if this option is used and the ESSID passes a UTF-8 validity check,
the ESSID will be displayed unmodified. this is potentially dangerous
on terminals not supporting UTF-8 when an attacker embeds terminal
escape sequences.
closes#207closes#206closes#205
output can be written to file using standard shell means:
e.g. wash > file.log
or wash | tee -a aps.log
having the option out of the way, allows us flexible use of stderr
and stdout, as requirements dictate.
This reverts commit f0b8b1dc862519cd6a68d870134928a4a4b0ce67.
This broke the broken driver on Mac OS X. basically having the driver
signal "bad fcs" means: "do not process, broken!", rather than
"the FCS checksum is bad, so ignore it".
git is meant to be used with sources, not binary products generated from them.
this makes it possible to actually look at the contents of the
(horribly outdated) manpage.
wash historically used -C for the --ignore-fcs option, but reaver uses
that letter already for other purposes. we now reintroduce the option
(removed in 31cd6da1c4ed55163443ea5da923b7fbb3e94f5d under wrong premises[0])
using the letter -F, so both reaver and wash can use the same.
note that reaver in the past did not even have a way to ignore FCS validation.
this fixes issues with the wrong checksum produced by the RTL8812AU
driver from https://github.com/aircrack-ng/rtl8812au .
note this only needed for drivers that are broken (i.e. it won't happen with
any official drivers).
closes#193
[0]: the code doing the FCS validation was buggy and produced false positives.
this caused the terminal to be spammed with warnings.
it was fixed in 1381ee2d9b544cfdcb25a3bc77b9b0242668fe4a
Added vendor OUI 00:09:86 (MetaLink) to detect properly Lantiq WiFi chips such as the "Xwave 300 Lantiq psb8231 11 bgn" from arcadyan ARV7519RW22 (livebox 2.1).
FYI MetaLink was bought by Lantiq in 2010 which explains why they have the same OUI
since SQLite support was removed, this option doesn't make any sense
anymore. with sqlite, a user could daemonize (i.e. send to background)
both reaver and wash, and the scan results from wash and an eventually
suceeding pin try from reaver would probably have made its way into the
database (but there didnt exist any way to see those results, without
resorting to the sqlite command line client...).
so without anything to back up the results automatically, losing the
text output of both reaver and wash defeats the entire purpose.
for people wanting to use the programs in the background, it is instead
recommended to use them with a terminal multiplexer such as GNU screen
or tmux, or via a script that redirects output to a file, and then
detaches from the controlling terminal...
even though usage of airmon-ng is not strictly necessary to put the
wlan interface into monitor mode (i personally use macgeiger[0] for that),
its naming convention of the renamed wireless device has changed from
monX to wlanXmon in the past so let's use the updated syntax...
[0]: https://github.com/rofl0r/macgeiger
i keep getting reports of people who are not capable of reading README
which states that ./configure must be used, and then get a build error
because they miss config.h.
since 10e76373a85a38b469ef59957a6953f87064a3a0 it is no longer needed
anyway, the only thing we took from config.h was the version numer
passed in the macro PACKAGE_VERSION.
the resend logic introduced in 569cc227a624a2ff00002f44d910f385fa73f370
should make it superfluous. i leave the functionality intact for the
time being.
both programs use 95% of the same object code, i.e. they're mostly
equivalent. that means merging them into a single binary will save
almost 100% harddisk space.
reaver is now the main binary, while wash is a symlink to it.
main.c inspects argv[0] and then decides which program to run based
on the name the binary was started as.
closes#168
1) make some functions in 80211.c static
2) fix warning about redefinition of IFNAMSIZ
3) remove header inclusion bloat
4) 80211.c: rename u_char to unsigned char
