Can’t sleep again due to too many geeky things flying about in my head, gah!
The next major feature to add to redbus will be bus stop data updates; I already have code to pull the stop information from lothian’s website. I need to make it check if anything has changed and automatically publish a new stop database to the website if so. The redbus clients then need to check for updates and download ’em when necessary. I’ve left this bit till last as its pure plumbing and not particularly interesting. However, it needs to be done so the map stays current. I reckon it’ll probably take a couple of evenings.
In the meantime I’m porting my Hero USB patchset to the leaked 2.6.29 hero git tree used by the villainrom guys. There’s a forum topic about it here. Hopefully this means it’ll be included in an "official" unofficial Villainrom release, opening it up to a much wider set of users. I’ve already done a fair bit of the porting this evening, I just have the USB function drivers left. And then I can try compiling it and watching the errors fly 🙂
I’m /also/ engaged in building my own Hero android open source build from scratch based on cyanogen’s userspace android tree. I’m removing as much proprietary cruft as I possibly can, reverse engineering+rewriting the stuff that isn’t too complex… for example, I should soon have a completely opensource replacement for libhtc-acoustic.so. I’m also going to kill the TI proprietary bluetooth stack in favour of the more normal bluez; my mate Colin has already had success getting "hciattach" to work properly on the hero. This means we can get things like bluetooth keyboards working, which is impossible with the closed source TI bluetooth stuff. Mind you, you could just plug a USB keyboard in now, but having the choice is nice 🙂
I also know how to (farily easily) add hardware MP3 and AAC hardware decode acceleration to the AOSP android builds, something that seems to have been missing up till now.
adq 