Developer blog: marco

Alright it has been a while but I am trying to pick up the pace again. This whole work thing is taking up a lot of time and those bills need to be paid. After lots of prodding from Theo I finally got around to figuring out why there were so many false-positives of IPMI on older boxes. It was quite silly actually, the BIOS scanning code returned sometimes -1 whenever it did not find anything useful however the function that called the signature check was looking for != 0. Thus a bunch of older boxes ended up returning -1 yielding a false positive. Silly bug and clearly a remnant from the initial user space implementation of it.

On an ACPI note Jordan and I have been busy working on that as well. Jordan did a major overhaul of the AML parser and interpreter. He made some assumptions that sounded good at the time but proved to be wrong. ACPI implementations use AML as a full blown language. For example, expressions can be built up of sub-expression, variables can be all kinds of wild weird weird combinations. These vendors are pretty darn odd by implementing such deep expressions. Anyway Jordan made the code a lot more dynamic and the parser and interpreter are becoming better. Not yet ready for consumption but we are making good progress. I am hoping we'll be able to start working some more on the little devices when the interpreter is slightly further along. My first goal is going to get some battery stuff going.

That said, we are nearing beta status and we need to concentrate in making the release as good as we can. So you guys need to start testing snapshots and report issues.