ciderpress/ReadMe.htm

314 lines
14 KiB
HTML

<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=windows-1252">
<meta name="GENERATOR" content="Microsoft FrontPage 4.0">
<meta name="ProgId" content="FrontPage.Editor.Document">
<title>CiderPress Source README</title>
</head>
<body>
<h1>CiderPress Source README</h1>
Release notes for v3.0.0<br>
Last updated: 25-Mar-2007
<h2>Contents:</h2>
<ul>
<li><a href="#intro">What's Ciderpress?</a>
<li><a href="#build">Building the Sources</a>
<li><a href="#notes">Source Notes</a>
</ul>
<hr>
<h2><a name="intro">What's CiderPress?</a></h2>
<p>CiderPress is a Windows utility for manipulating Apple II disk images
and file archives. It supports all major disk and archive formats used
on the Apple II and by Apple II emulators.
<p>CiderPress was sold by faddenSoft, LLC as a shareware product for about four
years, starting in March 2003.
<h3>Why Bother?</h3>
<p>Back in 2002 I decided it was time to learn how to write an application for
Microsoft Windows. I had been a professional software engineer for many
years -- including 2.5 years at Microsoft! -- but had never written a
Windows program more complex than "Hello, world!".
<p>I decided to write a Windows version of GS/ShrinkIt. I had already written
NufxLib, which handled all of the ShrinkIt stuff, so I could focus
on writing the Windows user interface code.
<p>Somewhere in the early stages of the project, it occurred to me that a
disk image isn't substantially different from a file archive. They're
both just collections of files laid out in a well-defined manner. The
decision to handle disk images as well as ShrinkIt archives seemed like
a simple improvement at the time. The rest is history.
<p>CiderPress has allowed me to explore a variety of interesting technologies.
It has five different ways of reading a block from physical media, depending
on your operating system and what sort of device you're reading from. I
was able to take what I learned from a digital signal processing textbook
and apply it to a real-world problem (decoding Apple II cassette data). It
is also my first Shareware product, not to mention the initial product of
my first small business venture (faddenSoft, LLC).
<p>I could have written other things. No doubt they would have made more money.
CiderPress is something that I find very useful, however, in the pursuit of
my Apple II hobby.
<p>Above all, this has been a labor of love. I have tried to get the details
right, because in the end it's the little things that mean the difference
between "good" and merely "good enough".</p>
<h3>License</h3>
<p>The source code to CiderPress is being made available under the BSD license:</p>
<blockquote><code>
Copyright (c) 2007, FaddenSoft, LLC<br>
All rights reserved.
<p>Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
<ul>
<li> Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
<li> Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in the
documentation and/or other materials provided with the distribution.
<li> Neither the name of FaddenSoft, LLC nor the
names of its contributors may be used to endorse or promote products
derived from this software without specific prior written permission.
</ul>
<p>THIS SOFTWARE IS PROVIDED BY FaddenSoft, LLC ``AS IS'' AND ANY
EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL FaddenSoft, LLC BE LIABLE FOR ANY
DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
</code></blockquote>
CiderPress requires three other libraries, all of which are included as either
source or binaries:
<ul>
<li>NufxLib, also available under the BSD license.
<li>Zlib, available under the Zlib license.
<li>libhfs, available under the GPL license.
</ul>
<p>The license allows you to do a great many things. For example, you could
take the source code to CiderPress, compile it, and sell it. I'm not sure why
anyone would buy it, but you're legally allowed to do so, as long as you retain
the appropriate copyright notice.</p>
<p>If you retain libhfs, any changes you make to any part of CiderPress must
be made available, due to the "viral" nature of the GPL license. If this is
not acceptable, you can remove HFS disk image support from CiderPress (look for
&quot;EXCISE_GPL_CODE&quot; in DiskImg.h).</p>
<hr>
<h2><a name="build">Building the Sources</a></h2>
<p>How to build your own copy of CiderPress.
<h3>Windows</h3>
<p>The sources are distributed with the necessary project/solution files for
Microsoft Visual Studio 6.0 and Microsoft Visual Studio 2003.&nbsp; All you
really need is the C++ compiler.&nbsp; The free or student editions of the
compilers will probably work.
<p>I usually work on CiderPress in Visual C++ 6.0,
and use that when building for distribution.&nbsp; I chose to stick with 6.0 for
CiderPress because you need an additional MSVC DLL when you build with
2003.&nbsp; Also, the newer tools cause the newer file selection dialogs to be
used, and the custom dialog code gets all wonky.</p>
<p>In 6.0, you need to select &quot;app&quot; as the build target (Build-&gt;Set
active configuration), and hit F7.&nbsp; This will build everything except MDC,
which you can then build by selecting &quot;mdc&quot; and hitting F7.&nbsp; You
can also configure a batch build.&nbsp; In 2003 just select &quot;build
solution&quot;.&nbsp; The necessary prebuilt libraries should be there for both
&quot;debug&quot; and &quot;release&quot; builds.&nbsp; (You can tell how it has
been built by looking at the version string in the About box.)</p>
<p>When the build completes, the necessary DLLs are copied around so you can
execute the binary by launching it within Visual Studio (hit F5).&nbsp; You will
get a warning about not being able to find the NiftyList data file unless you
copy that from &quot;DIST&quot; into the app source directory.</p>
<p>The distribution comes with prebuilt versions of NufxLib2 and zlib in DLL
form.&nbsp; If you like, you can download and compile your own.</p>
<p>There are two things you can't create with the standard tools:</p>
<ul>
<li>The help text was created with HelpMatic Pro. The source file is in a
proprietary format. It should probably be converted to HTML Help.</li>
<li>The installation packages are created with DeployMaster. The
".deploy" file specifies how the package is built. </li>
</ul>
<p>Compiled help files are included, so you can still generate a new version of
CiderPress even if you can't update the help.&nbsp; You don't strictly need the installer either, though it is
quite handy, and the uninstaller will clean up the registry entries CiderPress
creates.</p>
<h3>Linux</h3>
<p>The &quot;linux&quot; directory has a few command-line utilities and a simple
makefile.&nbsp; It will build the diskimg and HFS libraries, then the sample utilities:</p>
<ul>
<li><code>getfile disk-image filename</code> -- extract a file from a disk
image.&nbsp; The file is written to stdout.</li>
<li><code>makedisk {dos|prodos|pascal} size image-filename.po file1 ...</code>
-- create a new disk image, with the specified size and format, and copy the
specified files onto it.</li>
<li><code>mdc file1 ...</code> -- this is equivalent to the MDC application
for Windows.&nbsp; It recursively scans all files and directories specified,
displaying the contents of any disk images it finds.</li>
</ul>
<p>This are mostly intended for testing and illustration of the interfaces, but
they can be useful as well.&nbsp; Some other code is also provided:</p>
<ul>
<li><code>iconv infile outfile</code> -- convert an image from one format to
another.&nbsp; This was used for testing.</li>
<li><code>packddd infile outfile</code> -- the DDD code was originally
developed under Linux.&nbsp; This code is here for historical reasons.</li>
<li><code>sstasm part1 part2</code> -- the SST re-assembly code was originally
developed under Linux.&nbsp; The code is here for historical reasons.</li>
</ul>
<p>The &quot;prebuilt&quot; directory has a pre-built copy of NufxLib, so you
don't have to build your own.</p>
<p>If you're planning to hack on this stuff, use &quot;make depend&quot; before
&quot;make&quot; to fill in the dependencies.</p>
<hr>
<h2><a name="notes">Source Notes</a></h2>
<p>Some notes on what you'll find in the various directories.
<h3>Main Application</h3>
<p>This is highly Windows-centric.&nbsp; My goal was to learn how to write a
Windows application, so I made no pretense at portability.&nbsp; For better or
worse, I avoided the Visual Studio &quot;wizards&quot; for the dialogs.
<p>Much of the user interface text is in the resource file.&nbsp; Much is not,
especially when it comes to error messages.&nbsp; This will need to be addressed
if internationalization is attempted.
<p>It may be possible to convert this for use
with wxWidgets, which uses an MFC-like structure, and runs on Mac and Linux as
well. The greatest barrier to
entry is probably the heavy reliance on the Rich Edit control. Despite
its bug-ridden history, the Rich Edit control allowed me to let Windows
deal with a lot of text formatting and image display stuff.
<h3>MDC Application</h3>
<p>MDC (Multi-Disk Catalog) was written as a simple demonstration of the value of having the DiskImg
code in a DLL instead of meshed with the main application.&nbsp; There's not
much to it, and it hasn't changed substantially since it was first written.
<h3>DiskImg Library</h3>
<p>This library provides access to disk images.&nbsp; It automatically handles a
wide variety of formats.
<p>This library can be built under Linux or Windows. One of my key motivations
for making it work under Linux was the availability of "valgrind". Similar
tools for Windows usually very expensive or inferior (or both).
<p>The basic classes, defined in DiskImg.h, are:
<ul>
<li>DiskImg.&nbsp; This represents a single disk image, which may have
sub-images.&nbsp; Operations on a DiskImg are roughly equivalent to a device
driver: you can read and write blocks, detect image formats, and create new
images.</li>
<li>DiskFS.&nbsp; Paired with a DiskImg, this is roughly equivalent to a GS/OS
FST (FileSystem Translator).&nbsp; You can perform file operations like
rename and delete, format disks, see how much space is available, and search
for sub-volumes.</li>
<li>A2File.&nbsp; Represents a file on a DiskFS.&nbsp; This holds the file's
name, attributes, track/sector or block lists, and provides a call to open a
file.</li>
<li>A2FileDescr.&nbsp; Represents an open file.&nbsp; You can read or write
data.</li>
</ul>
<p>Sub-classes are defined in DiskImgDetail.h.&nbsp; Most applications won't
need to access this header file.&nbsp; Each Apple II filesystem defines
sub-classes of DiskFS, A2File, and A2FileDescr.</p>
<p>In an ideal world, the code would mimic the GS/OS file operations.&nbsp; In
practice, CiderPress didn't need the full range of capabilities, so the
functions have some basic limitations:
<ul>
<li>On ProDOS and HFS, you can only open one fork at a time. This allowed me
to use simpler data structures.
<li>Files are expected to be written in one large chunk. This reduced the
complexity of the task enormously, because there's so much less that can
go wrong.
</ul>
<p>Some things that can be improved:
<ul>
<li>The overall structure of the filesystem handlers evolved over time. There
is some amount of redundancy that could be factored out.
<li>The API, especially in DiskImg, could probably be simplified.
</ul>
<p>The library depends on NufxLib and zlib for access to compressed images.</p>
<h3>Reformat Library</h3>
<p>This is probably the most "fun" component of CiderPress. It converts Apple II files
to more easily accessible Windows equivalents.
<p>Start in Reformat.h and ReformatBase.h. There are two basic kinds of
reformatter: text and graphics. Everything else is a sub-class of one of the
two basic types.
<p>The general idea is to allow the reformatter to decide whether or not it is
capable of reformatting a file. To this end, the file type information and
file contents are presented to the &quot;examine&quot; function of each reformatter in
turn. The level of confidence is specified in a range. If it's better than
"no", it is presented to the user as an option, ordered by the strength of its
convictions. If chosen, the &quot;process&quot; function is called to convert the data.
<p>Bear in mind that reformatters may be disabled from the preferences menu.&nbsp;
Also, when extracting files for easy access in Windows, the &quot;best&quot;
reformatter is employed by the extraction code.<p>Most of the code should be portable, though some of it uses the MFC CString class.&nbsp;
This could probably be altered to use STL strings or plain.
<h3>Util Library</h3>
<p>Miscellaneous utility functions.
<p>For a good time, look at SelectFilesDialog.cpp.
<p>To enable debug logging for one of the applications, define _DEBUG_LOG in
MyDebug.h in this library. You will see "_DEBUG_LOG" in the version string
in the About box when this is defined. The log is written to C:\cplog.txt.
An existing log file will be appended to if the previous log was written to
less than 8 hours ago.</p>
<hr>
<p>Enjoy!</p>
<address>Andy McFadden</address>
</body>
</html>