**** BEGIN LOGGING AT Wed Apr 10 02:59:57 2019
Apr 10 04:36:38 <nyt> send an updated patch to nbd
Apr 10 04:36:43 <nyt> seems to work
Apr 10 04:41:45 <mangix> gch981213: any idea on controlling ethernet LEDs on ath79?
Apr 10 04:42:40 <nyt> https://pastebin.com/Wuu8HUjc
Apr 10 04:52:19 <gch981213> mangix: Ah... What do you mean exactly?
Apr 10 04:57:40 <gch981213> mangix: If you want to use a GPIO led for ethernet port on switch, just use swconfig led trigger. Alternatively, if that ethernet port is on the builtin switch for ar9344/qca953x/qca956x you can set pinmux register to make that GPIO pin act as switch led. (e.g. https://paste.ubuntu.com/p/tBW6qYVN22/ )
Apr 10 05:11:40 <mangix> gch981213: maybe. All I remember is on my archer c7, I was able to control the ethernet LEDs through LuCI
Apr 10 05:13:53 <mangix> in ar71xx that is
Apr 10 05:15:34 <gch981213> mangix: LEDs added by this one, right? https://github.com/openwrt/openwrt/blob/master/target/linux/ar71xx/files/arch/mips/ath79/mach-archer-c7.c#L144
Apr 10 05:17:01 <gch981213> mangix: There isn't a dts equivalent for that.
Apr 10 05:18:47 <mangix> yeah exactly
Apr 10 05:20:39 <mangix> whoa, this has to be a bug
Apr 10 05:21:04 <mangix> https://gist.github.com/neheb/4f26199722356be6b7966ee273579b95
Apr 10 05:24:36 <mangix> hmmm. can't figure out how to use swconfig to change LED state
Apr 10 05:28:19 <gch981213> mangix: In my previous reply I misunderstood your question :P
Apr 10 05:32:25 <mangix> well, all I'm looking to do is turn all the LEDs off. I can't turn off the power LED but good enough
Apr 10 05:36:40 <gch981213> mangix: If you don't mind to set it in dts and being unable to turn them back on unless you flash a new firmware, you can do this: https://paste.ubuntu.com/p/2fWBXf5F7v/
Apr 10 05:37:06 <pkgadd> d'oh, I finally got wpa3psk to work... I'm usually always using hex PSKs wherever possible, which is kind of a problem with wpa3 requiring a 128 hex values instead of just 64 (and wpa_supplicant 2.7 doesn't appear to accept a wpa3/ sae PSK with 128 hex values at all) works with a passphrase in all cases
Apr 10 05:37:33 <mangix> gch981213: :)
Apr 10 05:39:47 <abenz> hi. When flashing a fresh image, uhttpd/px5g generates a new self-signed certificate crt/key pair
Apr 10 05:40:58 <abenz> when one accesses one's router, they would need to add exception for self signed certificate. Now to avoid adding 2 sets (ie seeing browser warning twice), I figured I'd reuse the same uhttpd generated certificates for jsonrpc over https (aria2), which is running on port 6800
Apr 10 05:41:57 <abenz> so I converted the DER formatted crt to PEM, when checking them, I see that the two certs are identical (just one of them is binary)
Apr 10 05:42:27 <abenz> all seemed good, except I still can't use rpc over https unless I manually point my browser to https://192.168.1.1:6800
Apr 10 05:43:15 <abenz> upon doing that I get asked to add exception for the certificate (which I assumed would not be required since I'm using the same uhttpd cert I added earlier).. only after I add the cert to exceptions, I can use rpc over https
Apr 10 05:46:03 <abenz> Q: does anyone know a workaround besides disabling https redirect or using lets encrypt?
Apr 10 05:49:19 <mangix> gch981213: I'm guessing a future update to qca8k would make configuring the LEDs possible
Apr 10 05:49:52 <blogic> mangix: writing the code and sending a patch would do that
Apr 10 05:50:13 <mangix> or that :)
Apr 10 05:50:21 <blogic> only that
Apr 10 06:39:46 <abenz> blogic: I know patches welcome, I'm Just wondering if your interest shifted to something else or still working on it? :)
Apr 10 06:40:03 <abenz> dsa that is
Apr 10 06:44:16 <mangix> abenz: referencing ramips?
Apr 10 06:44:44 <mangix> if so, upstream got it working
Apr 10 06:45:04 <mangix> mt7621-eth got nuked from the staging directory
**** BEGIN LOGGING AT Wed Apr 10 06:45:24 2019
Apr 10 06:45:47 <mangix> I'm running the mainline driver on mt7621 actually. only problem is that it's slow
Apr 10 06:46:11 <abenz> mangix: how slow?
Apr 10 06:46:30 <mangix> around twice
Apr 10 06:46:52 <mangix> ~450mbps on iperf runs
Apr 10 06:47:31 <Rene__> mangix: from and to the device of NAT?
Apr 10 06:47:43 <Rene__> s/of/or/
Apr 10 06:47:55 <mangix> to and from. Have not tested routing
Apr 10 06:48:06 <Rene__> I get 940mbit
Apr 10 06:48:35 <abenz> you get 940mbit NAT with dsa?
Apr 10 06:48:57 <Rene__> from and to the device I get ~940mbit.
Apr 10 06:49:09 <mangix> Rene__: well, you have a lot of custom changes. I just backported patches
Apr 10 06:49:40 <abenz> if its from/to device doesn't that work as a normal switch/entirely in hardware?
Apr 10 06:50:08 <Rene__> mangix: I did not change the main stuff, phylink all about how to handle/setup GMAC and phy
Apr 10 06:50:11 <abenz> mangix: what device ?
Apr 10 06:50:50 <mangix> abenz: GnuBee PC1. mt7621 at 900MHz
Apr 10 06:51:10 <mangix> I bricked it trying to go to 1.1ghz
Apr 10 06:51:16 <mangix> and revived
Apr 10 06:51:41 <mangix> need to find the proper resistor to remove
Apr 10 06:52:24 <Rene__> mangix: main issue with backporting the mainline kernel is the other device like mt7620 and rt38xx they are not supported.
Apr 10 06:54:19 <mangix> yeah. i tried using the staging driver but gave up trying to get it to work.
Apr 10 06:54:34 <blogic> abenz: i am just busy with other stuff
Apr 10 06:56:28 <abenz> I see
Apr 10 06:57:39 <Rene__> I don't know if it can be an option for openwrt to have two mt7621 drivers. the old one but files should be renamed so they don't conflict with the mainline driver and the mainline/backported driver.
Apr 10 06:58:56 <mangix> Rene__: funny you mention that. it used to be named different. but there was a commit that matched the names
Apr 10 07:00:19 <blogic> what is a mt7621 driver ?
Apr 10 07:00:42 <Rene__> mt7621 ethernet and mt7530 switch driver
Apr 10 07:01:25 <blogic> there is no mt7621 ethernet driver
Apr 10 07:01:32 <blogic> there is a mtk_eth_soc driver
Apr 10 07:01:46 <Rene__> that is what I mean
Apr 10 07:01:50 <blogic> ok
Apr 10 07:01:56 <blogic> we can have 2 drivers in the tree
Apr 10 07:03:18 <Rene__> That would be great.
Apr 10 07:24:48 <mangix> wow
Apr 10 07:25:01 <mangix> i managed to boot openwrt with an arch linux kernel
Apr 10 07:25:35 <mangix> this makes no sense to me
Apr 10 07:26:38 <olmari> Is that good or bad? :D
Apr 10 07:27:02 <mangix> well, bad in this case
Apr 10 07:27:08 <mangix> i wanted to boot an initramfs
Apr 10 07:27:12 <mangix> *to an
Apr 10 07:28:20 <mangix> ugh I can't boot from USB on this device
Apr 10 07:38:53 <Tapper> ldir ping
Apr 10 08:18:25 <ynezz> blocktrron: https://git.openwrt.org/?p=openwrt/staging/ynezz.git;a=commit;h=8443c229f48332f9d56743f0fc6d6aff163619b6 can I get your Reviewed-by please?
Apr 10 08:20:12 <blocktrron> ynezz: are you sure you intended to highlight me? I never hears of this TP-Link router
Apr 10 08:23:12 <ynezz> blocktrron: yeah, sorry to bother you, but it seemed to me, that you simply have some knowledge about lantiq target and you're responsive here :)
Apr 10 08:23:32 <ynezz> but don't worry about it too much, it's not a big deal
Apr 10 08:24:20 <ldir> Tapper: pong
Apr 10 08:29:50 <blocktrron> as the userspace assignment seems to be gone (and this is the only issue) this should reactivae the led
Apr 10 08:30:32 <blocktrron> however, i'm for myself are a bit hesitant to state something is fixed unless i've tested it myself. However this change should not worsen things.
Apr 10 08:31:21 <ldir> kabooooooom :-)
Apr 10 08:57:24 <ynezz> blocktrron: yeah, I've asked for additional testing, but I'm not sure if the user could test the change himself, so that's better to ask for 2nd pair of eyes
Apr 10 08:57:53 <ynezz> then I'm going to sleep on it and push it eventually
Apr 10 09:17:18 <dansan> Is there a way to make more than one package out of a single build?  Like if I want to have a normal and -dev package, but I don't want to build everything twice?
Apr 10 09:55:40 <rmilecki> nyt: nice debugging work! thanks!
Apr 10 10:27:47 <Tapper> we got some spam on fly spray
Apr 10 10:40:55 <dansan> w00t!
Apr 10 11:06:07 <ldir> ynezz: ping
Apr 10 11:08:10 <ldir> ynezz: going back to last night & perf, curious what you meant by "I've removed the clashing module"
Apr 10 11:09:13 <ynezz> the one taking the irq 13
Apr 10 11:10:20 <ldir> ah, so USB didn't work...at the same time?
Apr 10 11:13:16 <ynezz> in my case it was gpio keys
Apr 10 11:14:21 <ldir> ah, that's interesting, must be dynamic/different for the v2 then as the interrupt is used by the ehci module on mine :-)
Apr 10 11:14:43 <ldir> gpio keys I can lose, usb I can't :-)
Apr 10 11:15:27 <ynezz> well, if you're perf tracing the usb
Apr 10 11:17:12 <ldir> perf needs to steal the interrupt before usb grabs it.... or share nicely
Apr 10 11:36:20 <rmilecki> nyt: interesting, i'm looking for some regression between 4.14 and 4.19 and I think it's also related to that netfilter patch
Apr 10 13:57:52 <ynezz> blogic: could you please give me your ACK for the last 4 commits https://github.com/ynezz/umbim/commits/staging ? Thanks
Apr 10 14:12:58 <brucethompson> Hello. It's Bruce Thompson. I am trying to finish the port of Python pynacl to OpenWrt and am running into build problems. I need some help getting to the bottom of the issue.
Apr 10 15:03:12 <brucethompson> Anyone there??
Apr 10 15:13:11 <valku> brucethompson: what are the build problems?
Apr 10 15:13:20 <brucethompson> Hi
Apr 10 15:13:45 <brucethompson> I have a dependancy on libsodium in my makefile
Apr 10 15:14:23 <brucethompson> The setup.py for nacl ends up executing a configuration script
Apr 10 15:15:21 <brucethompson> one second
Apr 10 15:16:14 <brucethompson> It's checking if the configuration is correct for libsodium.
Apr 10 15:17:43 <valku> I guess it finds that something is incorrect and complains, right? What are the error messages?
Apr 10 15:18:39 <brucethompson> checking whether we are cross compiling... configure: error: in `/home/brucet/bu
Apr 10 15:18:39 <brucethompson> ildbot/source/build_dir/target-x86_64_musl/python-pynacl-1.3.0/build/temp.linux2
Apr 10 15:18:40 <brucethompson> -2.7':
Apr 10 15:18:40 <brucethompson> configure: error: cannot run C compiled programs.
Apr 10 15:18:42 <brucethompson> If you meant to cross compile, use `--host'.
Apr 10 15:20:03 <brucethompson> I checked in setup.py and its checking the configuration for libsodium
Apr 10 15:20:31 <valku> It looks like the configure script attempts to compile and execute some code. The execute part obviously fails because you are cross-compiling.
Apr 10 15:21:06 <brucethompson> The makefile is about 10 layers removed from this error
Apr 10 15:21:11 <brucethompson> It created the makefile
Apr 10 15:21:14 <brucethompson> I
Apr 10 15:23:04 <valku> You probably need to (1) identify lines in your Makefile.am (or other source makefile, as appropriate) that cause this problem, (2) see if this test is indeed necessary for OpenWrt, and (3) if the offending test is not necessary, remove it from the source makefile by creating and applying a patch
Apr 10 15:23:34 <valku> https://openwrt.org/docs/guide-developer/build-system/use-patches-with-buildsystem
Apr 10 15:23:43 <brucethompson> Could these lines in the Makefile cause a cross-compile?
Apr 10 15:23:48 <brucethompson> PYTHON_PKG_SETUP_VARS:= SODIUM_INSTALL=system
Apr 10 15:23:48 <brucethompson> PYTHON3_PKG_SETUP_VARS:= SODIUM_INSTALL=system
Apr 10 15:25:19 <brucethompson> I will look at this doc.
Apr 10 15:25:26 <brucethompson> What is the definition of a patch?
Apr 10 15:25:45 <valku> Mmm... Could be, could be not. You probably need to dig deeper into the source makefiles. I wish I knew a shortcut. Others here might have some other ideas.
Apr 10 15:27:05 <brucethompson> No problem. But I am not sure that patches are relevant to this problem.
Apr 10 15:27:23 <brucethompson> This is a new package that has never been ported to OpenWrt.
Apr 10 15:27:37 <brucethompson> So what is the purpose of patches?
Apr 10 15:27:53 <valku> OpenWrt build system downloads an upstream package, unpacks it and builds it according to instructions provided by OpenWrt's package Makefile. If you need to modify the source makefiles of the upstream package after the build system unpacks the package but before it proceeds with building it, you have to create a patch
Apr 10 15:28:37 <valku> the build system will apply a patch **after** unpacking but **before** doing the actual build work
Apr 10 15:28:41 <brucethompson> Ah...
Apr 10 15:29:26 <brucethompson> OK. Let me check on this some more.
Apr 10 15:29:28 <valku> See for example feeds/packages/net/subversion/patches/301-cross-compilation-macos.patch
Apr 10 15:29:33 <brucethompson> Thanks greatly for the info.
Apr 10 15:29:40 <valku> np
Apr 10 15:39:59 <karlp> there's also CONFIGURE_ARGS+=blah and PKG_FIXUP:=autoreconf which might help too.
Apr 10 15:40:13 <karlp> (if you can just tell configure how to configure properly via it's own arguments)
Apr 10 15:40:26 <karlp> (which autohell promises, but doesn't always exactly deliver)
Apr 10 16:19:19 <starlays>  hello to all, I'm using a snapshot and I have build a kernel module for my corresponding release number. I have a strange thing in the dependency of my package and I don't know why there is another kernel required: https://pastebin.com/raw/41kVMuZJ How can I eliminate one of the deps and let only the necessary kernel version as dep? Thank you
Apr 10 16:21:44 <valku> starlays: are you on extroot?
Apr 10 16:22:55 <starlays> valku, no, I have build it in debian vm
Apr 10 16:24:29 <valku> The system where you run "opkg install kmod-fs-..." -- does this system use extroot?
Apr 10 16:24:35 <starlays> I have checked out to a new branch on the required sha and then fallowed the build instructions from the wiki. Have checked include/kernel-version.mk and in there is present only the correct kernel version 4.14.109-1
Apr 10 16:24:52 <starlays> no
Apr 10 16:25:39 <starlays> the corresponding git sha for my trunk build
Apr 10 16:25:54 <starlays> to be more specific
Apr 10 16:27:13 <valku> Looks like there may be some leftovers from previous builds causing ABI mismatches. Try "make clean" to start building from a clean slate.
Apr 10 16:27:43 <valku> Yes it will take time to build everything from scratch but it often solves these errors
Apr 10 16:27:52 <starlays> I have used ./scripts/getver.sh r9814 && git checkout -b r9814 7f3facfce29639ec50677aa2b0eecebe259aa4d3
Apr 10 16:27:58 <brucethompson> valku: I have another question
Apr 10 16:28:17 <starlays> I have removed the complete repo and redone it, same result , 2 kernel deps
Apr 10 16:30:15 <starlays> I'm wandering if there is another place where it is defined the second kernel dep
Apr 10 16:31:08 <valku> starlays: are you saying that you either ran "make clean" after getver.sh and git checkout, or you did not have any previous builds at all?
Apr 10 16:31:49 <valku> brucethompson: shoot your question, I will go offline soon for a few hours
Apr 10 16:32:20 <brucethompson> I am trying to directly patch files in the source directory (like setup.py).
Apr 10 16:32:40 <brucethompson> But when I run make compile, the patches get overwritten.
Apr 10 16:32:59 <starlays> valku, I didn't have any previous build, I have checkout, switch branch on my snapshot sha, run the feed update, make menuconfig, make download, make, c&P to the router and got the error message
Apr 10 16:33:25 <starlays> c&p the ipk file
Apr 10 16:33:31 <brucethompson> Is there a way to prevent the system from updating the source directory and just perform the build?
Apr 10 16:33:44 <valku> brucethompson: if you edit files in the source tree, your edits will be overwritten by the build system.
Apr 10 16:34:16 <valku> Again, there is no shortcut I am aware of. You should follow the process described in the wiki page to create a patch with quilt
Apr 10 16:34:17 <brucethompson> I know that. I am just trying some tests.
Apr 10 16:34:36 <brucethompson> OK
Apr 10 16:34:39 <valku> look at the bottom of the wiki page, there are some suggestions you may find useful
Apr 10 16:34:44 <brucethompson> OK
Apr 10 16:36:33 <starlays> I'm doing a new build right now again from scratch ...
Apr 10 16:37:16 <valku> starlays: running "scripts/feeds update" updates your packages git repo, which is separate from the openwrt git repo. "cd feeds/packages; git remote -v". It seems possible that some packages depend on a kernel version that does not yet exist as of your r9814
Apr 10 16:38:15 <starlays> kmod-fs-ext4_4.14.109-1_mipsel_24kc.ipk
Apr 10 16:38:35 <starlays> this should be dependent to only one kernel not 2
Apr 10 16:39:10 <valku> yes, it should be dependent only on one kernel
Apr 10 16:39:47 <starlays> I'm still looking for the makefile for this package, I will take a look inside
Apr 10 16:40:30 <valku> package/kernel/linux/modules/fs.mk
Apr 10 16:46:13 <valku> starlays: what do you get from "cat /etc/openwrt_version"?
Apr 10 16:47:08 <starlays> r9814-7f3facfce2
Apr 10 16:47:30 <starlays>  OpenWrt SNAPSHOT, r9814-7f3facfce2
Apr 10 16:48:30 <valku> looks correct
Apr 10 16:50:13 <starlays> nothing added for the LINUX_VESRION in the kernel module make file
Apr 10 16:52:33 <valku> check "uname -a", you should have 4.14.109
Apr 10 16:53:36 <valku> check also "opkg list-installed | grep kernel"
Apr 10 16:54:27 <starlays> uname -a > Linux OpenWrt 4.14.109 #0 SMP Sat Apr 6 18:29:35 2019 mips GNU/Linux
Apr 10 16:55:08 <starlays>   "opkg list-installed | grep kernel" > kernel - 4.14.109-1-2fbbf5991ca75130d18b6809f5795dfe
Apr 10 16:55:30 <valku> perhaps you should check "opkg list-installed | grep 4.14.111" also
Apr 10 16:55:48 <starlays> no output
Apr 10 16:55:52 <valku> odd
Apr 10 16:57:21 <valku> starlays: sorry, have to go offline for a while and hope that you'll get some other ideas here
Apr 10 16:57:44 <starlays> thank you for your help
Apr 10 17:03:24 <nyt> rmilecki, took me a minute to realize it was a patch causing the issue but otherwise wasn't too bad to debug
Apr 10 19:20:21 <igel> re: snapshots ?
Apr 10 19:22:40 <starlays> can someone help me with this problem: https://pastebin.com/raw/CnHK9PJg  I don't understand what is downloading if I'm targeting a local package that has no deps what is opkg downloading?
Apr 10 19:23:13 <igel> you have snapshots
Apr 10 19:23:21 <igel> so you need to upgrade your kernel
Apr 10 19:23:47 <igel> the kernel magic (checksum) must match when installing kernel modules
Apr 10 19:24:25 <starlays> igel, I have compiled the package for my specific snapshot sha
Apr 10 19:24:40 <starlays> 4.14.109-1
Apr 10 19:24:55 <starlays> r9814 7f3facfce29639ec50677aa2b0eecebe259aa4d3
Apr 10 19:25:09 <Borromini> you haven't, since your error message talks about kernel 4.14.111
Apr 10 19:25:20 <Borromini> and a different hash as well
Apr 10 19:25:21 <igel> i understand, but the kernel you have installed, or have compiled, is not the same
Apr 10 19:25:35 <starlays> it is
Apr 10 19:25:48 <ynezz> ldir: https://github.com/ynezz/linux/commit/2165f5cc846d23d00552d5286b26437c44dabc4f unbreaks perf on ath79 for me
Apr 10 19:26:13 <Borromini> starlays: don't argue with logic.
Apr 10 19:26:32 <starlays> I don't understand why it opk trigering and download from a source that is for kernel 4.14.111 if I give hime the specific package kmod-crypto-crc32c_4.14.109-1_mipsel_24kc.ipk
Apr 10 19:26:34 <igel> you should copy the kmod-crypto opkg from your bin/ to the device
Apr 10 19:26:47 <igel> then opkg install kmod*..
Apr 10 19:27:07 <igel> ooooh
Apr 10 19:27:09 <starlays> Borromini, Downloading http://downloads.openwrt.org/snapshots/targets/ramips/mt7621/packages/kmod-crypto-crc32c_4.14.111-1_mipsel_24kc.ipk
Apr 10 19:27:14 <starlays> igel, done that
Apr 10 19:27:22 <Borromini> starlays: nevermind me, saw that now.
Apr 10 19:27:23 <starlays> it is in my tmp dir
Apr 10 19:27:34 <igel> Borromini: hehe i did the same thing
Apr 10 19:27:54 <igel> ummm... thats odd
Apr 10 19:28:09 <Borromini> starlays: try passing the path
Apr 10 19:28:15 <Borromini> opkg install /tmp/$packagename
Apr 10 19:28:16 <starlays> hmm, ok
Apr 10 19:28:31 * igel never seen that
Apr 10 19:29:11 <starlays> same result, it is taking the source for the newer package
Apr 10 19:29:14 <starlays> :(
Apr 10 19:29:16 <Borromini> :-/
Apr 10 19:29:39 <igel> errr
Apr 10 19:29:44 <ldir> ynezz: oooh will give that a go :-)
Apr 10 19:29:50 <Borromini> starlays: maybe try playing with 'flag'
Apr 10 19:29:59 <starlays> witch one?
Apr 10 19:30:42 <Borromini> opkg flag user kmod-crypto-crc32c e.g.
Apr 10 19:30:45 <Borromini> then try installing again
Apr 10 19:31:19 <Borromini> if you just call 'opkg' you can see its help page so to speak
Apr 10 19:31:27 <Borromini> maybe something there that can help.
Apr 10 19:32:08 <starlays> --force-overwrite	Overwrite files from other package(s)
Apr 10 19:32:14 <starlays> maybe this ...
Apr 10 19:32:17 <starlays> will work
Apr 10 19:32:28 <Borromini> that's for conflicting files usually
Apr 10 19:32:33 <Borromini> not package versions.
Apr 10 19:32:45 <igel> like... this is odd
Apr 10 19:32:54 <Borromini> e.g. package A wants to install a file that package B already installed
Apr 10 19:33:13 <igel> can you clear the opkg file cache?
Apr 10 19:33:23 <igel> is it trying to just install the latest version?
Apr 10 19:33:37 <igel> eg; --downgrade option or something?
Apr 10 19:33:51 <starlays> --offline-root <dir>	offline installation of packages.
Apr 10 19:34:11 <starlays> igel, how can I clear the cache
Apr 10 19:34:12 <starlays> ?
Apr 10 19:34:21 <starlays> for opkg
Apr 10 19:34:34 <igel> i know 1 way, to reboot :/
Apr 10 19:34:48 <igel> um... not sure theres a flag, or a directory in /tmp/
Apr 10 19:35:33 <Borromini> everything is in /tmp
Apr 10 19:35:37 <Borromini> if there's a cache
Apr 10 19:35:52 <Borromini> good luck. gotta go.
Apr 10 19:35:53 <starlays> "/tmp/opkg-lists/"
Apr 10 19:36:49 <starlays> igel, bingo! it is doing the right thing now
Apr 10 19:36:57 <starlays> thank you
Apr 10 19:48:28 <igel> the link https://downloads.openwrt.org/releases/snapshot/ is now 404
Apr 10 19:54:45 <igel> its possibly messing up other builds/stuffs... eg; https://chef.libremesh.org/
Apr 10 19:55:15 <igel> only packages in stable/releases are avail.. no snapshotz0rs
Apr 10 19:56:34 <karlp> that's not the link that's been on downlaods.openwrt.org for... a while at least?
Apr 10 19:56:54 <karlp> http://downloads.openwrt.org/snapshots/targets/ ?
Apr 10 20:08:19 <igel> aah, crap. ok
Apr 10 20:13:51 <aparcar[m]> jow ping
Apr 10 20:17:05 <ldir> ynezz: seems to work on the archer c7 v2 under ath79 now :-)
Apr 10 21:44:01 <mangix> fork is evil
Apr 10 21:44:05 <mangix> intersting...
**** ENDING LOGGING AT Thu Apr 11 02:59:58 2019