PE> When you add to that the fact that pqwk is far more complicated
PE> than pktjoin, written by multiple authors, basically a heap of
PE> shit, had quite a lot of bug fixes go through already, whilst
PE> the pktjoin code is about 3 pages long,

RS> Sure, no argument there. Main problem tho is that everyone else
RS> is using PKT2QWK, have been doing so for the best part of a year
RS> now. Frank is the only one who is seeing the dud QWKs at the moment.
RS> So its less likely that it is PKT2QWK than his BAT files, which he
RS> has changed, or his machine config.

PE> HEY ROD!!!  Everyone else is using PKTJOIN too!!!!!!!!!!  Without
PE> problems!!!!!!!!!!!!!!!  Please reread this para!!!!!!!!!!!!!

Looks like time for a cold shower boy. You appear to have had a rash of
exclamation marks. The difference is that the mix of PKTs a particular
person processes in a particular PKTJOIN run will vary with the detail
of how he does his mail runs. PKT2QWK sees just a single combined PKT
every time anyone uses it. So its quite feasible for there to be a
difference there with PKTJOIN form user to user which can conceivably
make a bug visible.

PE> you would have to be on drugs to be saying the pktjoin is more
PE> likely to be problem than pqwk.

RS> But you keep studiously ignoring the situation where processing
RS> more than one archived file of PKTs in a single run produces a
RS> fucked QWK and processing them individually does not.

PE> Sounds like bullshit to me,

Who cares how it sounds to you ?  You werent the one who saw the bug.
And authors of software often fall back on this sort of ploy too when
their code doesnt perform perfectly 'I saw no problem, you must be
imagining it'. Funny, could have sworn thats precisely what Kimes said
to you too, and it got comprehensively up your nose when he said that
to you too. Its a natural reaction to deny its possible. Doesnt mean
that just because the author denys that it says much about bugs tho.

PE> but even if it was true, it's still just as likely to be a pqwk problem.

Nope. Only PKTJOIN sees something different when a single archive
of PKTs is processed instead of a number of them in one PKTJOIN run.
Yes, its certainly possible that that bug is in PKT2QWK, but its also
quite possible, and IMO more likely, to be in PKTJOIN. Who gives a
stuff anyway ? Its pointless to speculate with no new information,
what matters is where the bug actually is found to be.

RS> Particularly when the resulting QWK with the multiple archive files is
RS> nothing unusual size wise. THATs the symptoms which make it more likely
RS> its PKTJOIN, no matter how unlikely that looks from the code of PKTJOIN.

RS> AND I just dont buy this 'look at the code, its so simple' line, it
RS> wouldnt be the first time that some compiler has managed to produce
RS> an EXE from very simple code which has some subtle problem which
RS> causes it to go bang on some particular detail of its input files.

PE> We're getting into the realm of the incredibly unlikely here,

Nope, compilers do produce dud code from source code which looks and
is fine. You know that as well as I do. In fact the Borland compilers
are a bit more prone to it than some at times too.

PE> when we've got code that looks like something the cat dragged in.

I think its quite silly to be ignoring that vital bit of evidence about
the processing of the archive files of PKTs individually. You can do that
if you want, I wont be ignoring it tho.

PE> Do you know how to do axiom semantics?  I don't, but with the
PE> amount of time spent arguing about a trivial program like pktjoin,
PE> it could probably have been proven to be correct!

You may have noticed that its just a tad hard to prove a damned
thing about where the bug is without some situation which makes
the bug visible being available for careful debugging.

And you can do whatever you like analysing the source code, that
doesnt say a damned thing about whats actually run does it ?

RS> Corse in the strictest sense you can argue that it is always a
RS> PKT2QWK bug if its too stupid to notice a fucked input PKT and
RS> just blindly produces a fucked QWK from it too.

PE> It does do some validation BTW.

Sure, but clearly not enough if a dud NDX does ever get produced.

RS> Just what you call the reliance on no existing NDXs when PKT2QWK
RS> starts is also arguable. If you want to be totally hair splitting
RS> even that can be called a bug sort of.

PE> Mark made a conscious effort to append to the file rather than
PE> overwrite it.  The mind boggles as to the reason for this.

Yeah, I have actually had quite a bit to do with him by email and it
wouldnt surprise me in the slightest if there is no good reason for it
all, a classic case of a severe problem with dog shit between the ears.

--- PQWK202