We’ll Be Back Soon

I’m not even going to say where I saw this… but if Amazon did this every time they updated their offerings, they’d be out of business by now. I really wonder what the site build/deploy situation is like at Apple. I used to know the guy that pushed the magic button for the Japanese web site, but he would never spill the beans… now he’s at Google. Oh well.

BBM Music Rocks

I don’t have a BlackBerry, though I did a couple years ago (and having just come back from Japan, I literally thought the device was a sick joke). Anyway, the BBM Music service that’s rolling out and everyone is making fun of is quite interesting. Sure, you only get 50 songs a month; sure, you can only swap out 25 of those songs each month; sure, it’s spendy at $5 (er. $4.99)… but you get to listen to the music of all your friends on the service. I thought only Facebook could come up with this kind of forced virality. They have it all wrong though – all carrot and no stick.

Steve Jobs Resigns

Creating and destroying industries.

It’s everywhere, so you don’t need to read my thoughts on it… but I did feel like writing something, anything, as I feel a little bit of my soul has died. I know that sounds corny, but I’ve had a personal dream of working for the one man who is living firmly in the future. I didn’t pursue this as much as I could and now that chance seems to be gone. I’ve pieced together a lot of information and believe I know what this means for the man and his company. I won’t go into that here, figure it out yourself.

BTW, Steve Jobs Is Not Dead.

Updating Mail.app Bundles for Compatibility

I’m going to port some more of my technical articles from tumblr over here. This is an oldie but goodie as I still get a lot of traffic for it. Original post is here.

Sometimes when new versions of Mail.app come out, the plugin architecture isn’t always backwards compatible. To disable plugins that could potentially crash Mail.app, a plugin verification occurs via a PluginCompatibilityUUID check.

So, you need to (WideMail is used in this example):

  • add the Message.framework PluginCompatibilityUUID to the SupportedPluginCompatibilityUUIDs of each plugin
  • add the Mail.app PluginCompatibilityUUID to the SupportedPluginCompatibilityUUIDs of each plugin

Add the Message.framework PluginCompatibilityUUID

  1. Open up /System/Library/Frameworks/Message.framework/Resources/Info.plist and search for PluginCompatibilityUUID. In 10.6.5, this is 857A142A-AB81-4D99-BECC-D1B55A86D94E.
  2. Open up ~/Library/Mail/Bundles/WideMail.mailbundle/Contents/Info.plist and search for SupportedPluginCompatibilityUUIDs. Add a new string to this array with the value 857A142A-AB81-4D99-BECC-D1B55A86D94E.

Add the Mail.app PluginCompatibilityUUID

  1. Open up /Applications/Mail.app/Contents/Info.plist and search for PluginCompatibilityUUID. In Mail.app 4.4, this is BDD81F4D-6881-4A8D-94A7-E67410089EEB.
  2. Open up ~/Library/Mail/Bundles/WideMail.mailbundle/Contents/Info.plist and search for SupportedPluginCompatibilityUUIDs. Add a new string to this array with the value B842F7D0-4D81-4DDF-A672-129CA5B32D57.

Done.

This is a kluge and there is a reason Apple requires this check. It’s best to wait until the official developer releases an updated plugin.

WTF, Comcast?

First off, I’m going to start posting technical articles here again. I’ve been on hiatus doing a web curation project over on Tumblr. That was good fun and all, but I started tailoring the content more to what the audience wanted to see than what I was interested in. A brief example is that an image post would ALWAYS do better than a quote or text post. I like writing, so here I am.

Ok, Comcast and Level 3. I know a thing or two about peering, I wrote a graduate level economic analysis of the practice when WorldCom was set to take over everything, but I’m not going to go into any details because I’ll probably get something wrong. Almost every article bouncing around the echochamber has glaring flaws in it, but I’m not going to go into that either.

Here is what I want to know and that nobody is telling me. Level 3 is a backbone provider getting into the CDN business. As a backbone provider, it is capable of entering into a peering arrangement as traffic flows on/off of their network. Level 3 has many CDN competitors: Akamai, Limelight, etc. These guys don’t run networks, so peering is not an option. This leaves them having to pay Comcast and the various ISP it delivers traffic to and thus gives Level 3 a pricing advantage.

Now Comcast is stating that Level 3 is a CDN and must pay. This is all well and good, but Level 3 is BOTH a network provider and a CDN. The fees the other guys pay generally have to do with colocation at the ISP peering points and not some much (or if at all, with network traffic)… and Level 3 does not have to pay the colo fees as they own equipment in these same facilities. So the question is, how much is Comcast charging regular CDNs for colo services vs. how much are they charging for network traffic?

If the charge for network traffic is non-zero (or pretty damn close), I’m at a loss. The CDNs seem to be doing the ISPs a favor by bringing the content closer to their customers and removing unnecessary traffic costs. I can understand a traffic charge for content flowing over a pure transit network (as opposed to a peered network), but I doubt such traffic would cost the ISP anything as they are on peered connections.

If anyone can point me to a sane explanation of the above and why Comcast feels it can charge a hybrid CDN/Network to deliver traffic to its own customers, the ones that made the request in the first place, it’d greatly appreciate it.

This post is intentionally light on technical detail.