Bug List

globalScope> Global bt_close (void *)

It takes whopping 3 seconds for the COM port to be reclaimable. If opened before that, the device is unusable for a minute.

globalScope> Global bt_close (void *)

It takes whopping 3 seconds for the COM port to be reclaimable. If opened before that, the device is unusable for a minute.

globalScope> Global bt_error (void *device)

it's useless

globalScope> Global bt_error (void *device)

it's useless

globalScope> Global bt_read (void *handle, u8 *buf, size_t count, int milliseconds)

milliseconds is ignored

globalScope> Global bt_read (void *device, u8 *buf, size_t count, int milliseconds)

the milliseconds part needs to be tested more throughly

See also

implementation at bt-win.c and bt-unix.c

globalScope> Global bt_read (void *fd_, u8 *buf, size_t count, int milliseconds)

the milliseconds part needs to be tested more throughly

globalScope> Global bt_write (void *handle, const u8 *buf, size_t count)

the first byte is omitted for compatiblity with the leading report byte demanded by hid_write. Wrapping HIDAPI could fix this.

globalScope> Global bt_write (void *fd_, const u8 *buf, size_t count)

the first byte is omitted for compatiblity with the leading report byte demanded by hid_write. Wrapping HIDAPI could fix this.

globalScope> Global bt_write (void *device, const u8 *buf, size_t count)

the first byte is omitted for compatiblity with the leading report byte demanded by hid_write. Wrapping HIDAPI could fix this.

See also

implementation at bt-win.c and bt-unix.c

globalScope> Global info (const char *arg)

needs more detailed information for bluetooth/wlan should print COM port, device name

File packets.h

A more standard approach like libpack might be worthwhile.

globalScope> Global PREFIX_SIZE

hidLayer shouldn't be maintained here as it's non-sense to demand bluetooth to account for it Maybe wrap HIDAPI with a wrapper that prefixes zero automatically, instead of having bt_write() omit the first byte

globalScope> Global tcp_error (void *fd_)

it's useless. Could use wprintf and strerror

globalScope> Global tcp_error (void *device)

it's useless

globalScope> Global tcp_info (void *device)

it's useless

globalScope> Global tcp_read )(void *device, u8 *buf, size_t count, int milliseconds)

the milliseconds part needs to be tested more throughly

See also

implementation at bt-win.c and bt-unix.c

globalScope> Global tcp_write )(void *device, const u8 *buf, size_t count)

the first byte is omitted for compatiblity with the leading report byte demanded by hid_write. Wrapping HIDAPI could fix this.

See also

implementation at bt-win.c and bt-unix.c

globalScope> Global up (FILE *fp, const char *dst)

might not handle files bigger than 2gb :-)

globalScope> Global up (FILE *fp, const char *dst)

might not handle files bigger than 2gb :-)