[Part III] How Do I Upload, Download, Decode?
by LuvBunny and others

* INTRODUCTION

This document is a step-by-step guide for downloading, decoding, and
uploading binaries from and to ABMA. In order to use the information
in this guide, you will need to download at least some of the volumes
of the ABMA Toolkit. You can request the toolkit in the group at any
time. Or you can download the tools that you need from web and FTP
sites on the Internet by using the URLs in the "Links" section of this
guide; you might find that quicker and more convenient.

This guide is designed to be complete, concise, and comprehensible to
folks with little or no ABMA experience. This is not intended as a
replacement for the FAQ, nor is it an 'official' work. We put it
together based on the things people seemed to ask the most about.
Please feel free to redistribute this to anyone in the group who
requests help.

The information in this Guide is the product of many people's knowledge
and good will. I [LuvBunny] have compiled and arranged a lot of
information here. Most of it is stuff that I learned from other people
in ABMA. I didn't know any of it three years ago. So please remember
that ABMA is a group of people, not a single person. Some of them are
Lazarus Long, Webster, Pinochle, Alkoholic, Al Uminum, Shivering,
Tricky, Exception, sQuid Pro Quo, Fart Knocker, Owl, Jamaican, Mooch,
NetGuru, Freedom1, Grizzle, and LOTS of others who deserve credit for
helping to make ABMA what it is. [Editor's note: See the end of this
file for some more recent additions to the ABMA honor roll. Thanks
again to all uploaders.]


* DOWNLOADING

- INTRODUCTION

To download you will need a news reader, a set of decoding and
decompression utilities, and probably a registration number. A news
reader is a specialized and dedicated application for viewing
newsgroups and the contents of them; it will correctly download the
files that are attached to the news messages. Once you have downloaded
the files from ABMA, you will need to convert them into a form that you
can use. Some version of StuffIt is the minimum that you should have
in order to convert and decode the files; if you are not using
YA-NewsWatcher (see below), then you will also need a copy of
YA-Decoder or some other good BinHex decoder. Last, you will need
to acquire a serial or registration number to evaluate the application
that you have downloaded and installed. These can usually be found
in one or more databases that are periodically posted here in ABMA.
Also, check the newsgroup if
your news server carries it.

Posts are usually uploaded in segments. You will need to download all
of the segments, decode them, and then recombine (join) them.

We recommend that you use YA-NewsWatcher 4.2.6 or later (or 3.1.8 if
necessary) for downloading binaries. The following instructions assume
that you will be using YA-NewsWatcher to download. Use the latest
version possible that will work with your computer hardware and Mac OS
version. I highly recommend this application for downloading attached
files. Please make use of YA-NW's extensive online help; it should
answer most questions about the program. YA-NW also comes with a
documentation file (user's guide) that should be your main source of
information about the program aside from the on-line help.

As of 17 April 2000, the current version of YA-NW is 5.0.1, and the
last 4.x version is 4.2.6. Both version 5.x and version 4.x of YA-NW
require a Power Mac (PPC processor) and Mac OS 8.5 or later. Version
3.1.8 of YA-NW is still available for use on both 68k Macs and
Power Macs running Mac OS versions earlier than 8.5. See the "LINKS"
section below for URLs for downloading these programs.

######################################################################
# --- R E A D M E --- #
######################################################################
### VERY IMPORTANT WARNING ### [22 Apr 2000]: Several people have
reported in ABMA that binaries that were uploaded with YA-Newswatcher
versions 5.0 and 5.0.1 appear to be somehow corrupted and are not
decodable after downloading, no matter what decoding program is used.
This is definitely a problem. It's being caused by a bug in YANW that
corrupts BinHex uploads when two upload streams are active at once.
Until a new 5.x release of YA-NW (5.0.2) has replaced the old ones,
PLEASE DO NOT USE YA-NW VERSION 5 FOR UPLOADING BINARIES.
USE VERSION 4.2.6 INSTEAD. This does not affect downloading, so if
you don't do any uploading of binaries at all, then it's OK to install
YANW 5.0 or later.
######################################################################

WARNING: It is not advisable to either download or post (upload)
attached binaries with Outlook Express, Netscape Communicator,
Hogwasher, or most of the other Internet applications with news-reading
or posting abilities. Most if not all of them have minor to serious
problems with downloading and decoding binaries from ABMA. They are
mostly intended for downloading uuencoded binaries like JPEGs and GIFs,
not BinHex files. They are likely to corrupt your downloads and cause
problems. Instead, use one of the NewsWatcher programs (YA or MT);
they're free, and they're widely available on the net.


- SETTING UP YA-NEWSWATCHER

Here are some basic instructions for setting up the preferences
in YA-Newswatcher (in the Edit menu, select 'Preferences...').

IMPORTANT: In the section of YA-NW 4.x preferences that's titled
'News Server', be sure to check the two options 'Use Additional
Connection for Queued Transfers' and 'Use Additional Connection for
Queued Subject Opening'. That turns on multi-threading in YA-NW.
If you don't do this, you will be seriously inconveniencing yourself
and you probably won't like using YA-NW at all.

Next, go to the 'Saved Articles' section and check 'Save Encoded Text'.
Then go to 'Extracting Binaries' and check 'Keep No Attachment
Found and Partial Downloads', 'Decode Extracted Binary Attachments',
and 'Use Slack BinHex Decoding Rules'. You can also select a default
location for saving the downloads here. In this same section, you will
need to uncheck the option 'Add post info text to downloaded files'
if you intend to use CRCs to verify the files that you download.
(CRCs are described later in this document; if you haven't got that
far yet, then just uncheck that option in the prefs and forget about
it for now.)

There are a lot of other options that you can play with, but they are
not specific to ABMA downloading, so you should simply consult YA-NW's
on-line help (or, even better, its user's guide) regarding them.
A few of the selections that you checked in the prefs are only
important when doing certain types of downloading, but they are the
minimum that you should set for general use.

YA-NW has built-in decoding for several types of files, including
BinHex, uuencode, and MIME Base-64. If your version of YA-NW does
not decode a particular file, then you can use YA-Decoder or StuffIt
(more about that later).


- GETTING ALL THE PIECES

We suggest that you always open ABMA from the "full group list"
and not from a subscription list (sometimes called a "user group list").
The reason for this is that opening from the full group list allows you
to see all messages posted to ABMA, whereas opening from a subscription
list will exclude any messages that were marked as "read" during
previous sessions when the subscription list was open. The
subscription list is useful if you do want to exclude messages that
you have already read, but be aware that the exclusion could remove
segments from files that you might want to download. (The files will
still be there on your server, but you won't see them in the subject
list.) In YA-Newswatcher, in the Full Group List window, you can use
the Find command (Command-F) to go to a particular group by name,
without having to manually scroll up or down to it.

Parts and segments: There is often a lot of confusion about these.
Parts are not segments and segments are not parts.

Segments are created by DropSegment or the StuffIt Deluxe application.
A single segment is attached to a message and then posted or downloaded.
It is very useful to know how many segments are in a particular post.

Parts are created when a single attachment is uploaded by a news client
like YA-NewsWatcher. The size of the part is determined by the 'News
Part Size' setting in YA-NewsWatcher or, if some other program is used
for posting, by a similar setting in that program. As we mentioned
before, this should be set to between 200 KB and 495 KB. The number
of parts depends, then, on the total size of the attachment. How many
parts are in an attachment is not useful information. The number of,
creation of, and combining of parts should be handled by the news
client, and these are not things that you should need to be concerned
about when downloading. There is no need to post this information
when uploading.

You will not be able to download a complete attached file unless all
of the parts are available on the server. YA-NW will display a '<'
character just to the left of the author in the 'count' column if the
attachment is incomplete. More later on what to do if this happens.

Select the article that you want to download with the mouse
(single-click).

Go to the menu bar and select 'Extract Binaries' from the 'Special'
menu, or just press Command-B. You can select more than one article
at a time before giving the 'Extract Binaries' command, and YA-NW will
put them all into its queue and will then download them all, one at
a time.

YA-NW will create and write to a temp file as it downloads. This file
contains the raw BinHex text of the attachment. YA-NW should handle the
BinHex decoding transparently. When it is finished, you should be left
with a StuffIt segment or a complete archive or an .img file, or
whatever was uploaded.

You will need to repeat this downloading process until you have grabbed
all of the segments of the particular file that you are getting. Most
posters will indicate somewhere how many total segments there are in the
post. This information might be in a separate readme file; or very often
it will be in "part zero" of the post. If a post that you intend to
download does have a "part zero", then you should read that in your
newsreader before beginning the download. In addition to the number of
parts, it may have a serial number for the program or other information
that the uploader wants you to know. When you have that "part zero"
article open in your newsreader (or selected in the subject-list
window), you can press Command-S to save it to disk on your computer
for future reference.

When you have downloaded and decoded all of the segments of the post,
you will need to combine them. Make sure that they are all in the same
folder, which will generally be whatever download folder you have set
in your newsreader.

A note about attachments: If a segment has been properly BinHexed by
the uploader, then if you look at the article in a newsreader or if
you look at a text file containing it, you should see a line saying:
"This file must be converted with BinHex 4.0". That statement is
somewhat misleading. Although a BinHex 4.0 application does exist,
it is better to use one of YA-Decoder, StuffIt Expander with
Expander Exhancer, or MindExpander to unBinHex the segment, since
those apps are more versatile. And of course YA-NW can do the
unBinHexing itself if its prefs are set correctly, so no external
decoding program is needed in that case.


- PUTTING IT TOGETHER

Make sure that the segments all have the StuffIt icon. If they do not,
there could have been some problems in decoding the post, which we will
cover later.

We recommend that you use StuffIt Deluxe 4.0 or StuffIt Expander 4.0.
Try not to use version 4.5 or 5.x of these programs, because both of
those can cause some serious problems when attempting to decode.
However, some archives can only be unstuffed using version 5.0 or
later (these 'SIT5' archives have a red dot in their icons), in which
case you will have no choice but to use Stuffit Expander 5.x at least.
Double-click or drag and drop the first segment file onto the decoding
app. If you use Expander, make sure that you have let it know to
combine segments in the preferences.

If you do need to use version 5 of StuffIt to unstuff a 'SIT5' archive,
then it's a good idea to get version 5.5 or later of Stuffit Expander,
and possibly DropStuff/Expander Enhancer as well. Version 5.5 is
faster and less bug-ridden than earlier 5.x versions of StuffIt.
See the "LINKS" section below for URLs. Also, you may need to
increase the memory allocation of StuffIt Expander 5.x in its Get Info
window; if the minimum and preferred sizes shown there are 1024K
or less, then increase each of them to at least, say, 1500K, or else
the program might fail to unstuff some files.

Often, posters will rename the segments of their posts. This can cause
a little confusion with StuffIt, as it may ask you where a segment is.
Just help it out by finding the approximate segment name with the same
number as the one it is asking for. It's the segment number that's
important here, not the full file name (which doesn't have to match
exactly).


- EXPANDING AND INSTALLING IT

When StuffIt is finished, you will have a StuffIt archive (.sit or
.sea), or a disk image (.img or .smi), or an installer application,
or (rarely) some other kind of file.

.sit
StuffIt Deluxe 4.0, StuffIt Expander 4.0, or MindExpander is the
recommended application for dealing with a .sit file (but see the notes
in the previous section about the need for using StuffIt version 5
in some cases). If the .sit file does not open with StuffIt, then
try MindExpander. If it doesn't open with either, then it may be
corrupted. Check the file type and creator; if the creator is not
'SIT!' or the type is not one of 'SIT5', 'SITD', or 'SIT!', then it's
not a StuffIt archive or it was somehow corrupted in transit. Maybe
the uploader used uuencode rather than BinHex by mistake.

.sea
This is a self-extracting archive, which means that in fact it's an
application file. It should not require the use of an external
unstuffing utility; instead, it should extract itself when you open
it (double-click it, or select it and press Command-O).

.img
Use DiskCopy 6.3.2 or later, or ShrinkWrap 3.5 or later, with these
files.
Mounting a disk image creates a virtual disk on your desktop that
behaves
as though it were a real disk. Usually the mounted disk contains an
installer that you will need to run to install the program. You can
dismount/unmount the disk by dragging it (the mounted disk, NOT the
image file!) to the trash.

.img (again)
Some .img files that are posted to the group have been created by Toast
or another CD-burning program rather than by DiskCopy or ShrinkWrap.
(By default, Toast images will end with "쩳mage" or ".image", although
Toast itself does not provide this extension. Applications such as
CD-Copy are usually used to create the Toast image of a CD; Toast
is used to burn it.) To use these, you will need either Toast or
the stand-alone "Toast Image Mounter", which is a small application
that will mount the image file just as DiskCopy or ShrinkWrap will
do for the image-file formats that they handle. (Post a request
to the group if you need the Toast Image Mounter.)

.smi
This is a self-mounting disk image. Just as with a self-extracting
archive, it should not require an application, and should mount itself
when you open it.

temp
This is YA-NW's temporary file. It is a BinHex text file. YA-Decoder
is best for painless decoding of files. Sometimes you can use
StuffIt or MindExpander, but they are not as good. Usually, YA-NW
handles the decoding of this file by itself. If you are using some
other newsreader, then you will probably want to configure it to
automatically call an external decoding program like YA-Decoder to
decode the binaries that you download. (YA-Decoder can handle
uuencode and Base64 as well as BinHex, so it will also work in other
newsgroups for decoding images, movies, and other files.)

Sometimes your target will be wrapped up in a number of encodings,
like an onion. If this happens, be persistent; you will get to the
center of it eventually. This sometimes happens when the person doing
the uploading inadvertently encodes a file before posting it, not
knowing that his newsreader will do another encoding pass itself.


- TROUBLESHOOTING YOUR DOWNLOADS

CRCs, or "cyclic redundancy checks" (a fancy name for checksums), are a
means of verifying that the file that you have downloaded is the exact
file, byte for byte, that was uploaded. Many posters will include
CRCs in a separate post or in part zero of their upload, or are willing
to post CRCs on request.

After you have downloaded all of the segments, if you have problems
combining and/or decoding them, use List Files to make your own CRCs
of the download. Then compare your CRCs with the poster's. They
should be exactly the same. If not, then any file whose CRC is
different from the poster's CRC is part of the problem. (Don't stop
if you find one mismatch; check the CRC of every segment.) At this
point you might want to re-download the problem segment or segments
(possibly using another computer and/or server, if one is available
to you). If you fail to get a proper CRC match again, then ask for
a repost of the problem file or files, and indicate in your message
to the group that your CRCs don't match the posted ones.


- MAKING CRCS

Use List Files 2.6 (see the "Links" section for an URL).

From the menu "Worksets", select Checksum and CRC

Options: List Formatting: Uncheck every box except File name and
Global Checksum and CRC (although it comes in handy to let it generate
the files sizes too; that makes it easier to spot the corrupted file,
in some cases).

Drag and Drop the folder containing only the segments of the post onto
List Files.

List Files will work for a while and then spit out the CRCs, as a bunch
of text which you should compare to the uploader's CRCs.


- CHECKING CRCS

(This section was provided by t.v.c.15.)

Once you've downloaded all the segments, put them in a folder and drop
the folder onto a utility called ListFilesFAT. (This is the same "List
Files 2.6" that we mentioned earlier.) This will generate a list of
CRCs in the same order as the list of good CRCs. Save the CRC list
to a text file. Import the text file into a spreadsheet like Excel.
Then split that column of text into separate columns (in Excel, the
menu command is "Data->Text to Columns") using "comma delimited."
Import the original list of CRCs in the same way. Copy and paste the
CRC data together into one spreadsheet, aligning the columns of data
so that each original CRC and downloaded CRC are on the same horizontal
row. Use a simple true/false test on the two CRCs: Go one cell to
the right. Type an = sign to indicate you're entering a formula. Then
click on the original CRC. Then hit the equal sign again. Then click
on the CRC that you downloaded. Then hit Enter. What you've basically
done is to enter a formula that says and it will
evaluate to true or false. Then fill that formula down to the bottom
of the list of CRCs. If the cell is false, you've got a bad seg
(or perhaps the list is out of order due to a missing seg). Just
remember, if the CRCs are equal, the files match. Actually, really
intense mathematical types (like me) would only be satisfied if BOTH
the file length AND the CRC match; there is about a one in ten billion
chance of CRCs matching when files of different lengths are compared.
Strictly speaking, to be a valid file check, both the file length and
the CRC must match.


- AN ADDITIONAL NOTE ON CHECKING CRCS

(This section was provided by Zapf.)

I wanted to throw out some things I learned after my fiasco downloading
all 98 segments of OS 9. Here's what I found out:

Toolsetter did the original reposting that I latched onto. I missed
a lot and started grabbing others' re-reposts. Comparing sizes, I
noticed differences between those from TS and others.

I found out that you need to uncheck the YA-NW pref 'Extracting
Binaries: Add post info text to downloaded files'. I was doing
exactly that. After disabling that option, all new DLs were exactly
as listed. In the end it turned out that two segments were corrupted,
and were eventually replaced. Knowing how to read the CRCs was the
only way to narrow down which segments were bad.

But I was intrigued as to why so many previous downloads had worked
correctly for me, and reassembled just fine, but I had had my settings
wrong for months. What I learned is that if you have 'Add post info
text' selected, you will create an extra, unnecessary 'YANF' resource.
That contains the text, and it will throw off your size/checksum/CRC
numbers. Removing the 'YANF' resource from the .sit file segments
with ResEdit will not correct it. But the resource also will not
adversely affect your ability to reassemble the pieces.

Bottom line: Using the 'Add post info text' option (having it turned
on before starting a download) yields wrong size/checksum/CRC
numbers, but the segments are fine. You just can't rely on the CRC
list to find corrupt ones any more, so be sure to turn off that
option. And if you haven't been using CRCs for comparison, then
you probably shouldn't be reposting segments from other people,
as you might be creating differences.


- CREATORS AND FILE TYPES

Every Mac file has a "signature" by which the computer identifies it;
this signature has two parts, the creator and the type.

The creator tells the Mac what application the file belongs to (which
can be itself if it is an application file), and the type tells the Mac
exactly what kind of file it is (application, text, preferences,
StuffIt segment, and so on).

Sometimes this information is missing because the file is only slightly
corrupted and can be fixed by manually changing the type and/or creator.
Often, though, if this information is missing the file is useless
because more extensive corruption has occurred. Without the
type/creator information, the file will appear as a "generic document",
without any custom or application icon.

If your downloaded segments have bad CRCs, you should try to get
reposts of the bad segments before you bother with setting the file
type or creator.

Following is a table of applications, their native files, and the type
that each one has. You can set these manually using a utility like
FileTyper, File Buddy, or ResEdit. Note that a few of these, like
.cpt files and DiskCopy 4 images, are rarely seen these days.

FILE / TYPE / CREATOR / SOFTWARE
--------------------------------
.sit SIT5 SIT! StuffIt 5.x archive
.sit SITD SIT! StuffIt 3.x or 4.x archive
.sit SIT! SIT! StuffIt 1.5.1 archive
.sit SITD SITx Generic StuffIt archive
(the "creator" is the StuffIt Expander app)
.1 Seg1 SIT! StuffIt segment 1
.2 Seg2 SIT! StuffIt segment 2
.3 Seg3 SIT! StuffIt segment 3
.4 and up
SegN SIT! Stuffit segment, higher than 3
.sea APPL aust StuffIt self-extracting archive
.cpt PACT CPCT Compact Pro archive
.sea APPL EXTR Compact Pro self-extracting archive
.img dImg ddsk Disk Copy 6.x image
.img rohd ddsk Disk Copy 6.x read-only image
.smi APPL oneb Disk Copy 6.x self-mounting image
.img dImg Wrap ShrinkWrap disk image
.smi APPL sImg ShrinkWrap self-mounting disk image
.img hdrv Wrap ShrinkWrap drive container
.img dImg dCpy DiskCopy 4.2 image
.img [or 쩳mage or .image / any file meant to be a Toast image]
hImg CDr3 Adaptec Toast HFS image
HImg CDr3 Adaptec Toast Hybrid image
iImg CDr3 Adaptec Toast ISO 9660 image
DImg CDr3 Adaptec Toast CD Extra image
[Multitrack XA, CDDA with Data]
GImg CDr3 Generic Toast CDROM image
Sd2f CDr3 CDDA [CD-Digital Audio] Image
[very unlikely to appear on ABMA]
CDIR CDr3 Adaptec Toast VideoCD image
AIFF CDr3 Adaptec Toast AIFF file
UImg CDr3 Adaptec Toast UDF image
[generally not used]
-- APPL kajr Apple installer
-- APPL VIS3 Installer Vise 3.x installer
-- APPL Vis5 Installer Vise 5.x installer
-- APPL STi0 StuffIt Installermaker installer


- PARTIAL ATTACHMENTS

There may be times when you want to grab only selected parts of an
attachment (to combine with a partial attachment) or manually download
and decode the parts (if there is some error in the part order or
contents).

Open the thread of the attachment by clicking the blue, leftmost arrow
(the disclosure triangle) in YA-NW's subject window. This shows all
the parts.

Hold down the Command key and select the specific parts that you want.
They should then be highlighted.

Go to the 'File' menu and select 'Save As'; you will be presented with
a Save dialog.

Be sure to select the check box 'Save encoded text'.

Choose a location on your hard drive and select 'Save'.

When the download completes, what you will have is a BinHex-encoded
text file containing the parts that you selected. You will need the
"uucd" application to combine and reorder partial attachments, and
"YA-Decoder" to decode the resulting complete BinHex file. After
the complete file has been decoded, you should be left with an intact,
complete archive that StuffIt should be able to handle (or a complete
.smi file, or whatever was originally uploaded).


- DOWNLOADING MAC BINARIES FROM ABMA TO A PC

(This section was provided by Doctor.)

From a dyed-in-the-wool Mac user who also uses a PC, here's the correct
way to extract Mac binaries using Forte Agent:

1) Sometimes Agent will expand all the parts of a segment or file. To
re-combine them, ctrl-click all of the individual parts to select
them, and then go to the "Join Segments" button. This will bring
up a window that allows you to rearrange them into the correct order
using the "Up" and "Down" buttons in the window. Once they are in
the correct order, select "Join".

2) To save your segments or files in a format for the Mac, Ctrl-click
all of the files you want, even if they are unrelated, e.g.: Mac OS8
segments 3, 5, and 12, as well as Norton AntiVirus update, and
anything else you want.

3) Then press the "Save As" button. This will bring up a dialog window.
In that window select: "Save as type: (*.txt). In the check-box,
be sure to select: "Save raw (unformatted) message." Name the
file anything you want, and save it to a drive or folder that's
accessible to your Mac: Zip, Jaz, networked drive, or whatever
you have.

4) Take the file that you just saved and, on your Mac, drop it onto
any Stuffit Expander except version 4.5 (that version has problems
with this type of file). When Expander is done, it will leave you
with a folder with the same name as your file, containing all of
your Mac files complete with resource forks and all.