAL, not sugar.