summary refs log blame commit diff stats
path: root/doc/colorschemes.txt
blob: 905c7a3ecea2742cc76b33d121f68ff81eb3da06 (plain) (tree)
rner cases. The premise is</span> <span class="Comment">//: that understandability and rewrite-friendliness are related in a virtuous</span> <span class="Comment">//: cycle. Doing one well makes it easier to do the other.</span> <span class="Comment">//:</span> <span class="Comment">//: Lower down, this file contains a legal, bare-bones C++ program. It doesn't</span> <span class="Comment">//: do anything yet; subsequent files will contain :(...) directives to insert</span> <span class="Comment">//: lines into it. For example:</span> <span class="Comment">//: :(after &quot;more events&quot;)</span> <span class="Comment">//: This directive means: insert the following lines after a line in the</span> <span class="Comment">//: program containing the words &quot;more events&quot;.</span> <span class="Comment">//:</span> <span class="Comment">//: A simple tool is included to 'tangle' all the files together in sequence</span> <span class="Comment">//: according to their directives into a single source file containing all the</span> <span class="Comment">//: code for the project, and then feed the source file to the compiler.</span> <span class="Comment">//: (It'll drop these comments starting with a '//:' prefix that only make</span> <span class="Comment">//: sense before tangling.)</span> <span class="Comment">//:</span> <span class="Comment">//: Directives free up the programmer to order code for others to read rather</span> <span class="Comment">//: than as forced by the computer or compiler. Each individual feature can be</span> <span class="Comment">//: organized in a self-contained 'layer' that adds code to many different data</span> <span class="Comment">//: structures and functions all over the program. The right decomposition into</span> <span class="Comment">//: layers will let each layer make sense in isolation.</span> <span class="Comment">//:</span> <span class="Comment">//: &quot;If I look at any small part of it, I can see what is going on -- I don't</span> <span class="Comment">//: need to refer to other parts to understand what something is doing.</span> <span class="Comment">//:</span> <span class="Comment">//: If I look at any large part in overview, I can see what is going on -- I</span> <span class="Comment">//: don't need to know all the details to get it.</span> <span class="Comment">//:</span> <span class="Comment">//: Every level of detail is as locally coherent and as well thought-out as</span> <span class="Comment">//: any other level.&quot;</span> <span class="Comment">//:</span> <span class="Comment">//: -- Richard Gabriel, &quot;The Quality Without A Name&quot;</span> <span class="Comment">//: (<a href="http://dreamsongs.com/Files/PatternsOfSoftware.pdf">http://dreamsongs.com/Files/PatternsOfSoftware.pdf</a>, page 42)</span> <span class="Comment">//:</span> <span class="Comment">//: Directives are powerful; they permit inserting or modifying any point in</span> <span class="Comment">//: the program. Using them tastefully requires mapping out specific lines as</span> <span class="Comment">//: waypoints for future layers to hook into. Often such waypoints will be in</span> <span class="Comment">//: comments, capitalized to hint that other layers rely on their presence.</span> <span class="Comment">//:</span> <span class="Comment">//: A single waypoint might have many different code fragments hooking into</span> <span class="Comment">//: it from all over the codebase. Use 'before' directives to insert</span> <span class="Comment">//: code at a location in order, top to bottom, and 'after' directives to</span> <span class="Comment">//: insert code in reverse order. By convention waypoints intended for insertion</span> <span class="Comment">//: before begin with 'End'. Notice below how the layers line up above the &quot;End</span> <span class="Comment">//: Foo&quot; waypoint.</span> <span class="Comment">//:</span> <span class="Comment">//: File 001 File 002 File 003</span> <span class="Comment">//: ============ =================== ===================</span> <span class="Comment">//: // Foo</span> <span class="Comment">//: ------------</span> <span class="Comment">//: &lt;---- :(before &quot;End Foo&quot;)</span> <span class="Comment">//: ....</span> <span class="Comment">//: ...</span> <span class="Comment">//: ------------</span> <span class="Comment">//: &lt;---------------------------- :(before &quot;End Foo&quot;)</span> <span class="Comment">//: ....</span> <span class="Comment">//: ...</span> <span class="Comment">//: // End Foo</span> <span class="Comment">//: ============</span> <span class="Comment">//:</span> <span class="Comment">//: Here's part of a layer in color: <a href="http://i.imgur.com/0eONnyX.png">http://i.imgur.com/0eONnyX.png</a>. Directives</span> <span class="Comment">//: are shaded dark.</span> <span class="Comment">//:</span> <span class="Comment">//: Layers do more than just shuffle code around. In a well-organized codebase</span> <span class="Comment">//: it should be possible to stop loading after any file/layer, build and run</span> <span class="Comment">//: the program, and pass all tests for loaded features. (Relevant is</span> <span class="Comment">//: <a href="http://youtube.com/watch?v=c8N72t7aScY">http://youtube.com/watch?v=c8N72t7aScY</a>, a scene from &quot;2001: A Space</span> <span class="Comment">//: Odyssey&quot;.) Get into the habit of running the included script called</span> <span class="Comment">//: 'test_layers' before you commit any changes.</span> <span class="Comment">//:</span> <span class="Comment">//: This 'subsetting guarantee' ensures that this directory contains a</span> <span class="Comment">//: cleaned-up narrative of the evolution of this codebase. Organizing</span> <span class="Comment">//: autobiographically allows a newcomer to rapidly orient himself, reading the</span> <span class="Comment">//: first few files to understand a simple gestalt of a program's core purpose</span> <span class="Comment">//: and features, and later gradually working his way through other features as</span> <span class="Comment">//: the need arises.</span> <span class="Comment">//:</span> <span class="Comment">//: Programmers shouldn't need to understand everything about a program to hack</span> <span class="Comment">//: on it. But they shouldn't be prevented from a thorough understanding of</span> <span class="Comment">//: each aspect either. The goal of layers is to reward curiosity.</span> <span class="Comment">// Includes</span> <span class="Comment">// End Includes</span> <span class="Comment">// Types</span> <span class="Comment">// End Types</span> <span class="Comment">// Prototypes are auto-generated in the makefile; define your functions in any</span> <span class="Comment">// order. Just be sure to declare each function header all on one line. Our</span> <span class="Comment">// auto-generation scripts are too minimal and simple-minded to handle</span> <span class="Comment">// anything else.</span> <span class="PreProc">#include </span><span class="Constant">&quot;function_list&quot;</span> <span class="Comment">// by convention, files ending with '_list' are auto-generated</span> <span class="Comment">// Globals</span> <span class="Comment">//</span> <span class="Comment">// All statements in this section should always define a single variable on a</span> <span class="Comment">// single line. The makefile will simple-mindedly auto-generate extern</span> <span class="Comment">// declarations for them. Don't forget to define (not just declare) constants</span> <span class="Comment">// with extern linkage in this section, since C++ global constants have</span> <span class="Comment">// internal linkage by default.</span> <span class="Comment">//</span> <span class="Comment">// End Globals</span> <span class="Normal">int</span> main<span class="Delimiter">(</span><span class="Normal">int</span> argc<span class="Delimiter">,</span> <span class="Normal">char</span>* argv[]<span class="Delimiter">)</span> <span class="Delimiter">{</span> atexit<span class="Delimiter">(</span>teardown<span class="Delimiter">);</span> <span class="Comment">// End One-time Setup</span> <span class="Comment">// Commandline Parsing</span> <span class="Comment">// End Commandline Parsing</span> <span class="Identifier">return</span> <span class="Constant">0</span><span class="Delimiter">;</span> <span class="Comment">// End Main</span> <span class="Delimiter">}</span> <span class="Comment">// Unit Tests</span> <span class="Comment">// End Unit Tests</span> <span class="Comment">//: our first directive; will move the include above the program</span> <span class="Delimiter">:(before &quot;End Includes&quot;)</span> <span class="PreProc">#include </span><span class="Constant">&lt;stdlib.h&gt;</span> <span class="Comment">//: Without directives or with the :(code) directive, lines get added at the</span> <span class="Comment">//: end.</span> <span class="Delimiter">:(code)</span> <span class="Normal">void</span> setup<span class="Delimiter">()</span> <span class="Delimiter">{</span> <span class="Comment">// End Setup</span> <span class="Delimiter">}</span> <span class="Normal">void</span> teardown<span class="Delimiter">()</span> <span class="Delimiter">{</span> <span class="Comment">// End Teardown</span> <span class="Delimiter">}</span> </pre> </body> </html> <!-- vim: set foldmethod=manual : -->