Entries tagged atom

Use ssh-agent for atom.io remote-sync under MacOS

Posted on 22. Januar 2017 Comments

If you use the IDE atom.io with some sort of remote sync/scp/sftp package you can’t just use ssh-keys out of the box, if you set them up with a passphrase (which you should!). Here’s the 3 options most plugins offer:

  1. Use password (unless you disabled the option on the server)
  2. Use ssh-key
  3. Use ssh-agent

The problem with #2 is, that you would have to save your passphrase in a clear-text file which is never a good idea. But you can use the ssh-agent on MacOS the following way:

  1. create a ~/.ssh/config file
  2. add this line: AddKeysToAgent yes
  3. $ ssh-add .ssh/id_rsa
  4. restart atom

Flattr this!

Using nanoc for podcast feeds

Posted on 23. Juli 2015 Comments

Nanoc is a static site generator much like jekyll or octopress, but with a more minimalistic approach. These generators are not necessarily the most suitable choice for a podcast website, but it’s possible and you might save up on webspace and traffic when you use GitHub or Neocities.

Creating the podcast feed is basically like writing a normal Atom feed for the blog, since podcast feeds ARE indeed feeds with an enclosure tag in which the URL to the audio or video file is placed. This guide does not include the itunes tags. I might add it one day. Follow the instructions and use the documentation for the Helper Blogging. Tag your podcast episodes as kind:article.

The media files are placed in content/mp3 and content/opus, which is where the links in the feeds will point to later.

I invented the fields mp3 and opus, since these are the file formats I want to use. The values are the filenames. The header of a new episode/post would look like

---
title: 001 - Podcast Episode Title
created_at: 2015-03-14 09:00:00 +0000
kind: article
tags: [podcast,topic]
mp3: 001-podcast-episode-title.mp3
opus: 001-podcast-episode-title.opus
---

This has to be filled manually everytime, so make sure you have the exact filename, as some podcast clients won’t allow correction of the URL.

The next step is to write different feeds for the formats. For that, I’m using the new field format, which will be interpreted by the Helper class later on. For example, I called my normal feed blogfeed and the podcast feeds mp3feed and opusfeed. Create the file blogfeed.erb in the content folder and fill it with the following:

<%= atom_feed :title => 'repats podcast blog', :author_name => 'repat',
:author_uri => 'http://repat.de', :limit => 10, :format => 'blog' %>

The mp3feed.erb and opusfeed.erb are filled accordingly:

<%= atom_feed :title => 'repats podcast mp3', :author_name => 'repat',
:author_uri => 'http://repat.de', :limit => 10, :format => 'mp3' %>

<%= atom_feed :title => 'repats podcast opus', :author_name => 'repat',
:author_uri => 'http://repat.de', :limit => 10, :format => 'opus' %>

The next step is to use the Blogging locally in your nanoc installation. To do that you need to copy it from the gems folder into your lib folder. For me, that was

$ cp /var/lib/gems/1.9.1/gems/nanoc-3.7.5/lib/nanoc/helpers/blogging.rb lib/

It should be included like this in the lib/default.rb

include Nanoc3::Helpers::Blogging

Add the following attribute to the AtomFeedBuilder class

attr_accessor :format

If you don’t trust yourself to always remember the files you might want to  add this exception to the validate_feed_item function

if format.nil?
raise Nanoc::Errors::GenericTrivial.new('Cannot build Atom feed: no format(mp3,opus,blog) in params, item or site config')
end

After the # Add link comment is a good place to insert the  enclosure tag. File.size() will only work if the files are there and the exact same name. This code could probably be written a bit more safely, but I’m not a ruby developer and since I will have an mp3 file and and opus file in every post it’s not a problem this way.

# Add podcast enclosure
if format == 'mp3'
xml.link(href:"http://yourpodcast.com/mp3/" + a[:mp3],length:File.size("content/mp3/" + a[:mp3]), type:"audio/mpeg", rel:"enclosure")
elsif format == 'opus'
xml.link(href:"http://yourpodcast.com/opus/" + a[:opus],length:File.size("content/opus/" + a[:opus]), type:"audio/mpeg", rel:"enclosure")
end

To interpret the mp3 and opus attribute from earlier in the actual post, the last step is to add this line to the atom_feed function:

      builder.format            = params[:format]

You might need to install builder to let this run

$ sudo gem install builder

The only thing left to do is to edit the Rules file:

compile '/blogfeed' do
filter :erb
end
compile '/mp3feed' do
filter :erb
end
compile '/opusfeed' do
filter :erb
end
[...]
route '/blogfeed' do
'/blogfeed.xml'
end
route '/mp3feed' do
'/mp3feed.xml'
end
route '/opusfeed' do
'/opusfeed.xml'
end

 

You can find the blogging.rb and the Rules file on GitHub.

Flattr this!