aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorJohn Ankarström <john@ankarstrom.se>2020-10-21 23:02:01 +0200
committerJohn Ankarström <john@ankarstrom.se>2020-10-21 23:08:30 +0200
commit813559c0d0be87826dcf239a905e74c47529d6b4 (patch)
tree9081ae567cc5419d04d3909c3d96a41fe6ca9b13
downloadgit-save-readme.tar.gz
first commitreadme
-rw-r--r--index.html56
1 files changed, 56 insertions, 0 deletions
diff --git a/index.html b/index.html
new file mode 100644
index 0000000..8627146
--- /dev/null
+++ b/index.html
@@ -0,0 +1,56 @@
+
+<h1>git-save</h1>
+<p><big>git-save is a Git command designed for <em>commitless version
+control</em>. The instruction <code>git save</code> is equivalent to <code>git
+add . &amp;&amp; git commit --allow-empty-message -m ""</code>.</big></p>
+<h2>What is commitless version control?</h2>
+<p>Commitless version control is a software development method with two
+rules:</p>
+<ol>
+ <li> Make <i>saves</i> instead of commits. A save is like a commit,
+except
+ <ul>
+ <li>it has no commit message, and</li>
+ <li>it includes <em>all changes</em> in <em>all files</em>.</li>
+ </ul>
+ </li>
+ <li>Don't worry about keeping a clean, thematically organized commit
+history. A save should contain all changes made since the last save,
+regardless of theme.</li>
+</ol>
+<p>In summary, <h>eschew order for chaos</h>.</p>
+<h2>What is git-save?</h2>
+<p>git-save is a tool that makes it easy to use Git as a commitless
+version control system. It has a simple interface:</p>
+<pre>$ git save</pre>
+<p>In combination with a file watcher, saves can be performed
+automatically whenever you write to the files in your project:</p>
+<pre>$ watch example.doc | while read; do git save; done</pre>
+<h2>Benefits</h2>
+<p>For the single developer, commitless version control presents
+several benefits:</p>
+<ul>
+ <li><b>Lower commit hurdle.</b> Using git-save is as easy as saving a
+Microsoft Word document.</li>
+ <li><b>More comprehensive history.</b> Because saves don't need to be
+complete, isolated or even well-written, they can be made much more
+often. This results in a much more comprehensive history, where changes
+that normally wouldn't make it into a commit are included as well.</li>
+ <li><b>Lower risk of losing work.</b> Because saves are made more
+often than commits, and saves automatically capture all changes, not
+just a carefully selected few, potentially important changes rarely
+remain uncommitted, hanging in the air and easily lost.</li>
+</ul>
+<p>Additionally, if commitless version control were to become popular,
+it might have a few more far-reaching positive effects:</p>
+<ul>
+ <li><b>Improved tooling.</b> Lacking commit messages, developers
+would be forced to rely more on information about which files were
+changed and in which ways -- which version control tools would be forced
+to convey better. This information is ultimately more precise than a
+commit message.</li>
+ <li><b>Beginner-friendliness.</b> Without the requirement to keep a
+clean commit history, and the accompanying mental burden, it would be
+much easier for beginners to get into the habit of putting their
+projects under version control.</li>
+</ul>