[comp.sys.atari.st] New UUDECODE

NETWORK@FRSAC11.BITNET.UUCP (05/31/87)

To those true adventurers of the modern time:

There is a new version of uudecode to be tested, I will send it to you
upon request, if you provide a REAL address on EARN/BITNET, ARPANET, and
even UUCP if you can be reached from mcvax. (I will NOT try several time)

The thing is currently in .ttp arc + uuencoded form.
If nobody report real bugs, I will make the sources available.
I use MWC on the ST, and it can compile on UN*X too.
Try everything you believe a network + gateway can do.
If you get it, torture it with the most incredible, altough legal, files, then
report to me all problems encountered, with all meaningful informations, and
the offending file if possible. If you cant reach me, post the remarks to
INFO-ATARI16.

This program was written by a community of people, every time enhancing
the darn thing.

From the garden of:     Jean-Pierre H. Dumas
            Somewhere in France.

    network@frsac11 (EARN/BITNET)
    network%frsac11.bitnet@wiscvm.wisc.edu (ARPA)

Here is a short, but sufficient description.
================================================================================
This is the 30 May 1987 version of uud.ttp
It is named uud instead of uudecode, to avoid confusion between different
implementations of the uudecoding functionnality.

This version has the following capabilities:
1- Missing trailing blanks are OK. (I am on EARN/BITNET ...)
2- Some common internetwork transpositions are OK (e.g. grave accent instead of
   blank, tilde instead of circonflex) (I am still on EARN/BITNET)
3- Multi-file decoding is OK, by the 'include xxx' and 'begin part n' lines.
4- If the included files do not exist, uud.ttp will attempt to go on with the
   same file, in case the multiple files have just been concatenated upon
   reception.
5- There is a line sequence checking possible, allowing the detection of
   missing, duplicate, corrupted lines or files. (I have sloppy fingers when
   editing in emacs...)
   This feature is enabled if the first line of the uuencoded file
   is postfixed with a 'z', this is done automatically by the uue.ttp
   companion program, to be released soon.
6- A complete check of the uudecoding process is possible, by use of the
   -d flag. This will tell you about all whats going on in there, plus
   this is enabling a complete character by character check of the file
   while decoding. This should make some subtle file corruptions pop up,
   as all the uudecode do not check whats in a line before doing their job.
   The speed impact is minor: 1:03 with -d instead of 1:01 to fully decode
   a UNITERM.UUE file. (90+% of the time is I/O, not enough buffers for
   disk read/write) (this is done on my 1040 ST with SH204 HD)
7- If there is a character table present before each file, it will be used,
   this table is produced by the companion uue.
   The use of this table allows for some incredible recovery of files that
   went through some really brain damage gateway. No sweat.
8- You can decode a single file contaning several uuencoded files, at once;
   You give a complete month of INFO-ATARI16 digest, and voila all the
   nice PD .prg's, no editing, no sweat. (To be true, it's been a long time
   without *.prg in the digests, and I have not really tried it either, but
   you should do it.)
9- The files are completely compatible with true UN*X uudecode/uuencode.

The calls are:

uud uufile
uud -d uufile
uud -D uufile  (The desktop upper everything...)

If given no args, uud will tells you what version is this, and the usage.
In all case if there is anything else that a perfect decoding, you will
have some message and you'll need to press any key to exit. (I anybody
tell me how to tell if we are working from the desktop, or from a shell,
I will maque the pause valid only if run from the desktop. Some kind soul
with an exemple ? I suppose I have to look into the base page...)
================================================================================

hans@umd5.umd.edu (Hans Breitenlohner) (06/02/87)

In article <8705311430.AA00794@ucbvax.Berkeley.EDU> NETWORK@FRSAC11.BITNET writes:
>2- Some common internetwork transpositions are OK (e.g. grave accent instead of
>   blank, tilde instead of circonflex) (I am still on EARN/BITNET)
>

I would like to correct a misconception here.  The accent grave instead of
blanks is not a network transposition.  It is a conscious effort by the
latest U**X uuencode to avoid the problems associated with imbedded and
trailing blanks.  Unlike the real network transpositions it also does not
cause any problems if the decoding function is (take ascii character),
(subtract 0x20), (and with 0x3F).