android-llmservice: feature-roadmap TODO + llama.cpp b9941 + Kotlin build fix & installable APK#312
Merged
Conversation
… effort/feasibility) Map the 'Local AI / Offline LLM App' design brief onto what the llama-android AAR + llama-kotlin facade make possible today: per-feature purpose, t-shirt effort, status (done/partial/todo/blocked/out-of-scope), and notes. Flags the two gating decisions — (1) the OpenAI/Ollama API server is blocked on Android (needs an in-app Ktor/NanoHTTPD layer over the JNI core) and (2) model download needs the INTERNET permission (a change from the current zero-network posture) — plus a phased sequencing. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01HTS5FpMtBLJENTGoRUrp5m
Single-commit upstream range touching only ggml/src/ggml-cuda/mmvq.cu (a CUDA quantized mat-vec kernel fix) — no patch-target file (common/arg.*, tools/server/*) and no OuteTTS generator anchor (tools/tts/tts.cpp) changed, confirmed via the b9940...b9941 compare diff, so all eight patches (0001-0008) apply unchanged. No project-facing API change. Updated GIT_TAG + LLAMA_TAG (CMakeLists), README badge, CLAUDE.md pinned-version + prose references, and appended the two history rows. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01HTS5FpMtBLJENTGoRUrp5m
…lish installable release APK CI build-android-llmservice got past the Gradle-config fix and failed at :app:compileReleaseKotlin: Languages.kt had \"values-*/\" inside a KDoc block, whose \"*/\" closed the comment early so the rest parsed as code (\"Expecting a top level declaration\"). Reworded to \"values-*\" (backticked) — no stray */. (Checked all .kt: the other */ are single-line KDocs or the \"*/*\" MIME string, both fine.) Also answer \"where do I download the APK\": the job only uploaded the .aab (Play-only, not installable). Now also assembleRelease -> upload app-release.apk as the android-llmservice-apk artifact (sideloadable via adb install; debug-signed unless the upload-key secrets are set), downloadable from each run independent of any release. README updated. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01HTS5FpMtBLJENTGoRUrp5m
bernardladenthin
had a problem deploying
to
maven-central
GitHub Actions
Failure
bernardladenthin
had a problem deploying
to
startgate
GitHub Actions
Error
bernardladenthin
had a problem deploying
to
maven-central
GitHub Actions
Failure
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
android-llmservice/TODO.md— a feature roadmap mapping the "Local AI / Offline LLM App" design brief onto what thellama-androidAAR +llama-kotlinfaçade make possible: per-feature purpose, t-shirt effort, status, and notes. Flags the two gating decisions (the OpenAI/Ollama API server is blocked on Android — needs an in-app Ktor/NanoHTTPD layer over the JNI core; and model download needs theINTERNETpermission, a change from the current zero-network posture) plus a phased sequencing.b9940→b9941. Single-commit upstream range touching onlyggml/src/ggml-cuda/mmvq.cu(a CUDA quantized mat-vec kernel fix) — confirmed via theb9940...b9941compare diff. No patch-target file (common/arg.*,tools/server/*,tools/tts/tts.cpp) changed, so all eight patches (0001–0008) apply unchanged; no project-facing API change. UpdatedGIT_TAG+LLAMA_TAG(CMakeLists), README badge, CLAUDE.md pinned-version + prose refs, and the two history rows.build-android-llmserviceKotlin compile failure that surfaced onmainafter fix(android-llmservice): migrate Kotlin jvmTarget to compilerOptions DSL; Gradle 8.14.4; allowBackup=false #311:Languages.kthadvalues-*/inside a KDoc, whose*/closed the comment early (25× "Expecting a top level declaration"). Reworded to a backtickedvalues-*..apk. The job now also runsassembleReleaseand uploadsapp-release.apkas theandroid-llmservice-apkartifact — downloadable from each run's Artifacts (independent of any release) and sideloadable viaadb install(debug-signed unless the upload-key secrets are set). The existing.aabis Play-only / not directly installable.Test plan
build-android-llmservice) validates the Kotlin build fix, the release-APK build, and the on-device emulator UI test.Related issues / PRs
Refs #310 (introduced
android-llmservice) and #311 (Gradle-config build fix). This PR fixes the Kotlin compile error that only surfaced once #311's config fix let the build reachcompileReleaseKotlinonmain.Checklist
CONTRIBUTING.mdandCODE_OF_CONDUCT.md🤖 Generated with Claude Code
https://claude.ai/code/session_01HTS5FpMtBLJENTGoRUrp5m
Generated by Claude Code