chaospott-site/_site/tmp/var/www/chaospott.de
2014-10-26 19:35:05 +01:00
..
$basepath/www added a few articles, added media 2014-10-26 19:35:05 +01:00
api-www added a few articles, added media 2014-10-26 19:35:05 +01:00
cache added a few articles, added media 2014-10-26 19:35:05 +01:00
drafts added a few articles, added media 2014-10-26 19:35:05 +01:00
engine added a few articles, added media 2014-10-26 19:35:05 +01:00
example-hooks added a few articles, added media 2014-10-26 19:35:05 +01:00
example-templates added a few articles, added media 2014-10-26 19:35:05 +01:00
media added a few articles, added media 2014-10-26 19:35:05 +01:00
posts added a few articles, added media 2014-10-26 19:35:05 +01:00
templates added a few articles, added media 2014-10-26 19:35:05 +01:00
www added a few articles, added media 2014-10-26 19:35:05 +01:00
config.php added a few articles, added media 2014-10-26 19:35:05 +01:00
config.php.example added a few articles, added media 2014-10-26 19:35:05 +01:00
LICENSE added a few articles, added media 2014-10-26 19:35:05 +01:00
README.markdown added a few articles, added media 2014-10-26 19:35:05 +01:00
update.sh added a few articles, added media 2014-10-26 19:35:05 +01:00

Second Crack is a basic static-file blog engine using Markdown-formatted text files as input.

Warning

Second Crack should be considered an early alpha. Even though I've run Marco.org with it for a long time, it's still rough and unfriendly.

There are also still significant known bugs that I haven't had time to fix yet. (Feel free!) Among them:

  • The month archives sometimes only contain a few posts.
  • If you have a day with multiple posts, and you delete one of them in the middle, sometimes the others don't reorder themselves properly.
  • Editing slugs after publishing is unreliable.

These can usually be fixed temporarily by clearing the cache/ directory's contents and running the updater for a full rebuild.

Other minor bugs:

  • If you have multiple posts on the frontpage or an archive/tag page that use footnotes, the anchors will be repeated between posts and they won't work properly. Footnote anchors need some kind of unique per-post prefix.
  • If a template is edited, the changes won't propagate to the entire site automatically -- only new posts and updated pages will show the changes. To propagate a template's changes across the whole site, do a full rebuild by clearing the cache/ directory and running the updater.

Installation

Second Crack requires PHP 5.3+ and has only been tested so far on Mac OS X and CentOS 5.5, with Apache httpd. It's already fast with Apache, but it can be even faster and capable of even more burst traffic if you put a caching proxy such as Varnish in front of it.

Check this directory out on the server (or locally on Mac or Linux for testing, if you'd like).

Add a line like this to your crontab, replacing {SOURCE_PATH} (your blog source directory, see Basics below) and {SECONDCRACK_PATH} (this checkout) accordingly:

* * * * * /home/marco/secondcrack/engine/update.sh {SOURCE_PATH} {SECONDCRACK_PATH}

The updater script will append its output to /tmp/secondcrack-update.log.

I strongly suggest installing inotify-tools to get inotifywait. With it, the Second Crack updater will respond to file changes immediately as they happen. Without it, it only updates as often as the cron task runs. The round-trip Dropbox updating is a lot faster and more awesome with inotifywait.

Your webserver should use {SECONDCRACK_PATH}/www as its document root. It must support .htaccess files and mod_rewrite. Second Crack will install its engine/default.htaccess file into the web document root as .htaccess if an .htaccess file doesn't already exist.

Basics

Each .md blog-post file is structured like this:

This is the post title
======================
Tags: tag1, tag2
Published: 2012-01-03 10:22:31pm
Type: link

This is the post body, in **Markdown**.

It will continue for the rest of the file. See, the top is a lot
like an HTTP response with headers, but with a title and a row of 
any number of equals signs ("===") as the first two headers.

You can add arbitrary fields, formatted like HTTP headers. You
can then customize the templates to respond to them.

The engine expects a folder structure like this: (this is your {SOURCE_PATH} mentioned above)

(blog-source folder, name it whatever you want)/
    drafts/
        draft-post-slug.md
        another-draft-post-slug.md
        ...
    media/
        (images and other files however you want to arrange them. I do it like this:)
        2010/
        2011/
        2012/
            01/
                post-slug.png
                another-day-another-post1.jpg
                another-day-another-post2.jpg
                ...
    pages/
        about.md
        contact-me.md
        ...
    posts/
        2010/
        2011/
        2012/
            01/
                20120102-01-post-slug.md
                20120103-01-another-day-another-post.md
                20120104-01-something-happened-today.md
                20120104-02-something-else-happened-and-its-still-today.md
    templates/
        (this can technically be anywhere but it's easiest to keep it here)
        main.php
        rss.php
    hooks/
        (optional, code that executes on every new post, see example-hooks)
        post_twitter.php

If you're going to use Dropbox to publish, put that top-level blog-source folder somewhere in your Dropbox folder.

Set up config.php

Copy {SECONDCRACK_PATH}/config.php.example to {SECONDCRACK_PATH}/config.php and edit it to reflect your paths and preferences.

The blog-source folder ({SOURCE_PATH}) should be set as Updater::$source_path. The path to templates/ should be assigned to Template::$template_dir.

Posting

Make a file in drafts/ named as post-slug-you-want-to-create.md like this:

My draft title
==============

Draft content.

(Don't set a Published: header. The engine will add that automatically later.)

When the updater runs next, the engine will create a sister file in a _previews/ directory (and a _publish-now/ directory, if it doesn't already exist), like this:

drafts/
    post-slug-you-want-to-create.md
    ...
    _previews/
        post-slug-you-want-to-create.html
        ...
    _publish-now/

Whenever you want to preview your post, just save the .md file, run the updater (or wait for it to run itself), and open the corresponding .html file in your browser. When I'm writing, I just keep it open and hit Refresh to see changes.

To publish a post, either move it into the _publish-now/ directory, or add a header that simply says "publish-now", like this:

My finished title
=================
Tags: whatever, optional
publish-now

Here's my finished content.

Once that file is saved and the updater runs, it will publish the post immediately, removing it from drafts/ (and removing its preview file) and moving its final copy into posts/ in the appropriate year/month subdirectory.

To edit a post, simply edit its file in posts/. The updater will notice the change next time it runs and will update the published post.

Posting with bookmarklets

Second Crack comes with a pair of convenience bookmarklets, "Draft Link" and "Draft Article", that create draft posts from the current page and any selected text.

To enable this, create a second web root on an alternate domain, e.g. admin.myblog.com, using the api-www/ folder as its document root. Then you can navigate to, e.g.:

http://admin.myblog.com/add-draft.php

...and install the bookmarklets from there.

Your username and password to use these are set in config.php under "Meta Weblog API params". (I wanted to make a Meta Weblog interface in here too, but haven't gotten to it.)

Dropbox sync (strongly recommended)

Here's a video demo of why you really want Dropbox sync and those bookmarklets.

Second Crack's updater examines the blog source directory for changes, then writes HTML files into the web directory as needed.

If you're writing on a computer, you probably have a server hosting the blog somewhere, and need some way to update the content on the server. The best way to do this, and the setup that Second Crack is designed to operate in, is to set up Dropbox in both places: the native client for your computer, and the Linux CLI client on the server.

NOTE: If you don't want your server to have access to your entire Dropbox folder, you can do a dual-account trick: create a second Dropbox account, set up that one on the server, and "share" the blog-source folder from your primary account to the second account on Dropbox.

With Dropbox set up, the publishing and previewing workflow is streamlined, and you never need to run the Second Crack updater manually:

  • Have the preview .html file open in your local browser.
  • Make changes to the draft .md file. Hit Save in your editor.
  • Wait a few seconds. The changes will round-trip to the server and the .html file will update.
  • Hit Refresh in the browser and see the changes you made, rendered right in your site's layout by the server-side engine.

Even when you publish, just write the "publish-now" header, hit Save, and close the file. Within seconds, it will disappear from the drafts/ folder and the post will be published.

It's pretty awesome.

And if you want to edit your blog on the iPhone or iPad, you can do it with any Dropbox-capable text editor.

FAQ

Don't a lot of these static-file blogging engines already exist?

Yes.

Have you tried [existing solution]?

No.

Isn't this reinventing the wheel?

Yes.

Don't you have other things you could be working on?

Yes. (Don't we all?)

This name isn't unique.

You're probably right. Neither the domain nor the Twitter username are available, and it's probably trademarked in an industry I've never heard of.

Really, it's a terrible idea to launch a major project with such an unavailable name. But this isn't a major project, and I don't intend for it to get widespread enough that those problems will ever matter.

I needed a name. This came to mind. It's a coffee-roasting term for the moment in the roast that the bean audibly pops for the second time, indicating development of the strongest flavors and the point that you should stop the roast because it's done.

You're not entirely correct on that definition. And I prefer my roast to be [x] seconds (before|after) second crack.

I know. For the purposes of this FAQ, it's not really relevant.

OK, back to the static-file blog engine. What have you done differently from [existing solution]?

A bunch of small things, probably. I don't know enough about the other solutions to really say.

Why doesn't it have [feature]?

Because I didn't think [feature] needed to be there. Some anticipated frequent values for [feature]:

Comments: Use Disqus or Facebook comments. Or just go without comments. Do you really need them?
Stats: Use Google Analytics or Mint. (Or both.)
Widgets and dynamic page content: Use Javascript.
Dynamic rendering for automatic mobile layouts, etc.: Use CSS.

Why should I use this instead of [existing solution]?

I don't know. You probably shouldn't.

Will this make you, me, or anyone any money?

I doubt it.

So why did you make this?

Because I'm a programmer, and this is what I do.

Some people jog away from their house every day, only to jog back. Others walk on a treadmill, expending energy to get nowhere. In both cases, it may appear to others that they've accomplished nothing, but they've chosen to do these seemingly redundant activities on a regular basis to incrementally improve themselves. And it works.

That's not a perfect analogy. Programming another version of something with lots of existing solutions is nothing like daily cardiovascular exercise.

I know.