Skip to content

Latest commit

 

History

History
511 lines (346 loc) · 11.9 KB

CHANGELOG.md

File metadata and controls

511 lines (346 loc) · 11.9 KB

Voqal Changelog

Unreleased

2024.9.0 - 2024-08-12

Added

  • Viewing code problems to Idle Mode
  • Code smell correction to Edit Mode
  • Move Class Voqal tool
  • Goto Text Voqal tool
  • Rate limit retry loop

Changed

  • Open-source distribution
  • Upgraded to IntelliJ Platform Gradle Plugin 2.0.0
  • Default to including code with line numbers in prompts

Removed

  • Microphone audio streaming
  • Intent detection providers
  • Error reporting
  • Telemetry data collection
  • Trial mode
  • EAP distributions
  • Proguard obfuscation
  • WebXR client support

2024.8.4 - 2024-07-22

Changed

  • Improved edit mode directive shortcut

Fixed

  • Infinite loop in chunkText() on empty files
  • Exclude VCS ignored files from context prompt
  • Serenade intent detection configuration preview
  • MistralAI token limit defaults

2024.8.3 - 2024-06-28

Added

  • Fireworks AI language model provider
  • Groq Whisper speech-to-text provider
  • Speech-to-text language setting

Fixed

  • Preserve tab spaced indentation in edit mode
  • Groq LLM token limit defaults
  • Support multiple projects open simultaneously

2024.8.2 - 2024-06-24

Added

  • addUserContext() context extension function
  • getUserContext() context extension function
  • slurpUrl() context extension function

Changed

  • Set Anthropic default model to claude-3-5-sonnet-20240620

Fixed

  • Invalid role error in Anthropic language model provider
  • Invalid range error in chunkText() context extension function

2024.8.1 - 2024-06-21

Added

  • chunkText() context extension function
  • Voice probability indicator in Picovoice VAD settings
  • DeepSeek language model provider

Changed

  • Edit mode default to edit chunk size of 200 lines

Fixed

  • Picovoice VAD provider default configuration
  • Incorrect indentation edits in edit mode

2024.8.0 - 2024-06-13

Added

  • Voqal Idle Mode
  • Directive decomposition
  • Vertex AI multimodal provider
  • Google API multimodal provider
  • Anthropic language model provider
  • Custom language model provider
  • Project file structure to idle mode context
  • Custom URL prompt template provider
  • Chat window debug messages
  • View Source Voqal tool
  • Change Theme Voqal tool
  • Clear Chat Voqal tool
  • Ignore Voqal tool

Changed

  • Pause microphone stream on IDE focus loss
  • Voqal intents are now Voqal tools
  • Upgrade Picovoice Orca to 0.2.0
  • Prevent configuration changes in trial mode
  • Create file & create class tools auto-start edit mode

Removed

  • Voqal Command/Focus/Dictate Modes
  • Wake word detection providers

2024.7.0 - 2024-05-21

Added

  • Enhanced code mode
  • Enhanced dictation mode
  • Focus/Edit mode
  • Serenade intent detection provider
  • Ollama language model provider
  • Increased support for Swift programming language
  • SelectAll Voqal intent
  • SelectLine Voqal intent
  • SelectLines Voqal intent
  • StartOfLine Voqal intent
  • Unselect Voqal intent
  • Cancel Voqal intent
  • Delete Voqal intent
  • DeleteLines Voqal intent
  • LooksGood Voqal intent
  • Tab Voqal intent
  • PreviousTab Voqal intent
  • NextTab Voqal intent

Changed

  • Improved chat tool window
  • Ability to set model name per mode
  • Ability to see partial mode results
  • Ability to set speech silence threshold
  • Ability to set LLM temperature
  • Ability to set query params for Deepgram STT provider
  • Ability to view raw LLM response
  • Refactored live transcribe service
  • Tool confirm/execution messages

Fixed

  • EDT threading issues
  • Redo intent reporting as unavailable too early

Removed

  • Write code tools
  • XR tool window
  • Code mode keyboard accessibility

2024.6.3 - 2024-05-06

Fixed

  • Collection contains no element matching the predicate (#77)

2024.6.2 - 2024-04-25

Fixed

  • Mode switching failure in non-streaming speech-to-text providers

2024.6.1 - 2024-04-05

Removed

  • Automated usage of InsertText Voqal tool

Fixed

  • Slow operation error when viewing command prompt

2024.6.0 - 2024-04-03

Added

  • Early access distributions
  • Prompt library settings
  • Toggle Code Mode Voqal intent
  • EndOfLine Voqal intent
  • DeleteLine Voqal intent
  • InsertText Voqal tool
  • Voqal mode caret icon
  • Support for Rust

Changed

  • Default keymapping
  • Dictation mode prepends comments in appropriate files
  • Improved RunProgram Voqal tool
  • Renamed CreateNewClass Voqal tool to CreateClass
  • Improved PSI name change propagation

Fixed

  • Empty speech balloons in conversation tab (macOS)

Removed

  • Edit prompt status icon action

2024.5.0 - 2024-03-22

Added

  • Voice activity detection settings
  • Voqal voice detection provider
  • Picovoice voice detection provider
  • Language model token limit setting
  • Context cropping functionality
  • Groq language model provider

Changed

  • Improved error reporting for Helicone observability provider

Fixed

  • Redundant files in command context
  • Unsatisfied link error in WebXR client
  • Race condition in plugin configuration
  • Trace/debug level logging not outputting to log tab
  • Empty speech balloons in conversation tab

2024.4.1 - 2024-03-18

Fixed

  • Dictation mode does not trigger when using Picovoice intent detection provider

2024.4.0 - 2024-03-15

Added

  • Toggle Dictation Mode Voqal intent
  • Streaming speech-to-text & text-to-speech functionality
  • Picovoice streaming speech-to-text & text-to-speech providers
  • Deepgram streaming speech-to-text & text-to-speech providers
  • Deepgram intent detection provider
  • Deepgram wake word detection provider
  • Optional organization id setting for OpenAI provider

Changed

  • AI providers pre-initialized on plugin start
  • Re-organized plugin configuration settings
  • Decreased plugin logs font size
  • Improved error handling for network issues

Fixed

  • Add/remove breakpoints not working in non-JVM languages
  • Properly close microphone line on configuration change
  • Issue causing intents to be triggered on partial matches
  • Holding push-to-talk button causing multiple transcriptions
  • NPE when using WriteCode tools in Python files
  • No change on plugin configuration reset

Known Issues

  • Dictation mode does not trigger when using Picovoice intent detection provider

2024.3.0 - 2024-03-01

Added

  • WebXR client support
  • MistralAI language model provider
  • Whisper ASR speech-to-text provider
  • Optional data-sharing setting
  • Open files to system prompt
  • Enable/disable intents feature
  • Add Field Voqal tool
  • 16 new Voqal intents

Changed

  • Logo update
  • Improved LLM response parsing

Fixed

  • EDT threading issues

2024.2.4 - 2024-02-12

Fixed

  • Plugin crash on Mac (aarch64)

2024.2.3 - 2024-02-10

Added

  • Plugin error reporting
  • Plugin uninstall feedback form

2024.2.2 - 2024-02-03

Added

  • Logs tab for plugin troubleshooting

Changed

  • Intents are now additionally invokable via command

2024.2.1 - 2024-01-30

Fixed

  • NoClassDefFoundError in the write code tool

2024.2.0 - 2024-01-22

Added

  • Function signature change propagation
  • Active microphone configuration

Fixed

  • Issue upgrading Picovoice natives causing IDE crashes

2024.1.0 - 2024-01-11

Added

  • Ability to trigger intents (actions without wake words)
  • Ability to chat with Voqal via keyboard
  • Control language model from status widget
  • Modify Text Voqal tool
  • Stop Listening Voqal intent
  • Scroll Voqal intent
  • Goto Line Voqal intent
  • Toggle Zen Mode Voqal intent

Changed

  • Set default OpenAI language model to gpt-4
  • Upgrade Picovoice Cobra to 2.0.1
  • Upgrade Picovoice Leopard to 2.0.1
  • Upgrade Picovoice Porcupine to 3.0.1
  • Replace welcome message from voice to chat

Fixed

  • Issue causing audio directly after wake word to be cut off

Removed

  • Closed captioning setting

2023.3.1 - 2024-01-03

Changed

  • Updated logo
  • Improved LLM response parsing

2023.3.0 - 2023-12-26

Added

  • Ability to edit system prompt
  • Ability to add custom tools
  • Custom wake words
  • Hugging Face language model provider
  • TogetherAI language model provider
  • Helicone observability provider
  • Create New Class Voqal tool
  • Show Tab Voqal tool
  • Remove Breakpoint Voqal tool
  • Play Program Voqal tool

Changed

  • Improved LLM response parsing
  • Reorganized plugin configuration settings
  • Changed default wake provider to none
  • Increased support for Groovy
  • Disabled assistant mode

Fixed

  • Obfuscation bug preventing wake detection
  • Push to talk trigger issue

Removed

  • Unknown Command Voqal tool

2023.2.2 - 2023-12-17

Changed

  • Increased support for multiple code changes per transcription
  • Better invalid configuration error messages

Fixed

  • Issue causing configuration changes not to take effect
  • Missing pending status triggers

2023.2.1 - 2023-12-12

Fixed

  • Picovoice STT provider configuration

2023.2.0 - 2023-12-12

Added

  • Open File Voqal tool
  • Add Breakpoint Voqal tool
  • Play voice button to TTS settings
  • AssemblyAI to Speech-to-Text providers
  • Push-to-talk keyboard shortcut (default: CTRL+SHIFT+S)
  • Status notification sounds

Changed

  • Write Code Voqal tool now uses a PSI-based back buffer

Fixed

  • Erroneous flagging by Windows SmartScreen

2023.1.11 - 2023-11-18

Fixed

  • Issue booting Picovoice SST provider

2023.1.10 - 2023-11-17

Changed

  • Increased support for macOS

2023.1.9 - 2023-11-17

Changed

  • Increased support for macOS
  • Use Picovoice as default SST provider

Fixed

  • Issue saving model name configuration changes
  • Delayed installation issue

2023.1.8 - 2023-11-16

Added

  • Initial release