Elliott C. Back: Internet & Technology

I was talking to a friend the other at a dinner party about online mapping services, and she said that paper maps will never die. The dinner party people gave a number of GPS-took-me-along-a-bad-route examples and so I thought about it. Why are online maps bad at all? Aside from shortest-path bugs, isn’t the best way to get from A to B the shortest way?

No, it’s not. There are a few problems:

1) Not everyone prefers the same kinds of routes. Some people just like highways better than surface streets, even if the surface streets could shave 15% of their trip time. Therefore, mapping services need to associate user profile and preference data to use in calculating routes, to learn where people like to drive and deliver them routes that are straight or curvy, bumpy or smooth, consistent or sketchy depending on what they like.

2) Some kinds of terrain are unmitigable. You really don’t want to take that dirt road into the mountains when you could be on the regular freeway. You don’t want to have to ford rivers in the bush when there’s an access road, regardless of other factors. So, the type of terrain has to be known, and weights assigned based on its desirability.

3) Temporal problems. Things like police, traffic, an accident, or some kind of disaster, those events which affect the transportation system at random times, should play a major role in route selection. Currently they don’t.

4) Surroundings. Do you want to drive your car through a slum, or through the nice part of town? Do you want to drive along a scenic route or through a landfill? The surroundings a car takes should be a factor in choice of route.

So, Yahoo, Google, Mapquest, MSN, and co–the first of you to implement all of these wins.

As far as RSS goes, Feedburner is the hottest feed management technology. If you thought RSS was powerful before, it’s ten times as useful when powered by their propietary tracking and enhancement technologies. In this article, I am to show you how to convert your Wordpress feeds to work in Feedburner, and what value Feedburner adds to your site syndication.

Setting it up:

Sign up for a Feedburner account. Then, when it prompts you for the RSS url, give it the Wordpress absolute path:

This should be the wp-rss2.php file in the root of your WP installation directory. Why do we do this? Because the next step is to forward our RSS readers from the blog to the URL which Feedburner gives us with Apache mod_rewrite. When we do this, if told Feedburner to poll mysite.com/feed/ and mysite.com/feed/ to redirect to Feedburner, we’d have a recursive loop!

Now we just accept the title and url if they look ok, and activate our account:

Now you’ll see that they’ve given us the URL feeds.feedburner.com/VioxxNews . We want to set up our site to forward here, so go to the .htaccess file in the root of your Wordpress installation, and look for two lines that look like:

RewriteRule ^feed/(feed|rdf|rss|rss2|atom)/?$ /index.php?&feed=$1 [QSA,L]
RewriteRule ^(feed|rdf|rss|rss2|atom)/?$ /index.php?&feed=$1 [QSA,L]

These two rules take requests for your feed, and send them to the real file. However, we’re using Feedburner now, so we need to change these to:

RewriteRule ^feed/(feed|rdf|rss|rss2|atom)/?$ http://feeds.feedburner.com/VioxxNews [QSA,L]
RewriteRule ^(feed|rdf|rss|rss2|atom)/?$ http://feeds.feedburner.com/VioxxNews [QSA,L]

Now all requests for old feeds will be shunted through Feedburner! You should also add this ABOVE the # BEGIN WordPress block so that Wordpress doesn’t overwrite your rules with its own. Since yours are higher up, they’ll take precedence.

Adding value:

Now that we have Feedburner handling our rss/atom/rdf/xml feeds, we can leverage the best of its services to improve the quality of our syndication. Click on “Skip directly to feed management” to start! The first thing you want to enable is SmartFeed, under the Optimize tab:

Now your feed will automatically detect the type of agent visiting it and adjust its format between RSS and ATOM depending on its preferences and needs. This means that with one feed you can leverage a multi-format audience. Now there’s no distinction between RSS and ATOM readers: you just have a single XML feed.

Now go to the Publish tab, where we can add Pingshot–automatic pinging for your posts:

Everytime your RSS changes, you’ll send out pings. While this might seem a little spammy to me, it’s also a great idea to make sure ping services are getting notified about your content. The other cool Publicity tool to activate is the Feedcount widget:

Which will give you a publicity button like:

Discovering your feeds:

Feedburner also provides a number of tools to pry deep into the guts of your feeds, which fall under the “Analyse” tab of their administrative interface. Feed circulation gives the number of readers polling your feeds:

If you want more useful information, you should check out the Readership details:

Here you’ll see how many of your RSS hits are from real readers, search engines like IceRocket and Technorati, from spammers and scrapers, or other sources. This, in my opinion, is FeedBurner’s most powerful feature–the ability to satisfy curiousity about Feed membership. It’s easy to write a script to look for requests per day to /feed/, but that’s a lot more effort than just setting up FeedBurner and letting it do it’s thing.

Next steps:

If you’re a podcaster, FeedBurner will automatically do enclosures and make sure all the content types are correct. If you’re a big linker or big on Flickr, FeedBurner can splice your del.ic.ious feed into your blog feed, or splice in your photos. You can geotag, automatically add ticker symbols, create summary text, add a creative commons license, or adsense advertisments. If you can think of something to do to your feed, chances are Feedburner will help you get there.

Conclusion:

When I started using Feedburner, it was just for the tracking capabilities it has. But, I found a number of common useful options I’ve outlined about that add substantial value to your feed serving. In that regard, Feedburner adds more than just statistics and tracking. And, there are no competitors in its space. While every thing it does is relatively simple, it does a lot of them providing end-to-end RSS/ATOM/RDF/XML solutions for bloggers.

I don’t usually write about my life, but today I have a lot of things to do:

*CENSORED*

3) Chinese HW

Done. I will do this while in CS430, and the hour afterwards when I eat lunch. I have one thing to make up from yesterday, but oh well. I will still get this done. Interestingly enough, this is probably the only class I will bring with me over break because it’s so hard to memorize the hanzi (??)…

4) Get NYC tix, laundry, pack, determine when I am going… ;_;

DONE. I am going to Chinese class, then heading on out of here.

5) Start music project
6) Start 430 project

These two, sadly, haven’t happened, although I asked my prof just what the 430 project entails…

7) Make sure security project is good to go

DONE. Ah yes–need to sign the certificates. Open SSL is more complicated than you might think. First I used genrsa to create a public/private key pair, then I used req to generate a self-signed CA certificate with the -x509 flag, which is important because it creates a certificate instead of a certificate request! So, here are the commands in entirety:

touch index.txt
echo “01″ > serial.txt
openssl genrsa 2048 > ca.key
openssl req -config openssl.cfg -new -x509 -key ca.key -out ca.cert
openssl ca -ss_cert andrewPemCert -cert ca.cert -keyfile ca.key -config openssl.cfg -out andrewSignedPemCert
openssl x509 -inform PEM -outform DER -in andrewSignedPemCert -out andrewSignedDerCert

Sun Parser, mtg

DONE. Started writing the parse, but leaving regexes for a little later. The mtg was ok.

9) Blog feedburner…

DONE. this will be my next post.

10) Buy some things from the store

Heh. When I come back from t-break?

11) ???

Heh–this will be when I sleep.