I like this because it makes easy to keep my content organized.

Here is how you do it:

1. create a directory suffixed .b3 under your content dir. Ex. content/test.b3/
2. put all related files in this .b3 dir.
3. write your post text, and place it in index.markdown
4. to link to the other files, prefix references to the other files with @site, ex: <img src="@site/photo.jpg" />

For example, this post contains the bucket3 logo:

see the post source here.

It's a problem I've faced many times: you need to save your old blog, or your old website, in a form that will be accessible i n the future, and won't need any investment (in money or time from your part).

Some may say that the solution is to move your content to an open source content management system. I don't think so. You will have to keep up with the updates, new database versions, incompatibilities of future versions of the CMS with your current hosting environment etc. After a couple of years, you will realize that in order to keep up with the newest versions of the CMS you will have to upgrade your apache server, and then you will find out that you have to upgrade your linux distribution...

IMHO, the best format is static HTML pages. If you have a web server, you will be able to host them. And even if you don't you will be able to read them locally!

This is one of the main reasons I expect a lot from web publishing systems that create static pages in the first place: this content will most probably survive longer.

Here is some code in python that seems to work and notify Dave Winer's cloud server.

import httplib, urllib
params = urllib.urlencode({'url': "http://bucket3.com/blog/feed/rss2.xml"})
headers = {"Content-type": "application/x-www-form-urlencoded", 
    "Accept": "text/plain"}
conn = httplib.HTTPConnection("rpc.rsscloud.org:5337")
conn.request("POST", "/rsscloud/ping", params, headers)
response = conn.getresponse()
print response.status, response.reason
data = response.read()
print data
conn.close()

This is important for two reasons:

1. It makes it easy to create and use other tools (my favorite is bash scripts, but yours may vary) to create this content. bucket3 isjust the publishing system.
2. It makes it easy to distribute this content, since we have many excelent tools for filesystem content distribution, ranging from rsync, to git, to dropbox, that cover different needs.

Feel free to see how this blog's original content is organized at http://bucket3.com/blog_source/.

And, yes, I call this the "source" of the blog. Once you compile it using bucket3, you get the rendered HTML pages or the RSS feed. (From this point of view, bucket3 is a compiler, while wordpress is an interpreter :-)

This is why I decided to start, "bucket3 blog": to eat my own dog food, and use it as a sandbox for a real blog.

This is very useful, in case you want a post containing images, like this

• b3_markdown handles .markdown files.
• b3_html handles .html files
• b3_image handles images. (It will create a post from a single image)

Enabling new handlers

Edit bucket3/blog.py and add your own, after

from bucket3.handlers.markdown import b3_markdown
from bucket3.handlers.html import b3_html
from bucket3.handlers.image import b3_image

post.regHandler(b3_markdown)
post.regHandler(b3_html)
post.regHandler(b3_image)