Hi! Richard,

On 25 Nov 07 15:19, you wrote to me:

 RW> Appreciate all your work on this.  I"m surprised nobody's done it
 RW> before.

Trust me, it don't look anything fancy at the moment.  When it runs there's
all manner of output and it'll be worse for you if you get your hands on
the IfNumber tool, since the as-is executable spews all manner of stuff for
each call (IIRC); I hacked mine as soon as I got pissed off enough with it.

 RW> '97 when I took my fido board down.  WAs just getting ready to get
 RW> started with a fido board again before the big storm hit.

Groovy.  Another survivor, hey.  I didn't have dramas like yours, luckily. 
(We're apparently due for the 'cyclone of the century' any year, so the
locals say.)  I pulled my first BBS down mid-'98 but the PC had a major
meltdown later that October I think, just as my wife sat down to vege-out
in front of the TV, and, she saw the whole thing. The machine had been
running 24/7 for over 4 years, not only running FrontDoor and RemoteAccess
but also it was our household PC.  So, my node was 'down' for a couple of
months till we/I replaced it with one sporting a Binkley/w32 frontend.  (I
lost pretty much everything of the old configs on the HD, save for what was
on floppies.)

 PQ>> You'll also need a copy of "Replace.Com" by Gene
Garapic, back in
 PQ>> '88.  I still have a copy of the cscripted post by Bat Lang from
 PQ>> back in '97 which I use, so, I could simply post it here again.
 RW> THat might be an idea if it isn't a file available on bfds.

Actually I might speak with Ben Richley about that.  It seems to be a
fairly unique type considering that there are a few others.

 RW> I'd say do the replace functions once and use that cleaned up nodelist
 RW> for the searches.  IN fact, that could be part of one's weekly
 RW> nodediff processing automation imho. THe actual lookup function would
 RW> probably run smoother quicker and cleaner that way.

Yep, I thought so too.  Keep in mind that the fixed version might be
useless to run nodediffs against (does the CRC change?... it might).  It
would be a matter of copying the latest-diffed nodelist someplace that this
search BATch could use.

 PQ>> Ermmm... and, what version of Windows/DOS do you run (again)?
 RW> RUnning dos 6.22 standing alone on this box.   WHen we're dsl I'm
 RW> still not sure what I want to do with bbs, whether to be a masochist
 RW> and run it under windows 98 or something. Or, a linux dos emulator.
 RW> NOT sure how we're going yet but that's some time down the road.

Cool.  My Fido stuff has been running in Win9x-something since late '98. 
My Radius node is running with Win98SE on a P-90/64meg, along with a
JamNNTPd server and a POP3/SMTP server.  :)  (I do the right thing and
reboot it twice a week, Saturdays and Tuesdays.) It's one of six machines
on our LAN, all connected through a hub to ADSL.  IMHO, Win98SE has the
last -best- DOS available.

Which brings me to talk about some 'environment' matters.  This is where
DOS keeps variable information such as the PATH, TEMP and DIRCMD settings
in.  My DOS's environment space is set for 2048 bytes through a Config.Sys
setting, thusly: "SHELL=C:\COMMAND.COM C:\ /P /E:2048 /F". 
(There's more than just the environment space being allocated there, BTW.)

From my notes, the last search using this batch thing used 476 bytes all
up.  However, the default DOS environment space is 256 bytes.  With
anything involving nodelist processing you'll need, maybe, 137 bytes just
for a simple read operation on a single node's listing.  That's just one
variable this batch thing uses/sets.  (Also, according to some older notes,
even a variable name blows 5 bytes for each variable for DOS's
housekeeping.)

So, you will need to possibly adjust your machine's setting as well... say,
512 or 1024?

(BTW.  From personal experience... the DosBox linux emulator blows bigtime.
 It failed the first test: SET TMP=%TEMP%.  I ended-up with TMP=%TEMP%
instead of the -contents- of TEMP.  I may try VirtualBox sometime, as it
sounds a far better method but I need more RAM in this machine first.)

 RW> tHanks for all your work on this project Paul.  as I said, I'm
 RW> surprised nobody did it before .

I can't recall seeing anyone ask before.  :)

Cheers,
Paul.

... Only those who attempt the absurd achieve the impossible.
--- GoldED+/LNX 1.1.5-31012