XDJ-700 1.15 VBR upload regression fixture

Source: https://github.com/chrisle/rbxport/issues/43#issuecomment-6099327629
Capture: packet_capture_xdj700_e8306_no_file_rbxport_1.5.0.pcap.gz
The .bin contains only the client database request in TCP frames 3896 and
3897, concatenated in order. It contains no audio or library path.

[OBS, capture] Transaction 0x0580002d, request 0x2205, arguments:
[0x02080401, 148716872, 1604, <1604-byte VBR blob>]. Frame 3899 is RBX's
0x4003 error with argument 0x2205. A separate report reproduction uses transaction 0x05800031 and records
the player clearing its loaded track. The reporter observed E-8306 and a blank browser.

[OBS, static] AlphaTheta docs platform/link-export-database-contracts.md
records rekordbox 7.2.11 PSvDBMain::OnWriteCmd at 0x1017509d4 and
SavVbrInf: request 0x2205 is accepted without storage or a reply.
RBX previously routed it to the unsupported-command error branch.

[ASSUME] That unsolicited error is consumed by the next browse read, ending
the result list. The timing and firmware control flow support this hypothesis;
hardware testing of the fix must confirm the reported symptom is resolved.
The fix matches the inspected vendor's silent handling, without inventing
a VBR persistence format or acknowledgement.

Firmware investigation:
XDJ700.UPD 1.15 SHA-256:
73edec9802da51672257c2599efc04209dc92478fcbaa1a0425b3b122e33f99c
[OBS] The update carries separate MAIN and PANEL S-record streams, with
valid record lengths/checksums. MAIN's entry record names 0xa0000000;
the entry bytes decode as little-endian SuperH instructions. The standard
Linux UPD unpacker cannot read this packaging.
[OBS, static] The corrected decoder uses a 4078-byte space prefix in its
4096-byte dictionary. Both compressed-block checksums pass. The recovery and
all 18,601,864 application bytes match instruction execution of the firmware
decoder exactly. dbcl_RegVBRInfo at physical 0x08b76b3e sends 0x2205 without
waiting for a response. Instruction execution of dbcl_GetNextRecord at
0x08b74474 shows an unexpected 0x4003 returns 0
(no more rows) after one receive and logs an unsupported command; a 0x4101
control returns 1. These tests substitute mailbox and allocation helpers.
[OBS, static] The decoded-message callback at 0x08b72006 routes both captured
transaction IDs (0x0580002d and 0x0580002e) to mailbox 22. This was checked by
executing the callback instructions with synthetic messages and substituted
allocation and mailbox-send helpers.
[UNKNOWN] Exact mailbox consumption order and the E-8306 caller remain
untraced. No physical-player verification was performed.

Related older failure:
[OBS, issue #43] PR #260 fixed STALE refusals caused by the player's
20 trailing file-handle bytes, and the reporter confirmed successful track
loading on dev. That NFS fix is separate from this database-write mismatch.
