Magenta:Voice
The words magenta's voice uses: how a memo's transcription may mishear a name, the words it should expect, and how the narrator and read-aloud should say a word. The page is read every ten minutes or so (magenta's services/voice_words.py), so a change here needs no deploy.[unverified]
Only a trusted editor's revision is used (JMyles, SkymanJenkins): anyone may suggest a change, and it's used once one of them has saved or verified the page. A line in a comment, <nowiki> or <pre> is never read, so examples can sit there.[unverified]
Each list is a section of * lines. A pair is written word → meant (-> works too).[unverified]
Heard as
A name a transcription writes wrong, then the name it is. It's only read as that name where a name is plainly being said to someone ("Magnet, can you...", "at magnet").[unverified]
- magnet → magent[unverified]
Expected words
Words a memo's transcription should expect, spelled as they should come out. Everyone's names, the agents' and the Moods' are added by themselves.[unverified]
- banjo[unverified]
- mandolin[unverified]
- dobro[unverified]
- fiddle[unverified]
- upright bass[unverified]
- bluegrass[unverified]
- flatpicking[unverified]
- Scruggs[unverified]
- Cryptograss[unverified]
- PickiPedia[unverified]
- magenta[unverified]
- magent[unverified]
- Mood[unverified]
- Moods[unverified]
- hunter[unverified]
- maybelle[unverified]
- delivery-kid[unverified]
- Ethereum[unverified]
- blockchain[unverified]
Said as
A word, then how the voice should say it, wherever it reads: messages, intros, "is thinking". Whole words only, in any case; a capital is kept where the word had one. [unverified]