Exile From the Herd - Hacking tech

iPhone unlocked on the Fido network using Turbo Sim card

nospam@example.com (Mark Jeftovic) — Tue, 04 Sep 2007 23:40:06 -0400

I bought a turbo Sim card on the aftermarket (at a pretty inflated price), from PDAPlaza, it arrived pretty quickly.

I was going to wait for a software unlock but I broke down before the weekend and grabbed a couple turbo SIMs.

There are a lot of docs out there on what to do, I thought I'd post a quick summary of what you need to do to get iphone unlocked on a Canadian network like Fido or Rogers (I guess the point is this works on pretty well any GSM network)
Continue reading "iPhone unlocked on the Fido network using Turbo Sim card "

easySPF: an ajax enabled SPF wizard

nospam@example.com (Mark Jeftovic) — Tue, 21 Aug 2007 20:13:39 -0400

Creating Sender Policy Framework (SPF) data is a task that naturally lends itself to ajax. As you tweak the parameters of your policies, the spf record changes on the fly. This was a good opportunity for me to learn jquery and I was pretty impressed with how fast jquery enabled me to get a basic framework up and running.

What took me the longest was figuring out how to get the contents of my form processing via an external php script before being rendered into my inline <div>, after awhile I gave up and asked one of my programmers. Turns out it's the load() event handler, but the jquery page doesn't really make it evident in the documentation that what I want to do is this:



 $("#my_div_id").load("some_script.php", { myvar: myvar } );

Then I can access the myvar and anything else passed in the parameter list via the php $_REQUEST structure.

Anyhoo, the result is easySPF: an ajax enabled SPF wizard.

Migrating your feed subscribers to Feedburner

nospam@example.com (Mark Jeftovic) — Mon, 04 Jun 2007 22:33:07 -0400

I spent a few hours friday working on a blog widget that would parse the number of subscribers out of the HTTP_USER_AGENT strings of the various feedreaders so I could easily see the aggregate number of subscribers I have. It only took me a few hours to realize that what I was trying to do, wasn't gonna work.

My plan to use remote javascript may have worked fine for human readers to the main index, a la MyBlogLog (which I use), but of course, duh, the feedreaders won't load the main index and they likely won't interpolate the javascript. I doubted an iframe would work either.
So my idea for a widget wasn't going to fly, I realized the only way to do this was to actually run the entire feed URL through the counter system, the way FeedBurner does it. It was only a coincidence that this happened on the same day Google bought FeedBurner for 100 Million.

So the blog widget is a non-starter, I still want to easily know my aggregate subscribers but now I don't want to start counting from zero, I want to be able to factor in my existing readers.

My existing readers subscribe to my feed via http://mark.jeftovic.net/feeds/index.rss2, which is the same URL feedburner will use to load the feed. So simply redirecting my existing feed URL to my new Feedburner feed would create an infinite loop when FeedBurner tries to load the old feed.

We need to always redirect the old feed URL to the new feed URL except when the remote client is Feedburner.

My blog platform is serendipity, and it implements all the RSS and XML urls by a URL rewrite to rss.php, so at the very top of that file, we just add this:



if(!eregi("^FeedBurner",$_SERVER['HTTP_USER_AGENT'])) {

        Header("Location: http://feeds.feedburner.com/jeftovic/JVMu");

        exit;

}

This should allow Feedburner to load the RSS from the blog, and redirect all the other feedreaders to Feedburner. Hopefully I haven't overlooked anything. This is somewhat of a test post and if all goes well, it should show up in my Bloglines reader transparently.

To migrate to FeedBurner from other blog platforms including Wordpress and Drupal, there is this thread on the FeedBurner forums.

Useless mash-up between pdns pipe backend and md5

nospam@example.com (Mark Jeftovic) — Sat, 20 May 2006 16:20:33 -0400

I've had the idea to do some tinkering with powerdns' pipe backend for awhile and had decided to rig up a quick and dirty MD5 encoder/decoder database on a domain I've had for awhile MD5.ORG

So it was pretty easy. Now anybody can quickly get an MD5 hash on any string that'll fit inside a DNS query packet by doing something like:




host -t txt <string>.to.md5.org

or the slightly less readable




dig -t txt <string>.to.md5.org

Then for completeness you can try and see if we already know what string creates a given MD5 hash and retrieve it using




host -t txt <md5hash>.from.md5.org

There is not a lot of practical value to this, it was just a neat hack to learn the basics of the pdns pipe backend, which I like a lot.

Ebay's move to free and grokking PHP's Services_Ebay

nospam@example.com (Mark Jeftovic) — Fri, 16 Dec 2005 22:21:41 -0500

Ebay made a very smart move recently and removed most of the costs associated with their developer platform. I think this will bring a lot of innovative new services to the market as the barrier to entry is now largely technical and not financial.

One of the reasons Amazon's AWS and affiliate program is so ubiquitous across the net is largely due to two things: 1) the depth and stickiness of the Amazon site and 2) the free access to their affiliate and AWS services.

Ebay has one of the few sites whose depth and prevalence approaches that of Amazon, although a back-of-the-napkin comparison shows Amazon dwarfs Ebay in backlinks by about 28 million to 6 million.

I started playing with PHP's pear class Services_Ebay and after some struggling, I finally have a handle on it and thought I'd post some of the "aha" links which helped get me there.

Programming eBay Web Services with PHP 5 and Services_Ebay covers getting up and running. Here are a couple of handy calls to know:




$session = Services_Ebay::getSession($devId, $appId, $certId);


$session->setToken($token);



$call = Services_Ebay::loadAPICall('GetCategories');


$call->describeCall();


$result = $call->call($session);


print_r($result);

The describeCall() describes any method you care to load via loadAPICall() and you can see the complete list of methods in the package via getAvailableApiCalls()

The describeCall() output links to the eBay Developers documentation on the given call.

For example, getCategories() is documented here, which explains which parameters to pass to obtain the version number of the current category list and what to pass to dump the entire category tree.

One silly problem I ran into wass, how do I pass the parameters to the getCategory() call? It took me awhile because I couldn't find it via googling so I had to experiment.

Turns out it's




$session = Services_Ebay::getSession($devId, $appId, $certId);


$session->setToken($token);


$ebay = new Services_Ebay($session);



try {


                $args = array(
'DetailLevel'=>1,
'ViewAllNodes'=>1,


                          );


        $result = $ebay->GetCategories($args);


        print_r($result);


}   catch (Exception $e) {


        echo "Something went wrong.";


        echo $e;


}

If you're going to cut your teeth on the getCategories() call like I did, remember to lift the default 8M memory limit on your script when you dump the entire tree.

So I'm looking forward to experimenting with this API.