Nounoublog updated
February 06, 2010
Over the last few weeks this blog has changed dramatically. It looks pretty much the same as when it started but under the covers the code of the blogging platform, Nounoublog is very different. I am going to talk about three of the features that I have been working on lately:
- Archives
- RSS Feed
- Admin console
And I will show a few snippets of the actual code that powers the blog.
For those that visit the blog for the first time, Nououblog is a small blogging platform developed in Google App Engine. I started working on it basically for two reasons. First, I wanted to learn how to develop applications for Google App Engine. And second, because I wanted a simple but highly customizable platform, with free hosting and no ads.
Archives
Now the blog has an archives section. It is the small set of links in the right side. It will let you view all the posts since the creation of the blog.
For example, if you click in 2009, it will show you all the posts from the previous year. In order to enable this I had to update the url structure of the blog. Now paths have the form:
http://www.javiertordable.com/blog/2010/01/30/the-eternal-night
With slashes separating the different parts of the url. And all the following are valid urls:
http://www.javiertordable.com/blog/2010/01/30/ http://www.javiertordable.com/blog/2010/01/ http://www.javiertordable.com/blog/2010/
Each one will show respectively all the posts of the day, the month, and the year. Each one has its own handler. Here is for example the handler that returns all the posts in a year:
class BlogYear(webapp.RequestHandler):
"""Request handler for all blog posts in a given year.
This handler answers all requests for /blog/YYYY and /blog/YYYY/.
"""
def get(self):
# Get the year from the path.
path = self.request.path[len('/' + config.BLOG_PREFIX + '/'):]
(year, month, day, desired_url) = extract_url_parts(path)
# Get the list of all posts of the year.
posts = get_posts_in_date(year, month=None, day=None)
# And return an archives page with the posts.
s = pages.ArchivesPageGenerator()
self.response.out.write(s.generate(posts, str(year)))
There are similar handlers for all the posts in a month and all the posts in a day.
RSS Feed
An RSS feed is a specially formatted XML file, which includes data about the posts in a blog. It is updated automatically by the blog, so that when it changes, RSS subscribers know that there is new content. There are applications like Google Reader that are very helpful to keep track of many RSS feeds and alerting when there is new stuff to read.
In the Nounoublog blogging platform, you can access the RSS feed by clicking in the RSS link at the top of the page. In the browser you may see only a bunch of text, but if you add this link to your Google Reader subscriptions you will see a list of the most recent posts.
Same as before the RSS feed is powered by its own handler, which is very simple:
class RssFeed(webapp.RequestHandler):
"""Handler for the RSS feed.
This feed contains a list with all the blog posts, from last to first.
This list is subject to the maximum item retrieval limit of the DB.
"""
def get(self):
# Get the list of all posts.
posts = get_all_posts()
# Return the xml feed with the posts.
template_values = {'posts': posts}
self.response.out.write(template.render(template_path("rss_feed"),
template_values))
Where the template provides the XML structure of the feed, and inserts the data corresponding to the posts
<?xml version="1.0" encoding="iso-8859-1"?>
<rss version="2.0">
<channel>
<title>Javier Tordable Blog</title>
<link>http://www.javiertordable.com</link>
<description>
Javier Tordable blog on Software, Mathematics and Technology
</description>
<generator>Nounoublog</generator>
<docs>https://code.google.com/p/nounoublog/</docs>
{# Loop over all the blog posts. #}
{% for post in posts %}
<item>
<title>{{ post.title }}</title>
<link>{{ post.absolute_url }}</link>
<pubDate>{{ post.rss_pub_date }}</pubDate>
<guid>{{ post.absolute_url }}</guid>
<description>{{ post.escaped_content }}</description>
</item>
{% endfor %}
</channel>
</rss>
Notice that in the Django template the post elements appear as attributes while in fact they are method calls. Also I use as GUID the url of the post, as it is intended to be a permanent link.
Admin console
The last item that I have been working on is the administration console. This is still work in progress, but I expect that once I am done with it I will post more often.
The admin console will have options to:
- Add posts
- Edit posts
- Add static pages
- Edit static pages
- Edit the CSS
- Edit redirects
All these options seem very normal with the exception of the redirects. How does it work? For example, when going to:
http://www.javiertordable.com/blog/2009-12-01/my-first-blog-post
You are redirected to another url, which appears in the url bar. Notice how the dashes are now forward slash bars
http://www.javiertordable.com/blog/2009/12/01/my-first-blog-post
I added support for redirects because I changed the site several times (including the url structure), and I didn't want to serve 404 error pages for all old urls.
Keep visiting the blog or subscribe to the RSS feed for more news on Nouonublog!
The Eternal Night
January 30, 2010
My brother David is a film director. He has been making short film for a few years, he has even won a few prizes. Most of his work is at tpmpictures.com. Today I just wanted to show his last piece of work, a science fiction short about the end of the world. The short is in Spanish but with English subtitles.
There are a couple of things that are interesting in this short. First, the script is not wildly improbable. Check out this (humorous) list of the 5 scientific experiments most likely to end the world. And second, the special effects are pretty nice for a zero budget short. If you liked it, please go ahead and leave a comment at the Notodo film festival.
TRASCEND
January 15, 2010
A couple of days ago I started reading TRASCEND, the new book from Ray Kurzweil and Terry Grossman. The book starts from the principle that our knowledge of medicine and biology is increasing to a point where we can start to control effectively how fast our own bodies age. And even more important the amount of knowledge that we gather is increasing over time. If the trend continues we may reach a point where we can effectively reverse engineer our bodies in order to avoid aging.
Whether you believe that we will reach that point or not, the book is an interesting read. It is filled with healthy habits, complete diets and recipes, exercise programs and more. Even if you leave aside the supplements and the fancy biomedical technologies, there are plenty of actionable tips for inproving your quality of life. For example to check for food intolerances. There are millions of people out there that can't digest milk or wheat very well and are not even aware of it.
Disclaimer: I didn't get paid to write this post.
MapReduce Integer Factorization in arXiv
January 07, 2010
This Monday I published my article on MapReduce for integer factorization in arXiv. The article is essentially the same that can be downloaded in the research section of this site. So if you have already checked it out, you won't find anything new. However I am very excited because it is my first addition to arXiv.
In case that you are not familiar with arXiv, it is one if the greatest scientific sites in the web. It has over half a million articles, especially in the fields of mathematics, physics and computer science, and many relevant papers are published in the arXiv months before they appear in any peer reviewed journal.
MapReduce Integer Factorization released!
December 29, 2009
Recently I published the code of MapReduce for Integer Factorization. It is available under the Apache 2.0 License in Google Code. It includes everything necessary to run in Apache Hadoop, as well as the numerical libraries used. It has no dependencies apart from the last version of Hadoop.
This project is a proof of concept that shows how to use MapReduce, a framework for distributed computation to solve a purely numerical problem. The main conclussion is that it's possible to use MapReduce for problems that lie far ahead from its original area of application, for example number theory. Also in this case the difficulty involved in developing the MapReduce program is similar to the difficulty of creating a worksheet in a mathematical tool like Maple. But the performance of MapReduce is significantly higher.
If you have some time, please download it from here, and let me know how it works for you.