00:20:17 dsp-: because of its history, you need to disable a few of the fsck checks when cloning the repo 00:20:38 it's not isolated to illumos-omnios, the "offending" comments are in illumos-gate as well 00:20:55 https://github.com/pfmooney/dotfiles/blob/master/gitconfig#L23-L27 00:31:05 rude 00:40:15 yes, it's not cloning that's the issue necessarily, but pushing to github. or do these changes (even though they're for fetch.fsck) also apply for pushing? 00:43:26 (as in, create a new repo; add as remote in local clone; git push origin master -- at least for me this does not appear to work) 00:44:45 it isn't a blocker, i have workarounds. just unpleasant. (but unavoidable without noisy rebase i suppose.) 02:53:24 It's unfortunate, but fixing it would require rewriting repo history 02:55:20 clundered 03:54:52 i'll open an issue with github, maybe they can expose a switch somewhere. being unable to import repos is kinda crappy, even if they have invalid data. clearly they have no problem with the existing ones, or forks made from them, so. will mentally file as 'wontfix'. 06:51:31 dsp-: FWIW, you don't need a separate fork. You can push branches that are based on either illumos/illumos-gate or omniosorg/illumos-omnios to a fork of either one 06:51:57 e.g., I just have a fork of https://github.com/jclulow/illumos-gate with branches from all over the place 06:52:06 that is, a fork of illumos/illumos-gate there 14:45:40 jclulow: oh yes. just that my fork of illumos-gate was called illumos-gate, and I was using it as such, and I only went to fork as illumos-omnios because that's what the omni tool instructions said it expected. 14:47:35 just having two clones is just less to mess with (if it worked), no need to rename your master branch something else, and stitch stuff together, or rename my repo (probably it's configurable in the omni tool! i didn't get that far.) 14:49:40 just that my "this should work" hack didn't :) alas. is not a blocker. and have brought it to the attention of github, would be nice if it were possible to push the repo to the same place that it's hosted from, hah 14:53:51 and indeed, this fails with the same error: cd illumos-omnios && git remote remove origin && git remote add origin git⊙gc:dspearson/illumos-gate.git && git push origin master:master-omnios 14:53:58 i admit maybe i'm holding git wrong 14:55:20 (I remove the illumos-omnios as origin, switching it to the upstream illumos-gate fork i had. then push your master to master-omnios alongside the illumos-gate's existing stuff. remote: fatal: fsck error in packed object 14:55:42 dsp-, Is that like holding it in utter disdain? 14:55:46 Or is that just a me thing? 14:56:29 pretty much. ex darcs using lisp weenie that gets by living in magit :) 14:57:18 "gotta pay the bills", and all that. 19:19:52 dsp-: You might need to do a "git fetch --all --prune" after you change the origin? 22:05:42 jclulow: I did that, then push origin master:master-omnios, and the outcome is the same. :( 22:07:20 So what I usually do is this: git clone https://github.com/illumos/illumos-gate.git; cd illumos-gate; git remote add jclulow git⊙gc:jclulow/illumos-gate.git ; git fetch --all --prune; git checkout -b my-work ; ... commits ; git push -vu jclulow my-work 22:07:47 Or some combination of that sort of thing 22:08:33 I also add other upstreams; e.g., https://code.illumos.org/illumos-gate.git (our gerrit) and https://github.com/oxidecomputer/illumos-gate.git and https://github.com/omniosorg/illumos-omnios.git 22:08:36 if i want to push the omnios master alongside the upstream though, in this case, i guess this workflow wouldn't work. but let me try it anyway 22:08:53 I mean, you have to name the branches something unique 22:09:06 yes. in this case i named it master-omnios 22:09:53 * nomad read that as "master ominous" and was looking for the new archvillan. 22:10:44 the villain is fscking my branches :) 22:11:46 https://github.com/orgs/community/discussions/114208 - posted this at least 22:12:30 btw, i am extremely new to illumos still, and dipping my toes into the culture here 22:12:57 how open and tolerant are you all to complete novices when it comes to things like kernel work? 22:14:09 how I ended up down this rabbit hole was the lack of dprintf - I guess libc isn't kernel, but I would like to dig my teeth in eventually 22:16:19 (guile30 from pkgsrc fails to build since this functionality is not implemented. https://www.illumos.org/issues/16151 ) 22:16:20 → FEATURE 16151: want dprintf (New) 22:19:29 um, is there tshark in omnios? 22:22:26 wireshark 22:22:49 presuming you mean a network packet sniffer 22:22:50 /opt/pkgsrc/net/wireshark# grep tshark PLIST 22:22:50 bin/tshark 22:23:21 ah, ok, with pkgsrc... 22:24:15 maybe there are packages elsewhere. I'm just unfamiliar with that ecosystem, but used netbsd lots 22:29:43 : || lvd@fs1 ~ [509] ; pkg search -p snoop 22:29:44 PACKAGE PUBLISHER 22:29:44 pkg:/network/snoop⊙0 omnios 22:30:05 took me a while to remember the name. 22:30:20 SNOOP(8) Maintenance Commands and Procedures SNOOP(8) 22:30:20 NAME 22:30:20 snoop - capture and inspect network packets 22:39:56 unfortunately, snoop is rather broken and needs much work. 22:40:26 i'm just trying to build wireguard from pkgsrc now to see if it works 22:41:48 ofc, snoop is good to capture packets and so those can be analyzed where tools are present;)