**** BEGIN LOGGING AT Sun Apr 21 02:59:57 2019 Apr 21 07:52:21 puppies and kittens ... Apr 21 08:42:28 isn't it bunnies and chocolates? Apr 21 14:02:58 ffs it pissed me right off when package bumps are done without even doing basic things/checks that patches have been refreshed to avoid hunk offsets. It tells me all I need to know about the level & quality of testing done. ie. 'it compiles - f* it' Apr 21 14:06:53 * ldir bets it's a driveby github manual edit pr Apr 21 14:08:00 Well... Context would be nice :) Be it "this again" versus "first time ever trying" :P Apr 21 15:34:22 hmm... today compiled openwrt.. got some interesting DHCP related log Apr 21 15:34:24 Sun Apr 21 18:33:50 2019 daemon.info dnsmasq-dhcp[1660]: DHCPREQUEST(br-lan) 10.5.0.129 54:ee:75:45:19:b0 Apr 21 15:34:25 Sun Apr 21 18:33:50 2019 daemon.info dnsmasq-dhcp[1660]: DHCPNAK(br-lan) 10.5.0.129 54:ee:75:45:19:b0 lease not found Apr 21 15:34:55 rinse, repeat every 5-10 seconds Apr 21 15:36:34 I mean... client asks for IP, shouldn't it give it? Apr 21 15:53:46 ldir: as olmari I would like to hear more about that Apr 21 15:55:48 Ah... I see.. https://github.com/openwrt/openwrt/commit/13ffdf44824e8e8d276c2cb3b1cc181471c33101 yeah OpenWrt repository miss something like CircleCI which is in packages repository and when you send pull request to OpenWrt it gets closed instead of merged, so this could be also improved. Sometimes you need to read whole PR to see why it wasnt merged. Apr 21 16:13:48 circleci will not fail on wrong offsets. neither will git apply Apr 21 16:14:04 fuzz is the only issue Apr 21 16:16:24 maybe this should be applied to masted? https://git.openwrt.org/?p=openwrt/staging/jogo.git;a=commit;h=182e8e9643046b74884f5cc3cf3aa1f17fabb7df Apr 21 16:16:35 *master Apr 21 16:18:23 but in logs, you will see that there are hunk offsefs. Right now, you can to compile it yourself to see if there is anything wrong. **** BEGIN LOGGING AT Sun Apr 21 17:13:29 2019 Apr 21 18:05:00 any mini-pcie to pice adapters with 4 antennas for ac wave 2 radios? Apr 21 18:09:21 ldir: How does the automatic refresh work for parts of the toolchain? Apr 21 18:10:33 Hauke: it worked for me with make toolchain/binutils/{clean,refresh} Apr 21 18:11:40 ldir: ok, I thought that there was sonething not working for toolchain parts Apr 21 18:54:44 updated openwrt/upstream, https://sdwalker.github.io/uscan/index.html Apr 21 18:58:43 can someone say what functionality is lost when dropbear would be built without dbclient? Apr 21 19:26:03 Borromini: dbclient is an ssh client as opposed to ssh server. The default ssh client (on my install at least) is dbclient...ie the thing that allows me to ssh *from* the router to other devices. dropbear is the thing that allows me to ssh *to* the router in the first place. Apr 21 19:43:37 ok so it's short for 'dropbear client' Apr 21 19:44:19 thanks :) Apr 21 19:46:00 yes it appears so - and I learned something today Apr 21 19:47:42 the commit message is a bit dense, in my humble opinion: https://git.openwrt.org/?p=openwrt/openwrt.git;a=commitdiff;h=524810ce6ddff85a79af9ed979452bb8a8ed1b6d Apr 21 19:49:51 Borromini: why? Or I mean header says the rest Apr 21 19:50:08 i was reading dbclient as some database client, and i was wondering why dropbear kept a database Apr 21 19:50:40 i guess i've seen too many 'mariadbclient' debian upgrade messages :D Apr 21 19:50:44 Well naming is always an art ;D Apr 21 19:51:12 The commit message sort of answers the 'why' question but assumes you know the what. Apr 21 19:51:32 ldir: yep there is some sousentendu as the french would say Apr 21 19:52:54 If it said something like 'make building the ssh client (dbclient) optional' in the text it would have made life easier. Apr 21 19:53:12 Then again, context of "what" should be kind of known to readers of that caliber Apr 21 19:53:29 or even 'dropbear ssh client (dbclient)' Apr 21 19:54:03 that's almost an argument for not having any documentation - you can read the source can't you ;-) Apr 21 19:54:58 ..while ofcourse the clearer the better, but if that commit is dismissed solely of wording of messsge, then some faith in devs is lost :p Apr 21 19:55:30 olmari: assumptions are easily made. Apr 21 19:55:59 It says program name, it says what part.. and desc says what is achieved Apr 21 19:56:50 Again, I'm not saying that is perfect etc.. but it isn't THAT bad of desc either Apr 21 19:56:51 it never hurts to not take things for granted. Apr 21 19:57:21 neither was i saying it's a *bad* description Apr 21 19:58:20 ..if "we" start to nitpick too much, then we start to lose commits too ;p Apr 21 19:59:02 ..and no, nothing I said was meant to anyone personally :) Apr 21 20:06:57 * ldir calls it a night. "It's a night" Apr 21 20:07:19 Hello this is n..patrick Apr 21 20:19:01 Borromini: the bigger question not answered by the commit message: why does this need to be a compile time option and can't just be moved to its own .ipk Apr 21 20:19:27 KanjiMonster: even better point **** ENDING LOGGING AT Mon Apr 22 02:59:56 2019