<blogic>
enyc: they FOB china price is 89$ and that is fixed
<blogic>
then depending where you live there will be shipping and import tax
rsalvaterra has quit []
rsalvaterra has joined #openwrt-devel
<enyc>
blogic: hrrrrm yes and aliexpress nightmare :O
<mrkiko>
I am seeing a double mount of /boot in OpenWrt x86-64 qemu. Anyone noticed something like that?
<blogic>
enyc: they started poping up on amazon in various countries
<stintel>
mrkiko: seeing it too
<dvn>
enyc: aliexpress is nice because you're buying close from the source
danitool has joined #openwrt-devel
n3ph has quit [Ping timeout: 480 seconds]
Starsfall has quit [Remote host closed the connection]
Starsfall has joined #openwrt-devel
<slh>
enyc: at least over here (.de), ordering from aliexpress (as long as you stay below 150 EUR, so the platform can take care of customs and taxes) is rather painless for the last ~2 years (before that customs/ taxes were wild-west, with misdeclarations' galore).
<enyc>
hrrm, alas here in UK eurgh mess
<ynezz>
I've sourced my last BPI-R4s from conrad.de, so maybe just write them about One?
<ynezz>
having better prices then on aliexpress and basically 2 days delivery to .cz
gromero has joined #openwrt-devel
<blogic>
I am considering to rename jffs2reset to factoryreset and then printing a deprecation warning when the old command is called
<blogic>
jffs2reset feels outdated
<KanjiMonster>
sounds reasonable
<ynezz>
I've never used jffs2reset, always using `firstboot -y`
<ynezz>
ok, I see now, it calls jffs2reset under the hood
<blogic>
ynezz: ok, let me craft a patch and push it
<blogic>
I also plan to move this repo to git.openwrt.org
<blogic>
and then send a patch adding the repo to feeds.conf.default
n3ph has joined #openwrt-devel
<KanjiMonster>
anyone ever dabbled in 802.1x and mac based authentication? I wonder how broadcasts (from other ports on the bridge) are handled for the latter case
<blogic>
so option auth 1 means that netifd tracks the port as port of a bridge but does not add it until it gets told so
<KanjiMonster>
new-ish linux supports locking ports via IFLA_BRPORT_LOCKED and IFLA_BRPORT_MAB (for mac based authentication)
<blogic>
also upon a carrier-down event netifd will autmatically remove the port again
<blogic>
ok, good to know
<KanjiMonster>
if you enable MAB, then you can have "locked" FDB entries (i.e. seen on a port, but not used for forwarding), so software listening knows which mac addresses are seen without having to enable traffic to cpu
<blogic>
oh ok
<blogic>
which would also make radius based mac-auth easier to do
<KanjiMonster>
LOCKED and MAB is also wired up in switchdev
<ynezz>
blogic: BTW maybe today it would make more sense to consider using github.com/openwrt/uconfig as origin and Git mirror it to git.o.o/feed/uconfig, as its being done with other feeds?
<blogic>
ynezz: also works for me
khrbtxyz has quit [Ping timeout: 480 seconds]
<blogic>
ynezz: also less stress on the owrt git server
<blogic>
or pwd protect gitweb and post the credentials on the website
<tmn505>
unconventional, maybe worth a try
danitool has joined #openwrt-devel
<ynezz>
that bunch of 500 was caused probably by some bug in gitweb or such
<ynezz>
accept4() failed (24: Too many open files)
<ynezz>
and we were only rejecting ';sf=tgz' requests, so '&sf=tgz' would be still possible, changed that, lets see how long it takes to another 500s
<ynezz>
its really like someone is doing complete internet backup /?a=blob&f=target%2Flinux%2Fs3c24xx%2Fconfig-2.6.30&h=30cb86d14e2803164fcefb303810d147571b442a&p=openwrt%2Fsvn-archive%2Farchive.git