G'day Bob,

Replying to a message of Paul Edwards to Bob Lawrence on 19/05/97:

 BL>> Each of the .d files has a .x equivalent. What does this mean? Does
 BL>> the x-file identify the d-file, and can I rely on the .x extension
 BL>> on the X File? I know the truth is out there somewhere...

From my Watergate manual. It's a good explanation

>>

UUCP spool directory
--------------------

These batches are stored in your "spool directory", which is a 
sub-directory on your hard disk, for example C:\SPOOL\. Your UUCP 
uplink system has a sub-directory there, as well as your UUCP style 
downlink systems. Not your FidoNet style downlink systems, they have 
the inbound and outbound directories.

In these spool directories you will find files with the names *.X, 
*.D, *.XQT, *.DAT and *.CMD. The first two are incoming (inbound, 
received) files. WaterGate processes these files. The last three are 
outgoing (outbound, to be sent) files. The XQT file will end up on 
the other system's hard disk like a .X file and the .DAT file as a 
.D file. The .CMD file is used by the UUCICO (UUCP mailer) program 
and tells it which files to transfer.

The .X file is the so called "envelope" file and the .D file the 
letter itself. Each e-mail file has a .X file in which WaterGate 
finds the e-mail address of the recipient and a reference to the .D 
file, amongst others. In case of news, the .X file contains the 
recipient name "rnews".

Each e-mail message has its own .X and .D file. The news is bundled 
and you will find one .X file for each .D file with a number of news 
message in it. The .D file is mostly limited by size, not by number 
of news messages.

Compressed news and batch headers
---------------------------------

To reduce the transfer time, news batches are mostly compressed. 
E-mail is never compressed. There are two forms of compression used 
with UUCP: the older 12-bit or 16-bit "compress" and the nowadays 
more common "gzip".

Because of this compression, you cannot read the .D files with news 
directory. You have to decompress them first.

To make it easier for a script-based UNIX machine to detect the 
compression format, a special "batch header" is added to the start 
of the compressed file. When the file is compressed with normal 
compress, you will find the header "#! cunbatch" there. When it is 
compressed with gzip you will find the header "#! gunbatch" or "#! 
zunbatch" there.

WaterGate automatically detects all these headers and compressed 
formats and decompressed the .D files.

>>

 BL>> Or to put it another way... why the FUCK don't you use .z!

IOW, it's  not Paul this time :-)

 BL>> How do I identify actual mail? The mad "#! cunbatch"
line is not on
 BL>> all mail... 

See above. 

 BL>> Has the whole world gone mad?

What you are seeing is the culmination of puting all of those RFC documents
together, tumbling them around and assuming the first page spewed out was
Gods word on uucp connects.

It's as good an explanation as you'll get anywhere else.

 BL>> We're getting there... now you need an Areafix so I can leave all
 BL>> that C-programming crap and get a bit more AVT. This seems to be 
 BL>> incredibly easy at this end...

Unfortunately, there isn't an areafix for uucp :-( You've more or less got
to ask Paul to get it for you and then get yourself hooked in.

 BL>> THESE SMARTARSE UNIX-WANKERS GIVE ME THE SHITS!!!!!!

Refer to RFC comment above :-)

 BL>> First, they put a #1 cunthook header their compress doesn't handle,
 PE> Doesn't fxuucico do that by any chance?

Bob, try and look for a package with waffle in it's name. This might do
everything instead of just getting the mail as fxuucico does. If it does,
you shouldn't have to worry about the .d or .x extensions. From what I've
read, fxuucico is a replacement for waffles uucico, but it doesn't include
all the other goodies. Does that make sense?

Regards...Jeff (jeff{at}matra.com.au)

--- WtrGate+ 0.92.p5 gamma sn 500