Archive for November 4, 2006

Saturday, November 4, 2006 [Tweets] [Favorites]

Microsoft Office Stats

Brad Post says that the Mac version of Office has 30 million lines of code (428,000 per developer), with an average age of about 12 years, and that the development team has had 50% turnover over the last two years. So they’re working hard, but it’s a lot of work.

Disco, FlexTime, and the HIG

Rory Prior comments on the usability of a new HIG-violating app, and Daniel Jalkut illustrates the process of developing the “snazzy” new interface for FlexTime 1.1. It’s a nice illustration of the domino effect that making a few changes can have. I’d like to see what his new layout look like using standard controls.

Update: Paul Kafasis and Erik Barzeski discuss “The Delicious Generation.”

I do agree that there’s a need for streamlined disc burning, which is why you can Shift-drag a file or folder onto DropDMG (Shift because a regular drag creates an image), insert a disc, and click Burn. Disco lets you add a bunch of files to its window before burning them, but if you want to collect files from different places, I think Finder burn folders work pretty well, and the Finder window doesn’t constrict you. The Finder and DropDMG will also let you burn multiple discs at once. Disco’s Finder-format Spandex feature is a good idea, and I’ll probably add something similar to DropDMG, but it’s not as amazing as it sounds. It simply divides the files so that they’ll fit on multiple discs. This doesn’t make efficient use of your disk space, and if a file or subfolder is too large to fit on a single disc, it won’t work at all. With DropDMG and Toast you can span arbitrary items across multiple discs. DropDMG does this in two steps: first create a segmented .dmg file, then burn the segments. Toast does it in one step—you just feed in discs one after another—but the data is stored in its proprietary format.

I have mixed feelings about Delicious Library. The sizzle doesn’t bother me because I do think it looks good, and it does this while being quite functional. What I don’t like is that the iSight scanning was over-sold. First, contrary to popular belief, this appeared in Booxter first. And second, it just doesn’t work very well with built-in iSights, taking much longer than entering the items manually, in my experience. The other problem is that it gets really slow when the library contains a few thousand items. The friendly support person showed me how to run a regex on the XML file to remove the “related” items that it was storing, and this made a big difference as they outnumbered my real items by several times, but an otherwise friendly consumer app shouldn’t make me do that.