Stars of Shareware: DFÜ und Kommunikation/SOS-DFUE.ISO/programm/internet/explorer/binary.xfr

Text File | 1992-08-06 | 11.6 KB | 258 lines

Beginner's Guide to Binaries, Version 1.2
-----------------------------------------

          HOW TO GET BINARIES VIA NEWS AND E-MAIL

(c) Copyright 1990 Brian O'Neill. Permission to distrbute this file freely
is granted, so long as it remains unmodified.

   One of the more troubling things to new users is using programs posted to
groups like comp.binaries.ibm.pc. This is in part due to the fact that
different utilties are needed to decode these files into something more
familiar to them. This manual explains how to retrieve binary files posted
to UseNet, and also received by e-mail.

   This manual is divided into several sections, each dealing with a
different aspect of what is needed. Here is the basic setup:

     I.   Formats and Standards
     II.  Minimum utilities, and how to get them
     III. Using UNIX to do the work
     IV.  Using MS-DOS to do the work
     V.   Handling archived files
     VI.  Downloading files from UNIX to MSDOS

   This manual was written with the new user in mind. If any section is
unclear to you, please respond to me via e-mail. Let me know what section,
paragraph, and what it was that was unclear. My address is below:

          Internet: [e-mail]
          UUCP: ...!ulowell!oneill

              I. FORMATS AND STANDARDS

   UseNet, and other networks, user groups, and BBS's, all work on some set
of standards for sharing files between a diverse arangement of systems. For
PC's, the most common standard for files is the archive. This is a binary
file which allows a user to place several files into one, while compressing
them to save space. Several utilities exist for making and extracting files
from archives. The most common are ZOO from Rahul Dhesi, and PKZIP from
PKWare. The ZOO standard has been adopted for use in the UseNet newsgroup
comp.binaries.ibm.pc for the distribution. Some BBS systems use ZIP, and
others use ARC, another format from System Enhancement Associates which is
slowly becoming obsolete, and is generally avoided due to recent litigation
in court (which I will not discuss here). You will need the appropriate
extraction program for the particular format you are dealing with, which
usually can be told by the extension name on the file (.ZOO, .ZIP, or .ARC).

   UseNet and e-mail are based on ASCII text. In other words, they are only
able to transmit ASCII files. In order to transmit binary files, they are
transformed into ASCII files first. This is accomplished using a program
called 'uuencode', another fairly common standard. The file can then be
decoded back into binary form using 'uudecode'. These programs are fairly
common among UNIX systems, and are increasingly available for MS-DOS. Once
again, you will need a version of uudecode to properly handle these files.

           II. MINIMUM UTILITIES, AND HOW TO GET THEM

   You will need some form of uudecode, either UNIX or MS-DOS, to decode the
binaries. If you do not have it for either of these, you must secure a copy
of the source code. If you have FTP access, you can get MS-DOS source code
from wsmr-simtel20.army.mil (File PD1:<MSDOS.STARTER>UUDECODE.PAS). If not,
you will have to get it from someone else.

   Also, you need an archive extractor. Small extract-only programs are
available, such as LOOZ by Rahul Dhesi. It will be assumed you can secure
one of these without problem, as with uudecode.

   Also available is the CBIP Starter's Kit, which contains BASIC, and DEBUG
source for uudecode, a uuencoded version of LOOZ (ZOO file extractor), and
instructions on how to make use of these files. This is enough to get you
started in dealing with the comp.binaries.ibm.pc newsgroup on UseNet.

            III. USING UNIX TO DO THE WORK

   UseNet is a network comprising mostly of UNIX-based machines, so
therefor it is quite common to use the system connected to UseNet to do most
of the work, rather than downloading from that system and doing all the work
on a PC. To do this, you will need the following basic utilities:

               uudecode
               editor (vi, emacs, etc.)
               cat (not necessary, but useful)
               combine (Also not necessary, but makes
                     everything easy)

   The first thing you must do is use your news reader program to save the
articles to files. If the uuencoded file spans more than one article, save
then to different articles. Check your manual pages for the news reader you
have for details. If you are using 'rn', do the following:

   From within the article, or at the end of the article, type 'w
filename'. Answer 'n' to the mailbox format question, and then continue. 's
filename' could also be used, but it saves additional header information
that is not needed, and would have to be edited out anyway.

   For one single file, things are easy. Use your favorite editor and delete
all lines before the 'begin' line, and all lines after the 'end' line. Then
you can give the command 'uudecode file', and the new file will be created
for you. Download the new file to your PC.

   For multiple files, you must edit each file. For the first file, delete
all lines before the 'begin', and go to the end of the file. A line of
uuencoded text looks like this:

M'YV01N2\:0-BSAPV:<Z4D0-B2ATW((+`80AB!@@8.73`F*$Q!X@8.7#@4""E

   You then must delete any lines after the last uuencoded line. Then
proceed to the remaining files, deleting all lines before and after the
uuencoded lines. When you come to the last file, delete all lines after the
'end' line. Save each file after it has been edited.

   Now, to decode them, give the following:

     cat file1 file2 file3 ... | uudecode

   Users of csh can use range specifiers to simplify the typing, such as
if you have a five part file, you can say:

          cat file[1-5] | uudecode

   This will create a file specified in the begin line. Then you can
download this file to your PC.

   Currently, Rahul Dhesi, the moderator of comp.binaries.ibm.pc, has made
available a short, two-line script which will take all parts of a program
posted to comp.binaries.ibm.pc, edit, combine, and uudecode them automatically.
Here is the shell script:

#! /bin/sh
cat $* | sed '/^END/,/^BEGIN/d' | uudecode

Type this in, do a 'chmod 755 combine' to make it executable, and run it
like this:

          combine file1 file2 ...

You will not have to do any editing at all of the files, as combine will do
this for you. As long as the files are fed in the proper order, all should
be well.

                 IV. USING A PC TO DO THE WORK

   There are several versions of UUDECODE for MS-DOS available. You need some
version of UUDECODE to proceed. Check the local archives, or the CBIP Starter's
Kit.

   After dowloading the files to the PC, edit out unwanted lines in the
file. If using a simple version of UUDECODE, concatenate all the files into
one large file. Then type:

          uudecode <filename>

You should then end up with the program you went through all the trouble
for.

               V. HANDLING ARCHIVE FILES

   An archive is simply a file containing several other files. Usually
these other files are compressed in some fashion, in order to save disk
space. These files are used generally for easier handling of several files.
There are many types currently in existence, such as .ARC, .ZOO, and .ZIP.
The instructions below can be used for most any type of archive, using the
appropriate programs.

   To extract files from an ZOO file, you must have some sort of extractor.
LOOZ from Rahul Dhesi (to be included in Starter's Kit) is an extract-only
program for dealing with ZOO files. The ZOO program itself (also by Rahul
Dhesi) can be used to both create and extract ZOO archives. To unpack an
entire archive, simply type one of the following:

               zoo -extract <archive>
               looz <archive>

This will unpack the entire contents into the current directory. You need
not specify the .ZOO extension. To extract particular files, simply specify
the filename after the archive name.


             VI. DOWNLOADING FILES FROM UNIX TO MSDOS

   To download the files, you will need the following:

     1) A file transfer protocol on your UNIX system (i.e. Kermit,
        Xmodem, Zmodem, etc.)

     2) A PC communications package that supports the same protocol

   Due to the large amount of protocols and communications packages
available, it would be next to impossible to describe all of them. choosing
the proper set up often depends on your situation. Some protocols are much
faster than others, yet cannot be used over some networks. For the purposes
of examples, I will use Kermit as the protocol (specifically C-Kermit), and
ProComm as the communications package, as they are in wide use and very
reliable over most networks. A more generic method for downloading follows
afterwards, but you must read the manuals to all programs to operate them
properly.

   First off is to determine what type of file you are downloading, whether
it is binary or ASCII. Usually, if you can read it, it's ASCII. Files with
extensions EXE, COM, and archive files such as ZOO are almost always binary.
On UNIX, you can say:

          file <filename>

this will usually tell you what type of file it is.

   Downloading ASCII text is easy. On UNIX type:

          kermit -s <filename>

the "-s" puts C-Kermit into "send" mode. You then instruct your terminal
program to receive a Kermit transfer. On ProComm, you would hit PgDn, and
then select Kermit off the menu, selection 2. ProComm will take care of the
rest, and you can watch it's progress. When it's done, ProComm will return
you to UNIX.

   Downloading binary files are a little more difficult. If you can dial in
to a UNIX host using 8-bit communications (such as 8-N-1), do so. Sometimes
the Login: prompt may look weird, but once the host knows, it will fix
itself. If you cannot use 8-bit settings (the host insists on 7 bits to be
readable), you may wish to transform the binary file into ASCII, using the
uuencode program, downloading as above for normally ASCII files, and
uudecoding the file on the PC.

   If you are able to use 8 bits, on UNIX type:

          kermit -is <filename>

the "-is" puts C-Kermit into "send, image" mode. Basically, "tell it like it
is". As above, hit PgDn, then 2 and ProComm will do the rest.

   For a more generic explanation of what to do, here is a step-by-step
version of the above, without specifics:

     1) Determine what type the file is (ASCII or binary)

     2) Initiate transfer on UNIX end. Usually accomplished by executing
     a program and giving the filename of what you wish to download. If
     it is a binary file, specify 'binary' or 'image' mode, usually as a
     switch on the command line.

     3) Escape back to the PC, and set for receiving a file using the
     same protocol as being sent with. Often done by hitting some Hot Key
     (usually Pg-Dn), and then specifying the protocol you are using.

   You really should read the manuals to any programs you wish to use for
downloading files. Programs change, and they are not all used the same way.
If you can't seem to get a program to work, consult someone using the same
programs, and see what it might be that you do differently.

--
=======================================================================
Brian O'Neill - Systems Manager, Computer Science, University of Lowell
Internet: [e-mail]                  [phone]
UUCP: harvard!ulowell!oneill

==== <g GUIDE>                       9 links in glossary topic