summaryrefslogtreecommitdiffstats
path: root/submitting-patches
diff options
context:
space:
mode:
Diffstat (limited to 'submitting-patches')
-rw-r--r--submitting-patches44
1 files changed, 44 insertions, 0 deletions
diff --git a/submitting-patches b/submitting-patches
new file mode 100644
index 00000000..78974458
--- /dev/null
+++ b/submitting-patches
@@ -0,0 +1,44 @@
+HOWTO: Submit a patch
+---------------------
+
+This document is here mainly to make my job easier, and is more of a guideline,
+and not a strict set of rules. Please try to follow as much as you can.
+
+NOTE: Much of this is paraphrased from the kernel documentation's
+"SubmittingPatches" file.
+
+Creating your patch:
+
+* Use "diff -up" or "diff -uprN" to create patches.
+
+ These options make the diff easier to read for those of us who try to review
+ submitted patches.
+
+* Please try to make patches "p1 applicable"
+
+ This means that if you are patching file "lib/libalpm/alpm.h", I should be
+ able to apply your patch while passing the -p1 argument to 'patch'. The diff
+ header should look like so:
+
+ --- ORIGINAL_DIR/lib/libalpm/alpm.h
+ +++ NEW_DIR/lib/libalpm/alpm.h
+
+ With '-p1' the ORIGINAL_DIR and NEW_DIR arguments are stripped.
+
+Submitting your patch:
+
+* No MIME, no links, no compression, no attachments. Just plain text.
+
+ Patches should be contained in the actual body of the email. There are many
+ reasons for this. Firstly, it makes them easier to read with any mail reader,
+ it allows easier review "at a glance", and most importantly, it allows people
+ to comment on exact lines of the patch in reply emails.
+
+ It is important to know that the diff format ignores plain text before (and
+ after) the main diff itself. If you directly insert your patch into an email,
+ you can safely add lines above it describing your patch.
+
+* Describe your patch.
+
+ TODO...
+