I guess Google really does use Divmod as their source for new ideas!

First, as we were implementing our email system, they launched gmail. Then while we were working on VoIP they acquired Grand Central. Now that we're focusing on infrastructure... they've released their infrastructure!

(If you are familiar with Axiom, the appengine database model example should be good for a laugh.)

Of course, I'm wondering: should Divmod, or me by proxy, be worried about this? I think I've settled on "no", for a few reasons:

1. Google's release of appengine is an indication that this is a hot group of ideas to be working on right now. It sets a precedent for our open source offering to be idiosyncratic. When describing Mantissa's scaling model, I've often heard the objection "But with EC2 I can run whatever code I want!" While that objection is somewhat inaccurate and I can explain with some effort that it's not true, now I have the much snappier retort: "but if you want to use Google's stuff you have to write to a specific API too".
2. You can't get all the code to appengine, at least for now, so there's a big category of applications that can't use it. We can keep working to address it — and those were really the applications we were already focused on anyway.
3. appengine doesn't include Athena. It doesn't include Vertex. And it doesn't include Imaginary. There's plenty of fun stuff we're working on outside the realm of simply deploying a web app.

All in all, I'm thrilled that AppEngine is in Python. I hope that we'll be able to get Twisted into the mix at some point, but there's always the possibility that we'll need to integrate with anything a major player releases, and I'm relieved that this (very) major player has decided to go with something that will be easy for us to talk to.

It seems to me that most of my friends are the sort of busy, busy people who always have a half a dozen side-projects which they have the skills but not the resources to complete.

Today, while chatting with such a friend, he mentioned something about a project which, if he had a 2-week vacation, he'd probably be able to get to. After spending a week sleeping, doing nothing, and puttering around, of course, he'd pick up the project in the second week. Of course the likelihood that he would have such a vacation is basically zero, as is the likelihood that the project get done. (Of course, anything can happen, and some things often do.)

I propose a term. This is a second week project — a project that only takes place in the second week of a hypothetical mini-sabbatical. I feel like this could be a useful piece of shorthand for a concept which would otherwise require a few tedious sentences. Am I right in thinking that we all have at least a few of these?

Since the acquisition, I have been wondering if I'd leave my LiveJournal account behind. While I've been an LJ fan for a long time, Blogger has some features which make it appealing:

• Nicer looking templates
• More reserved "toolbar" area
  
• More diverse client software
• Not owned by the mob
• Far more customization options and "widgets"
• All my friends are doing it
• Custom domain names (I've been waiting for 5 years, "glyph".lj.com isn't going anywhere...)
  
• "Blog" is a distinct entity from "Account"
• Blogs which have multiple contributors aren't a special thing
• Nicer profile support
• Backlinks
  
• AdSense support
• Server-side drafts
  

While I plan to do more of my blogging here from now on, I don't plan to completely abandon LJ any time soon because it does have some cute other features:

• the "friends" page
• userpics
• moods
• GPG public key publishing

I want to say something about their file-hosting feature, but that's always been kind of broken, so I've never used it.