Changelog
What changed in each version of Vociferous, newest first.
Version 0.3.3
Labels that apply themselves, and bulk actions that no longer wait.
- Labels can apply themselves. In Settings, Labels, the Automatic Labels card sets how sure the app must be, and a few minutes after a new transcript is saved, each of your labels whose suggestion reaches that floor goes on by itself. Take one off and it stays off. Apply Now, after asking, does the same for the transcripts you already have. The automatic part is off until you turn it on, because we have met software before.
- Deleting many transcripts at once is much quicker, and so are restoring them, deleting them for good and emptying Deleted. Previously the app handled seventy transcripts the way a toddler handles seventy peas.
- A label is drawn in the colour you chose for it. What a bold concept.
- An open artifact's bar and menu offer only what can be done to the artifact, and none of its transcript's own actions.
- The capture summary is tidier, and “Copy as” is now a setting on your account, Plain Text to begin with.
- A transcript says when label suggestions are not available, instead of showing nothing and hoping you would not ask.
- How We Count says what the median wait measures and what a take is, and the Daily Usage title in Insights now sits centred over its map.
Version 0.3.2
Labels, and a phone that works as a companion.
- Labels are called labels everywhere, and a label can sit under another label, so “shop” can hold “orders” and “suppliers”.
- On a phone the app is rebuilt as a companion: a bar along the top for the main places, a record view whose own tabs you can swipe through, and titles that wrap instead of cutting off.
- Filler words such as “um” and “uh” are kept in transcripts from this version on, so every word stands as it was first transcribed.
- Insights' Overview says how long you spoke, out of how long you recorded, and how many times faster that was than typing.
- On a computer the navigation rail lines up with the page below it, and the logo in the corner opens vociferous.app in a new tab.
Version 0.3.1
Label suggestions, and takes that keep recording.
- Vociferous can suggest labels for a transcript from what it says, once you have accepted the terms that name the service it uses; turning suggestions off sends nothing.
- A transcript offered several labels stays in the review queue until each one is answered.
- A take left recording is never paused for a break, and a dropped sign-in is asked for again instead of ending the take.
- Export everything carries your library beside your transcripts, and says what it leaves out.
- How We Count says what the counting actually does, and a take the speech service refused tells you so.
Version 0.3.0
The redesign.
Every view is rebuilt on one design system, Minty Forest by day and Minty Forest at Midnight by night, with a light mode and a dark mode that follow your computer or your choice.
- Transcribe, Transcripts, the transcript itself, Grove, Instruments, Insights and Settings each have a new layout that uses the whole screen.
- Deleting a later take of a continued transcript says which versions go with it, and removes its words and its recording.
- Search finds a transcript by words that only its current version holds, and the preview shows what matched.
- A speaker's colour can be chosen from the same palette as a label's, and is kept.
- The terms are asked for again when a sentence in them changes, showing the sentence that changed.
Version 0.2.3
Steadier sessions and long transcriptions.
- A signed-in tab never sends a request without its session, and an ended session says so plainly instead of failing quietly.
- A long transcription is woken the moment the speech service finishes, and an hourly check re-rings any job left waiting.
- Every export includes your original transcript by default, and you can untick it.
- A transcript already in an account is recognised when it is synced again, so nothing is copied twice.
Version 0.2.2
Augment, Insights and plans, made clearer.
- Augment This Capture uses every Instruction you picked, in the order you picked them, and a capture can pick several.
- A long augmentation that runs out of time is sent once more, keeping the parts that finished.
- Insights counts words, sessions and active days from your whole history, and each figure says what it compares.
- Your plan's meters say what they count, and a refused transcription says what fell short and when your minutes reset.
- The session-expiry warning says what each button does.
Version 0.2.1
The first numbered version.
The app and the API now report one version, read from one file.
- Insights counts each transcript's current version, voice minutes as whole minutes, and a personal record on the first day it was reached.
- Exports write lengths in the app's one duration form, and a continued transcript's speaking pace counts every take.
- A skipped augmentation gives one true reason, and a budget reason links to Augment.
- Erasing an account also removes the upload chunks of its imports.