GDNote isn’t quite ready to launch yet, but we have started giving out access to the application (email if you want in). Yesterday we got our first real user. Obviously one user alone isn’t enough to make a successful service, but it’s a big step towards going live - this is feeling more like a real service every day.

The work we are doing on the application is changing too. The mobile clients are still in full active development - developing mobile apps is a lot harder than anything on the server as you have to start from scratch for a lot of stuff. However, on the website itself I think we are done with new features and performance improvements until after launch. Over the last few days we’ve made some bigger changes like cutting the initial load time (on a slow connection, slow vm, lots of data) from 10 minutes to 10 seconds and removing a 5 second delay on every button click. But now it’s actually quite usable so we have to focus on all the little details that don’t affect my own usage too much but still need to be done.

Mostly this means writing content and making small style fixes - nothing too serious, but it’s amazing how many small tasks are necessary to get something ready for release.

There’s also cross browser testing - the one thing that made IE having 95% market share seem good.

FireFox and Safari are pretty well tested already, since I’m using Firefox for my own notes and Safari for the test user account - more than one google login in a single browser is more trouble than it’s worth. That probably covers a fairly large proportion of the people who would be interested in something that is like OneNote but doesn’t require Windows. We are going to have to support IE though - I’ve done a few tests in IE7, and it seems to work ok beyond some issues with the meaning of 100% that were easy enough to fix. What we haven’t done though is testing with IE6 - I wish we didn’t have to, but unfortunately there are a lot of locked down machines out there with no other choice of browser.

At least it’s not as bad as when I had to write scripts that would run on IE3 and Netscape 3 - they had completely different ways of handling layers, and Netscape 3 was just as hard to get rid of completely as IE6 is now.