FIxes
Spoolman connection
-
Spoolman now works when your URL has a trailing slash. If your
Spoolman URL ended in a slash, FilaBridge built request paths like
//api/v1/spool, which Spoolman does not route to its API. It hands
those to the web UI instead and returns a 200, so you got
"invalid character '<'" errors and no spools. The connection test made
this harder to spot. It only looked at the status code, so it happily
passed against that same web page. Settings told you Spoolman was
connected while every real call failed. It now checks the response
is actually JSON. When something DOES come back as a web page,
the error now tells you to check your Spoolman URL instead of
showing you a JSON parser message. Confirmed against Spoolman
0.26.1. (fixes #50) -
Locations import again on Spoolman v0.26. FilaBridge read locations
from thelocationssetting, which the v0.26 client never writes, so
the Location Tags tab came up empty even though your spools had
locations. It now reads the location list directly, which is where
they actually live. Thanks to @mmydland for finding and fixing this.
(#44)
Printing
-
Slot picker for single-filament MMU prints. When a single-filament
print is sliced for a multi-toolhead printer, the file does not record
which slot it used, and PrusaLink does not report one either. You now
get a warning at the start of the print with a dropdown to pick the
slot, so usage lands on the right spool. (#36) -
Scan a toolhead tag first to unload it. Optional workflow where
scanning a toolhead's NFC tag on its own clears the spool loaded
there, instead of needing a spool scan first. (#42)
Interface
-
The redesigned interface is now the default. If it disrupts you,
FILABRIDGE_OLD_UI=true brings back the old one. It is a short term
options however and will be removed after a few releases. Only the
styling differs, the behavior and functionality is unchanged. -
Frames are centered again rather than pinned left, which was a
max-width typo.
Bambu Lab (still experimental, developer mode only)
-
Sliced file downloads now work on the A1 Mini. Filament usage
was never recorded on these printers because the download failed every
time. Their FTP server does not answer EPSV, so the connection falls
back to PASV, and the firmware replies with the address 0.0.0.0
instead of its own. FilaBridge now ignores that and uses the printer's
real address. -
Low filament warnings and auto pause now work for Bambu printers, not just
Prusa. Both printer types now run the same warning logic, driven
over MQTT for Bambu. -
A Bambu printer that is switched off now reads as offline. It used to keep
reporting whatever it was doing when you unplugged it, sometimes for
hours. If FilaBridge started while the printer was off it read as idle
instead. Bambu printers also connect at startup now, so they no longer
read offline for a poll interval after launch. -
Better capture tool for gathering printer data. The -bambu-probe flag
now collects everything in one run: full printer state from MQTT
including the AMS block, the SD card listings, job descriptor files,
the filament summary from each cached plate gcode, and slice_info from
the project file with every plate's weight. It works on an idle
printer and only reads, so it never starts, pauses or stops anything.
Known issues
-
AMS slot mapping is positional. FilaBridge assumes slicer filament 1
is AMS slot 1 and so on. If your slicer's filament order does not
match your AMS slot order, usage gets deducted from the wrong spool
with no warning. -
Multi-plate projects always read the first plate's numbers. On the A1
and P1 this means a wrong number when you print any plate other than
the first due to their lack of multi plate support. The X1 can print all plates
in one job, where only one plate of several would be counted at all on A & P. -
If you have a Bambu printer with an AMS attached, running -bambu-probe
and sending us the output is the most useful thing you can do to help.
AMS support is the last piece blocking a wider Bambu release.