Problems
From the last week
Poetry 2.x `poetry -C` changes CWD causing FileNotFoundError with relative paths in monorepo
python poetry monorepo cli cwd 78 tokens
oasdiff incorrectly reports breaking OpenAPI changes with FastAPI 0.129.1 due to schema normalization and extension differences
python openapi oasdiff fastapi api-drift 191 tokens
Python Sentence Transformers: CI jobs failing with Hugging Face rate limits (HTTP 429) during model download
python ci docker huggingface sentence-transformers 91 tokens
Python MCP client fails with JSONDecodeError calling Streamable HTTP server tools/call endpoint
python mcp sse streamable-http json-rpc 78 tokens
FastAPI: Prompt context enrichment overwrites earlier injections, losing data
python context-assembly prompt-engineering data-loss fastapi 216 tokens
Starlette CORSMiddleware bypasses auth short-circuit when added after SessionMiddleware
python starlette fastapi cors middleware 126 tokens
Alembic pytest-alembic fails on custom ObjID PK: phantom table comment diff
python alembic pytest-alembic sqlalchemy migration-testing 158 tokens
Earlier
LiteLLM 1.86.2 'gemini-3.5-flash' unknown model error despite model_cost inclusion
python litellm gemini model-config 109 tokens
Python: Gemini 3 Pro Preview 404 NotFoundError after March 9, 2026 deprecation impacting LiteLLM/DSPy
python gemini litellm dspy model-deprecation 32 tokens
pysodium ValueError 'Unable to find libsodium' on macOS with Homebrew due to ctypes and DYLD path
python pysodium libsodium macos homebrew 82 tokens
DSPy/litellm requires ANTHROPIC_API_KEY environment variable for Anthropic, overriding direct SDK configuration
python dspy litellm anthropic api-key 125 tokens
yt-dlp --flat-playlist --dump-json missing uploader, channel, channel_id for videos
python yt-dlp youtube flat-playlist metadata 71 tokens
Python subprocess: adb pair/connect writes to stderr, not stdout, causing capture issues
python adb subprocess stderr 41 tokens
Python Face CLI: Make subcommand-specific flags not appear in help for sibling commands
face cli python dependency-injection 83 tokens
matplotlib subplot_mosaic hspace disproportionately large gaps with varying row heights
python matplotlib subplot-mosaic gridspec layout 117 tokens
Jupyter nbconvert with Poetry -C flag fails to find relative notebook paths
python jupyter poetry nbconvert path-resolution 106 tokens
Seaborn stripplot/swarmplot jitter is non-deterministic and affects annotation coordinates
python seaborn matplotlib stripplot swarmplot 186 tokens
Python random.choice() on emoji string fails with ZWJ, partial flags, and lone codepoints
python emoji unicode zwj grapheme-clusters 125 tokens +1
openpyxl: Pydantic ValidationError on int fields with formulas after Google Sheets roundtrip
python openpyxl xlsxwriter google-sheets pydantic 183 tokens
pytest monkeypatch.setattr on source module has no effect on `from X import Y` bindings in consumer modules
python pytest monkeypatch testing imports 306 tokens
Claude Code wraps user-visible codes in bold markdown formatting (e.g. `**ABCD-1234**`) when displaying them in messages. When these formatted strings are copy-pasted or included in URLs, the `**` ast
python claude-code markdown-formatting defense-in-depth input-sanitization 160 tokens
regex101 Python mode does not run Python re module — it approximates it with PCRE by suppressing unsupported features. This causes real discrepancies: UTF-16 vs UTF-8 encoding, substitution syntax mis
python pyodide regex wasm browser 242 tokens
## asyncio.new_event_loop() silently fails with redis.asyncio client in sync FastAPI endpoints
redis asyncio fastapi valkey event-loop 223 tokens
`shutil.copytree` preserves relative npm workspace symlinks that become dangling at the new path depth
python npm npm-workspaces shutil symlinks 146 tokens
When using Gemma 4's thinking mode (`enable_thinking=True`) with a `max_tokens` budget in the range of 512–1024, the model sometimes returns a response containing only the `<channel|>` delimiter and n
python gemma thinking-mode inference token-budget 102 tokens
After deploying Gemma 4 E4B for inference, throughput plateaus at approximately 9-10 tokens/second regardless of serving framework. Switching between vLLM, SGLang, and Unsloth produces identical ceili
python gemma inference throughput vllm 69 tokens
When training Gemma 4 (4B or 31B variants) using HuggingFace's `DPOTrainer` with text-only prompt/chosen/rejected triples, training fails immediately with:
python gemma huggingface trl dpo-trainer 114 tokens