I’d been hoping for some speech recognition improvements in Leopard. Not because I plan on laboriously commanding the mac to do things I could do by myself easily enough using the native mouse interface, but because I’ve long wanted to write a ruby program that takes a (say) mp3, reads it to the speech interface, and gets something useful, like text, back. It might be too error-prone to really be useful, it might be much harder than I think – but at least I’d like to know the basic capability is there!
Alas, not. In a spirit of freewheeling experimentation, forgiveness of past sins and water under the bridge, I turned on Speakable Items and spent the next few minutes trying to make the useless thing recognise a single thing I said. It, of course, could not, no matter how slowly and careful I enunciated each word, or imitated Steve Job’s californian drawl.
Speakable Items is such a worthless piece of crap I wonder why they even bother including it. Ah well .. roll on 10.6!
UPDATE: Amusingly, 10.5 underlines “Speakable” as I type this, claiming it’s a spelling error. Apple are so ashamed of their retarded child they didn’t even teach its name to the internal dictionary!