Previously the "Assign Spool" button only appeared on configured slots,
forcing users to manually configure first — redundant since assignment
auto-configures the slot. Now shows on empty slots too. Also fixed the
AMS hover card showing generic material type instead of the spool's
slicer preset name after assignment.
- Sanitize project notes with DOMPurify before rendering via
dangerouslySetInnerHTML (ProjectDetailPage.tsx)
- Replace hand-rolled HTML sanitizer with DOMPurify in ProjectPageModal
to prevent attribute injection via crafted 3MF href values
- Block /api/v1/auth/setup when auth is already enabled to prevent
unauthenticated clients from disabling authentication remotely
The Debian ffmpeg package uses GnuTLS, whose hardened defaults reject
TLS renegotiation and legacy ciphers that some Bambu printer firmwares
(notably P2S) rely on — causing RTSP sessions to drop after a few
seconds.
Add a local TLS termination proxy (Python ssl/OpenSSL) that handles
the TLS connection to the printer and exposes a plain RTSP port to
ffmpeg. The proxy rewrites RTSP request-line URLs (rtsp://proxy →
rtsps://printer) while preserving Authorization headers so Digest
auth hashes remain valid.
Also:
- Reduce RTSP reconnect delay from 1.0s to 0.2s
- Add ffmpeg fast-start flags (-probesize 32, -analyzeduration 0,
-fflags nobuffer, -flags low_delay)
- Fix external camera double rate-limiting causing choppy streams
- Apply TLS proxy to external camera rtsps:// URLs and snapshot capture
- Update orphan ffmpeg cleanup to match rtsp:// (proxied) URLs
- Add unit tests for RTSP URL rewriting and proxy lifecycle
The Bambu Cloud API returns the base filament_id for versioned
setting IDs (e.g. GFSL99 → GFL99 for all "Generic PLA" variants),
so assigning a spool with a specific variant like "Generic PLA Silk"
(GFSL99_01) would configure the AMS slot with the base "Generic PLA"
profile (GFL99) instead of the correct one (GFL96).
Added a post-resolution cross-check: if the resolved filament_id maps
to a different name than the spool's stored preset name, reverse-lookup
the correct filament_id from the built-in filament table.
Printer File Manager Doesn't Auto-Refresh ([#704](https://github.com/maziggy/bambuddy/issues/704)) — The printer file manager (SD card browser) only fetched the file list once when opened. Files uploaded from BambuStudio/OrcaSlicer while the modal was open wouldn't appear until the user clicked the refresh button or reopened the modal. Now auto-refreshes every 30 seconds while open. Reported by @shadowjig.
Database Connection Pool Exhaustion Under Load ([#704](https://github.com/maziggy/bambuddy/issues/704)) — Background tasks (print scheduler FTP uploads, camera captures, notification sends, timelapse stitching) held database sessions open during slow network I/O, consuming connection pool slots for seconds at a time. With the default pool of 15 connections (size 5 + overflow 10), concurrent operations during print start/complete events could exhaust the pool, causing `QueuePool limit reached` errors and `greenlet_spawn` failures in RFID spool auto-assignment. Doubled the pool to 30 connections (size 10 + overflow 20). Reported by @shadowjig.
RTSP stream URLs (rtsps://bblp:<code>@<ip>:322/...) were not covered
by the credential sanitizer, leaking access codes in support bundles
and bug report logs. Extended the URL regex to match rtsps:// and added
access codes to the sensitive string collection for exact-match
redaction in both export paths.
Ambient drying: automatically dry filament on idle printers when
humidity exceeds threshold, regardless of queue state. Separate toggle
from queue auto-drying — both can run simultaneously. Uses the same
presets, humidity threshold, and power constraint detection.
Fix: block mode (wait for drying) previously skipped the humidity
auto-stop check for already-drying printers, causing drying to
continue indefinitely. Now only prevents starting new drying.
Queue auto-drying: scheduler automatically starts drying on idle printers
with scheduled queue prints when AMS humidity exceeds the configured
threshold. Uses conservative parameters (lowest temp, longest duration)
for mixed filaments. Drying stops when humidity drops below threshold
(30-minute minimum prevents oscillation), when scheduled items are
removed, or when the feature is disabled. Optional "block queue" mode
delays the next print until drying completes.
Configurable presets: temperature and duration per filament type,
editable in Settings → Print Queue, used by both manual drying popover
and queue auto-drying. Separate presets for AMS 2 Pro (n3f) and AMS-HT
(n3s) reflecting different heating capabilities.
PSU detection: drying button disabled with tooltip when dry_sf_reason
indicates insufficient power. Parses drying status bits and dry_sf_reason
from AMS info hex string via MQTT.
Backend: print_scheduler.py (+316 lines), bambu_mqtt.py, printer_manager,
schemas, settings route. Frontend: PrintersPage drying presets prop,
SettingsPage drying config UI, i18n (7 locales). Tests: 27 new tests in
test_scheduler_auto_drying.py covering conservative params, presets,
state sync, stop logic, minimum drying time, and auto-stop regressions.
Fix P2S camera stream dropping and snapshot capture race (#661)
P2S firmware's TLS renegotiation is rejected by Debian's hardened GnuTLS
defaults, causing ffmpeg RTSP sessions to drop after ~3 seconds. Add
GnuTLS config allowing unsafe renegotiation and legacy ciphers. Also add
ffmpeg fast-start flags, reduce reconnect delay from 1.0s to 0.2s,
remove double rate-limiting on external camera streams, and fix orphan
cleanup killing snapshot capture ffmpeg processes (exit code -9).
Or as a single combined commit:
Fix P2S camera streaming, snapshot race, and energy stats (#661, #695)
Camera: P2S firmware's TLS renegotiation rejected by Debian's hardened
GnuTLS defaults, dropping RTSP sessions after ~3s. Add GnuTLS compat
config, ffmpeg fast-start flags, reduce reconnect delay to 0.2s, remove
external camera double rate-limiting, and register snapshot ffmpeg PIDs
with the orphan tracker to prevent SIGKILL during capture.
In "total" energy tracking mode, the stats endpoint queried smart plug
lifetime counters which can't be filtered by date range. Energy costs
and kWh stayed the same regardless of timeframe selection. Fall back to
per-print archive data when date filters are active.
parse_version() misclassified "0.2.2b4-daily.20260313" as a release
because the daily suffix made the last dot-segment ("20260313") contain
no alpha chars, bypassing prerelease detection. Strip -daily.YYYYMMDD
suffix before parsing so daily builds compare as their base beta version.
The migration that removes the UNIQUE constraint on smart_plugs.printer_id
used an exact substring match ("printer_id INTEGER UNIQUE") to detect the
constraint. Databases created with older SQLAlchemy versions may express
the constraint differently (quoted column names, table-level UNIQUE clause,
or separate UNIQUE indexes), causing the migration to silently skip.
Users hit "IntegrityError: UNIQUE constraint failed: smart_plugs.printer_id"
when assigning a second HA switch to a printer.
Replace the exact string match with regex pattern matching that handles
inline constraints, table-level UNIQUE(printer_id), quoted column names,
and standalone UNIQUE indexes.
Printers send many undocumented HMS codes (e.g. 0501_0002, 0501_0004,
0300_0001) that don't correspond to real errors — calibration status,
firmware state, etc. The frontend already filters these out (only shows
codes in ERROR_DESCRIPTIONS), but the backend notified for all codes
with severity >= 2, causing spurious email/push notifications on every
backend restart when a printer has active phantom codes.
Flip the logic: only send notifications for HMS codes with a known
description in the backend's HMS_ERROR_DESCRIPTIONS dictionary. Also
fix the log message to report the actual number of notifications sent,
not the pre-filter count.
Non-script HA entities (switch/light/input_boolean) on the printer card
always sent turn_on when clicked, which is a no-op when the switch is
already on. Now sends toggle for non-script entities so the badge click
actually changes the switch state. Script entities still use turn_on.
Also fix SmartPlugManager._get_plug_for_printer() using
scalar_one_or_none() which crashes with MultipleResultsFound when
multiple plugs are assigned to the same printer (e.g., Tasmota + HA
switch). Now fetches all plugs and returns the main non-script plug.
Add tests for both fixes: _get_plug_for_printer multi-plug handling
(5 tests) and canonicalFilamentType/filamentTypesCompatible/autoMatch
PA-CF equivalence (8 tests).
Bambu Lab firmware treats PA-CF, PA12-CF, and PAHT-CF as interchangeable,
but the print scheduler and filament override UI used exact string matching.
A 3MF requiring PA-CF wouldn't match PA12-CF in the AMS, leaving the
override dropdown disabled and preventing scheduler assignment.
Add a filament type equivalence system (both backend and frontend) so
these PA variants are treated as compatible in scheduler assignment,
AMS slot matching, force color match validation, and the override dropdown.
Also fix the "Force color match" checkbox click target being too wide
(entire row instead of just the checkbox/label) by changing flex to
inline-flex on the label element.
The embedded camera viewer only handled mouse events (mousedown,
mousemove, mouseup) for window dragging and resizing. On iOS/iPadOS,
touch input doesn't trigger mouse events, so the camera window
couldn't be repositioned — touch just scrolled the page underneath.
Add touch event handlers (touchstart, touchmove, touchend, touchcancel)
alongside mouse handlers for both the header drag handle and the
bottom-right resize handle. Uses preventDefault on touchmove to
prevent page scrolling during drag.
When all higher-priority slot-to-tray mapping methods failed (MQTT
mapping, print command mapping, queue mapping, color matching), the
internal usage tracker fell back to slot_id - 1 which can never reach
external spool IDs (254/255) or AMS-HT IDs (128+). Added position-based
resolution using sorted available tray IDs, matching the Spoolman fix
from PR #686.
When inventory spool assignments had stale weight_used values, the fill
level fallback chain (Spoolman → Inventory → AMS remain) used nullish
coalescing (??), which doesn't fall through on 0. A stale inventory
fill of 0% permanently shadowed the correct real-time AMS remain value.
Viewer users were unaffected because their group lacked the
inventory:view_assignments permission, so the query never fired.
Now when inventory says 0% but AMS hardware reports positive remain,
the inventory value is bypassed in favor of the live AMS data.
Notify users when the first layer finishes printing so they can check
adhesion remotely. Triggers once per print when layer 2 begins
(layer_num >= 2, capped at <= 5 to handle reconnects). Includes a
camera snapshot attachment. Adds the on_first_layer_complete toggle
to all notification providers, with backend/frontend/i18n support
across all 7 locales.
When creating a virtual printer in proxy mode, the model was always set
to X1C because the frontend hides the model dropdown and the backend
used a hardcoded default. Now auto-inherits the model from the target
printer in proxy mode — on create, on target printer change, and on
mode switch to proxy.
Start, monitor, and stop drying sessions for AMS 2 Pro (n3f) and
AMS-HT (n3s) directly from the Printers page. Flame icon in AMS card
header opens a popover with filament-based temperature/duration presets
from BambuStudio. Live countdown status bar shows time remaining.
Backend: cache module_type from MQTT get_version, expose dry_time and
module_type in both WebSocket and REST paths, add supports_drying()
with firmware gating, add server-side guard on start_drying endpoint.
Frontend: drying button, popover with filament select + temp/duration
sliders, status bar, start/stop mutations. i18n for all 7 locales.
Supported: X1/X1C (fw 01.09+), P1P/P1S (fw 01.08+), H2D (fw 01.02.30+),
H2D Pro, X1E. Not supported: P2S, A1, A1 Mini, H2S, H2C.
Cloud credentials were stored globally — one Bambu Cloud account per
Bambuddy instance. When auth was enabled, any user logging into Cloud
overwrote everyone else's credentials. Credentials are now stored
per-user: each user gets their own independent Cloud login.
Also fixed cloud data endpoints (settings, fields, preset CRUD)
requiring settings:read/settings:update permissions instead of
cloud:auth — users who had "Cloud Auth" enabled but "Settings"
disabled couldn't load profiles after logging in.
The P2S firmware drops RTSP sessions after a few seconds with an I/O
error. The backend treated this as fatal, ending the MJPEG stream and
forcing the frontend through a full reconnection cycle. Added transparent
auto-reconnection: when ffmpeg's RTSP connection dies, it respawns
immediately and continues streaming MJPEG frames to the browser with
only a brief freeze (~1s). Up to 30 reconnections before giving up.
The Docker image (python:3.13-slim) didn't include iproute2, so the
`ip` command wasn't available. The code fell back to ioctl-based
enumeration which can only return one IP per interface — aliases like
eth0:1 were completely invisible. Added iproute2 to the Dockerfile.
The compatible_printers filter on local presets parsed the stored JSON
array (e.g. '["BBL X1C 0.4 nozzle"]') as a semicolon-delimited string,
so matching always failed and every local preset was silently skipped.
Removed the filter entirely — user-imported profiles should be usable
on any printer.
Three bugs: (1) home_flag bit 18 is set on all printers regardless
of connection type, so every ethernet-capable model showed the badge
even on WiFi. Replaced with wifi_signal heuristic — ethernet printers
with WiFi disabled report a hardcoded -90 dBm sentinel. (2) The lazy
import used `from app.utils.printer_models` which crashes with
ModuleNotFoundError in paho-mqtt's background thread (correct path:
backend.app.utils.printer_models). This killed the thread, causing
all printers to go stale after 60s and disconnect/reconnect in a
loop. (3) WiFi-only models excluded via model-based gating.
home_flag bit 18 is set on all printers regardless of hardware,
causing the ethernet badge to appear on WiFi-only models (A1, P1P,
etc.). The previous fix removed the feature entirely, so ethernet
printers lost their badge too.
Now only trusts bit 18 on models with an ethernet port (X1C, X1E,
P1S, P2S, H2D, H2D Pro, H2C, H2S). WiFi-only models always show
the WiFi signal badge.
The GET /api/v1/support/debug-logging endpoint returned 500 when the
database contained a timezone-aware enabled_at timestamp written by
0.2.2b3. The duration calculation mixed offset-aware and offset-naive
datetimes, raising TypeError. Strip tzinfo when reading the stored value.
Sends persistent notifications to the HA dashboard using the existing
HA connection from Settings. Zero config — just select "Home Assistant"
as provider type. Users can forward notifications to mobile via HA
automations.
Two bugs prevented bed cooldown notifications from working:
1. Stale temperature data: After print completion, the printer sends
partial MQTT updates without bed_temper, leaving the cached value
frozen at the end-of-print temperature. The monitor polled for 30
minutes seeing the stale value above threshold, then timed out.
Fix: send periodic pushall commands to force fresh temperature data.
2. Missing fields in provider create endpoint: on_bed_cooled and all 7
queue event toggles were omitted from the NotificationProvider
constructor, silently discarding user selections on create (update
worked fine via dynamic setattr).
Also added debug logging to the bed cooldown polling loop.
maziggy