What's Changed

• Set Qdrant on_disk_payload to true by @ggordonhall in #1195
• Additional chunks based on references and definitions in code and proc tools by @rmuller-ml in #1180
• UI fixes by @anastasiya1155 in #1197
• add scope-queries for cobol by @nerdypepper in #1196
• Refresh token on 401 error in LLM gateway, and quota requests by @calyptobai in #1215
• Multi-repo by @anastasiya1155 in #1153
• Minor UI fixes by @anastasiya1155 in #1219
• fix ts issue with file-icons-js import by @anastasiya1155 in #1220

Full Changelog: v0.5.12...v0.6.0

What's Changed

• Tweak rlimits by @ggordonhall in #1181
• bump version to 0.5.12 by @ggordonhall in #1183
• Fix error when pasting to the input by @anastasiya1155 in #1184
• Up soft rlimit by @ggordonhall in #1185
• Add more try/catch blocks to autocomplete input by @anastasiya1155 in #1187
• Allow dot in autocomplete mentions by @anastasiya1155 in #1188
• Qdrant keyword path filter in proc by @ggordonhall in #1190
• Log rlimits by @ggordonhall in #1191
• Do not reduce limit if > 10,000 by @ggordonhall in #1193

Full Changelog: v0.5.11...v0.5.12

macOS (Apple Silicon) - bloop_0.5.11_aarch64.dmg
macOS (Intel) - bloop_0.5.11_x64.dmg

What's Changed

• Allow quick & shallow copies of Git repos to select which repos to index by @rsdy in #1115
• Pin 1.73 rustc by @rsdy in #1148
• Improve scrolling in chat and studio by @anastasiya1155 in #1149
• Automatically open bloop app after authentication with GitHub by @anastasiya1155 in #1146
• Get user quota after login with GitHub by @anastasiya1155 in #1151
• Capture migration errors in Sentry by @calyptobai in #1156
• Run tests on GitHub runners by @ggordonhall in #1158
• Bump tiktoken-rs version by @ggordonhall in #1159
• Hybrid search (lexical + semantic) merged with RRF by @rmuller-ml in #1154
• bump ort to latest by @nerdypepper in #1140
• Revert change to Qdrant startup wait time by @ggordonhall in #1155
• Improve autocomplete input by @anastasiya1155 in #1152
• Save user preferred font size in local storage by @anastasiya1155 in #1162
• Update answer and studio to gpt4-turbo 24k by @rmuller-ml in #1160
• Deduplicate after merging lexical results by @ggordonhall in #1161
• Assorted search fixes by @ggordonhall in #1165
• Filter out noisy chunks by @ggordonhall in #1163
• Fix folder mentions in chat by @anastasiya1155 in #1169
• Kill qdrant on supervisor drop by @calyptobai in #1166
• Bump version by @ggordonhall in #1171
• Update semantic log levels by @ggordonhall in #1168
• Update Qdrant to 1.7.1 by @ggordonhall in #1170
• Fix build-time TS error by @anastasiya1155 in #1172
• Proxy isPastDue from quota API by @calyptobai in #1167
• Show warning message when subscription is past due by @anastasiya1155 in #1173
• Mark qdrant for x86 macOS executable by @calyptobai in #1175
• Fix heap out of memory in CI by @anastasiya1155 in #1176
• Use gpt-4 model for answer generation by @anastasiya1155 in #1178
• Add additional llm_gateway error types by @ggordonhall in #1177
• Remove unused summary/conclusion logic by @ggordonhall in #1179

Full Changelog: v0.5.10...v0.5.11