Push vs Pull

Email works on a ‘push’ model: you write your message, then push (send) it out to some list. When you write a new blog entry and want others to know about it, you may think to ‘push’ that to them by sending an email: “Hey, check out my new blog entry…”

The ‘push’ model works well in many cases, but it has one major problem: how to manage the list of poeple to send to? What if someone new wants to read your blog? What if 50 new people want to read your blog? Or 20 people decide your daily adventures really aren’t that interesting? That’s a lot of work for you to manage all those changes.

The ‘pull’ model allows the individual reader to choose which blogs or other sites he wants to view. When you think about it, that’s the way the whole internet works, actually — I decide when to go read my mom’s blog, or my sister’s, or my brother’s, or my cousin’s, or my other cousin, oh, and then there’s my aunt, and my brother-in-law, and the daily Dilbert cartoon, and the personal finance blogs, and the various geek sites… man, that’s a lot of work for me to visit all those sites to see what’s new.

Feed Me

Fortunately, some very smart (and lazy) people invented ‘feeds’ to help me and you — both the person publishing a blog, and the person trying to stay on top of 50 different blogs every day. A feed is just a feature built into a website (such as a blog) that helps to automate the ‘pull’ model.

Let’s look at an example. This blog’s address is http://dadeo.wordpress.com/. When you point your browser there, the wordpress.com server sends back the content of the blog, formatted for your reading pleasure. This blog has another address as well: http://dadeo.wordpress.com/feed/. If you were to point your browser there you’d get something different — an index of the articles available in the blog, with some information about each one, such as the date it was published and a snippet of the article.

That second link is the feed for this blog, and it automatically updates anytime I publish something new on my blog. That feed feature is built in to pretty much every blog you might run into across the Internet, and many other sites use them as well — such as online newspapers or news organizations.

The real trick happens when you pair a feed link such as that with a piece of software that knows how to do something with it. That would be a program called a feed reader, and all it does it continuously monitor the feeds you tell it you’re interested in, looking for new content. When it sees that the feed is updated with a new article, it will typically download it and notify you there’s something new to read.

There are a few different names and flavors of feeds, but they all do pretty much the same thing. The names you are likely to see are RSS and Atom. On a website, a feed is often indentified with an orange button or one of these symbols:

Okay, now you know what a feed is, and you want one. Here’s a bit on how to actually use it.

For Blog Writers

This is the easy part: if you write a blog, you can pretty well count on the fact that your blog already has a feed that automatically updates for every change you make. It’s built into all blogging software.

Usually all you have to do is invite people to subscribe to your feed so you can stop sending those emails out The hardest part about that is likely to be educating those who don’t know what you’re talking about. Feel free to direct them to this little tutorial…

For Blog Readers

I mentioned feed readers above. Well, there are gazillions to choose from. I’m going to simplify your life and just point you to my favorite: Google Reader. In this case, it’s an online service, provided for free by Google.

As an example, here’s all you need to do to subscribe to my personal (non-tech) blog:

• Find the link for my RSS feed – it’s labeled as ‘Entries (RSS)‘ at the bottom of the page. Copy the link (right click –> Copy link location)
• Visit http://reader.google.com/
• Click the ‘Add subscription’ link in the left-hand sidebar.
• Paste the feed link address into the entry field and click the Add button.

Now you’re subscribed, and anytime you come back to Google Reader it will tell you if there are new articles to read on any of your subscribed feeds. You can unsubscribe at any time. Note — subscribe / unsubscribe has nothing to do with paying for the feeds.

If you use Firefox — and you should: Firefox has some built-in features to help with subscribing to feeds. It can be set to add new feeds to your Google Reader account with just a click or two. Very cool.

But I Really Prefer Email

If you’d rather do all your blog reading via email, there’s an option for that as well. There are services such as FeedBurner that provide an RSS-to-email gateway service. Typically a blog writer/publisher will register his feed with FeedBurner, and then FeedBurner monitors that feed for new content. Anyone else can go to FeedBurner and register to receive email (how quaint) when the feed is updated. It’s another avenue to explore if you’re interested.

Password Protected Blogs

Unfortunately, password-protected blogs are a problem for syndication / feeds. I’m not aware of feed readers that are able to support password authentication in conjunction with the feed. So at least for now, email blast-o-grams are probably still the best bet to advertise your new posts. Yes, it’s a hassle, but it’s the trade-off for privacy. I’m sure some more smart (and lazy) people will have a solution to that problem before too much longer.

More Information

What is RSS?

How to: Read RSS Feeds

Wikipedia: RSS

http://www.skrenta.com/2007/05/14_rules_for_fast_web_pages_by_1.html

I read through the slides, and I’m looking forward to reading the whole book on safari.

When firebug is running, firefox seems to freeze up for a second or more, and the cpu load spikes up. In a new firefox session, everything seems to work fine. But over time, the problem gets worse, and I usually end up having to restart firefox at least once daily. Really a bummer for productivity.

I think I found at least a partial workaround: by disabling firebug when I’m not actively using it, the problem is much less severe. I can’t say that it goes away completely, but at least firefox is still usable.

On the firefox menu: Tools -> Firebug -> Disable Firebug

% rake db:migrate VERSION=0 && rake db:migrate

I finally got content filtering fully configured for my home network. I’m using DansGuardian + squid for the filtering, and the setup process was extremely smooth and easy (ubuntu breezy badger):

% apt-get install dansguardian

I already had squid running and tested, and DansGuardian just plugs right in. I had to change my WPAD proxy config (/wpad.dat) from this:

return "PROXY 192.168.1.10:3128; DIRECT";

to this:

return "PROXY 192.168.1.10:8080; PROXY 192.168.1.10:3128; DIRECT";

And it just worked — all the other machines in the network automatically got the new filtering. I also customized the ‘access denied’ html template a bit, in /etc/dansguardian/languages/template.html

There are probably other such online services — want to recommend another?

Easy solution: dnsmasq on my debian(ubuntu)-based server. Very easy to install:

% sudo apt-get install dnsmasq

This installs and configures the local DNS server, and sets it up to always run on the machine as a system service. A little bit of manual configuration is required as well: /etc/resolv.conf has the real nameserver addresses

search hsd1.co.comcast.net.
nameserver 68.87.85.98
nameserver 216.148.227.68
nameserver 204.127.202.4

And then I also put all the aliases for local machines into /etc/hosts

192.168.1.10 diener
192.168.1.100 doingo dad
192.168.1.101 velum vellum
192.168.1.200 gulliver boys

That way, all of the local network machines can resolve each other via this local DNS, and I don’t have to maintain /etc/hosts on each one.

Once the server (diener) is setup, all the other hosts on the network have to be configured to point to diener for DNS.  Currently everything is configured with static IP & DNS; dnsmasq also supports DHCP, and I need to still investigate using that for some hosts.

sudo apt-get install mysql-server-4.1 mysql-client-4.1

After that, tried to do some operation in rails, and it failed:

Mysql::Error: Lost connection to MySQL server during query: ....

That problem had a simple fix:

 sudo gem install mysql

The ruby mysql gem had to be rebuilt with the new mysql libs, then everything was peachy!

Lesson 1: Firefox – Opera – IE

After getting things working in Firefox, I found the next best platform for testing is Opera. Opera’s JavaScript engine seems a bit stricter on some things that Firefox, so it generates errors on things that Firefox handles fine. While not nearly as nice has FireBug, Opera does have an Error Console feature that is pretty clear on both the location and the meaning of the errors.

Finally it comes to testing Internet Explorer — IE6 to be precise. The tools for JavaScript debugging in IE6 are awful, it is a painful experience. So do all your testing on other platforms first.

Lesson 2: The real issues

Once I figured out a good debugging strategy, it wasn’t too hard to fix my problems. It came down to a few syntactic constructs that I now know to avoid in the future:

• Don’t use assignment in an if() statement
  • Okay in Firefox and Opera:
    if ( foo = something() ) { ... }
  • Alternative that works in IE6 as well:
    var foo = something(); if ( foo ) { ... }

• In hash / object construct, don’t leave dangling ‘,’ after last pair
  • Okay in Firefox:
    { a: 1, b: 2, c: 3, }
  • Alternative for Opera and IE6:
    { a: 1, b: 2, c: 3 }