pantheraspeech 0.95.0 — Lion, and the speech data manager. It's all here. V1.00 will be a feature-complete release with Snow-leopard driver support included, too.
Tools → Mac OS X speech data. Point it at an install disc image you own and it does the rest.
• Recognises Tiger, Leopard, Snow Leopard and Lion from Apple's own version file — not from the file name
• Extracts to the right folder, with a progress bar that says where it is
• No Python, no 7-Zip, no unpacking software, and nothing is downloaded
• Asks before writing over speech data you already have
It also refuses, and says why:
• the Tiger build whose engine loads and then crashes — found by hashing the engine off the disc in a few seconds, before anything is written. The version number does not predict this one; the disc that works reports an older version than the disc that does not.
• 10.8 and later — Apple dropped the 32-bit build of MacinTalk after 10.7, so there is no code on those discs this can run
• anything that is not a Mac OS X installer at all
Snow Leopard is recognised and its engine is on the disc, but there is no Snow Leopard synthesizer yet, so the tool says so rather than filling a folder nothing reads.
https://github.com/tgeczy/panthera-speech/releases/download/pantheraspeech%2Fv0.95.0/pantheraspeech-0.95.0.nvda-addon
https://www.eurpod.com/synths/lion.zip for the Lion data, hosted temporarily. Or just, you know, test the new tool with an archive.org image!
1
Reply
2
Boosts
also temporarily up on my site: Sequoia-alex. You can copy the folders inside: contents, which contains two things: resources, and info.plist. Both match Lion's voice format.
https://www.eurpod.com/synths/sequoia-alex.zip
Again, hosting over 2.5 Gigs of Apple voice content is a lot of bandwidth, so any mirroring on archive sites? please do it. for all the speech voice data, before I yank them off at the end of this week.