From 37498f1ce86049ba95473fc417e80d05d45d4201 Mon Sep 17 00:00:00 2001
From: Mike Lynch <m.lynch@sydney.edu.au>
Date: Sun, 10 Dec 2023 10:03:34 +1100
Subject: [PATCH] Put more details in the README

---
 README.md | 42 ++++++++++++++++++++++++++++++++++++++----
 1 file changed, 38 insertions(+), 4 deletions(-)

diff --git a/README.md b/README.md
index e85b875..084880e 100644
--- a/README.md
+++ b/README.md
@@ -1,9 +1,43 @@
 # autoradio
 
-Auto download of mp3s from podcast feeds using a little Python script.
+A little Python script to download audio files from RSS feeds. Installation
+is messy and I have tested this on exactly two feeds, so it's my no means
+finished software.
 
 ## Installation
 
-I haven't worked out how to intall pycurl with poetry, as it looks like it
-required a command-line flag to specify the SSL backend, and poetry doesn't
-seem to support that. So for my local install I pip installed it by hand
\ No newline at end of file
+I packaged this project with poetry and then discovered that poetry can't
+install pycurl properly, as the latter requires command-line options to compile
+it with the correct SSL backend. Here's the manual workaround:
+
+    > git clone https://git.tilde.town/bombinans/autoradio.git
+    > cd autoradio
+    > poetry install
+    > poetry shell
+    > pip uninstall pycurl
+    > pip install --compile --install-option="--with-openssl" pycurl
+    > poetry run autoradio --config config.json
+
+The config file looks like this.
+
+    {
+        "feeds": {
+            "Utility Fog": {
+                "url": "https://www.frogworth.com/utilityfog/feed/",
+                "dir": "/media/pi/Storage/Music/Utility Fog"
+            },
+            "RA Podcast": {
+                "url": "https://ra.co/xml/podcast.xml",
+                "dir": "/media/pi/Storage/Music/RA_Podcast"
+            }
+        },
+        "max": 10
+    }
+
+Each feed has a URL and a directory where the files will be downloaded. If a
+file with the same name is already present, it won't be re-downloaded.
+
+The value in "max" is the number of entries to look at from each feed, starting
+from the latest. The script parses the RSS feed and looks for links which 
+seem likely to contain audio and tries to fetch them using pycurl (this handles
+things like redirects, hopefully).
\ No newline at end of file