Subj : Re: MMTerm?
To   : Jim Hanoian
From : William McBrine
Date : Mon Apr 16 2001 12:55 am

-=> Jim Hanoian wrote to Ben Carpenter <=-

 JH> In the beginning, Sparky invented QWK to be able to transfer
 JH> messages in bulk from one PCBoard BBS to another one.  A
 JH> network developed, and QWK was the chosen standard.  Blue
 JH> Wave and Silver Xpress were developed to fix issues that BBSs
 JH> in Fido experienced with trying to network using QWK.

Erm... Fido is a networking technology unto itself; you wouldn't be
_networking_ using QWK _in_ Fido -- QWK would only be an offline reader
format there. And neither Blue Wave nor OPX has ever been anything but an
offline reader format, AFAIK.

Otherwise, I agree: BW and OPX were designed for Fido, as QWK had not been.

Re: QWK being a networking technology before it was used in readers, I've
often heard this repeated (mainly by you IIRC), but not authoritatively. Do
you have a reference for that? I don't think it's true, because the
existing QWK networks use kludged extensions to the format, and because
this snippet from the 1st Reader docs suggests a reader origin:

  Before we go into the details of the 1stReader API it might be
  of some benefit to review what makes up the QWK format.  The
  QWK format was created in 1987 by Mark "Sparky" Herring for a
  friend who moved to a small town in Texas.  The idea was to
  gather up all of the new messages posted on a bulletin board
  system and "quickly" transmit them to the user.  Then the user
  could disconnect from the BBS and read the messages at their
  leisure using a QWK mail reader.  If the user entered any new
  messages they could call the BBS back and upload their replies
  to the system.

 JH> inherit many limitations.  Quite a few QWK implementations still
 JH> require header information in ALL CAPS.  That riles a lot of
 JH> folks, including those whose names get mangled (like McB...)

I thought it was just me. :-) Anyway, this isn't inherent in QWK; it's just
a question of stupid implementors. But back in the early days of QWK, a lot
of BBSes were uppercase-only.

 JH> Some QWK implementations require messages to be less than 100
 JH> lines long.

Again -- stupid, but not actually part of the format. In such cases, we can
only blame the authors of the specific implementations. The QWK format per
se has no such limit. You're actually agreeing with Ben here. :-)

 JH> Further, QWK is limited in all header fields to 25 characters,

This, by contrast, is a real limit of the format, and a serious flaw.

 JH> reader demanded it.  OTOH, George added QWK to the Blue Wave
 JH> reader and Hector added QWK to the Silver Xpress reader only
 JH> because people whined and cried long enough so they could use
 JH> their favorite reader on more BBSs.  Neither wanted to do it,
 JH> but both thought they had to in order to keep their product
 JH> popular in the shareware arena.

Ah, but the reason neither wanted to do it was because each preferred to
push his own then-proprietary standard over the relatively open standard of
QWK. That desire wasn't really based on technical merit, although there
were valid technical arguments for it. Rather, each was more or less trying 
to lock up the market.

 JH> Yes, Wildcat has never supported Blue Wave.  Primary on a WildCat
 JH> BBS is Silver Xpress (Hector owns SX and WildCat, so is it any
 JH> big surpise?) and secondary is QWK.

I don't think we can really say that SX is primary on Wildcat, when the OPX
door for WINS is incomplete and defective. (I dunno about older versions of
Wildcat.) QWK, on the other hand, had an important place in Wildcat's
history -- the QWK interface, originally a standalone third-party door
known as TomCat, was one of the best ever. And QWK networking made great
inroads on Wildcat long before Fido did.

 JH> Blue Wave is much more consistent because most BW products were 
 JH> developed by George initially.

Also, once he finally opened the spec, he wrote very thorough documentation
for it. And because it already includes the sorts of features that the
various QWK extensions try to kludge in, there's no need to rewrite them in
incompatible ways.

 JH> SX products are most consistent because the doors and the readers are 
 JH> all owned/produced by Hector.

Oddly enough, despite what should be his unifying control, I don't find
them consistent at all. This could be due to his multiple personality
disorder.

 JH> The QWK standard was not a standard until Sparky released 1stReader
 JH> version 2.0 on July 20th, 1995.  That's something like SEVEN YEARS
 JH> from when QWK started.  In the interim, everybody else relied on
 JH> reverse engineered information, some better than others, and some
 JH> programmers implementing the standard better than others.

The reverse-engineered information is better than the "official" spec
anyway. Sad but true.

.... I used to be a cynic, but I don't believe in that crap any more.
--- MultiMail/Linux v0.40
 * Origin: COMM Port OS/2 juge.com 204.89.247.1 (281) 980-9671 (1:106/2000)

.