| lkcl | Las[m], i forgot to ask: you are running the dev-env-setup script for the coriolis2-chroot? | 08:50 |
|---|---|---|
| lkcl | not, "read the instructions and follow them manually"? | 08:50 |
| lkcl | toshywoshy: openpowerbot's happy with irc but not mattermost (#microwatt, #openpower all not up-to-date) | 09:06 |
| Las[m] | <lkcl> "Las, i forgot to ask: you are..." <- Yep. | 10:00 |
| Las[m] | I ran the script manually lkcl | 10:01 |
| *** kylel1 is now known as kylel | 12:40 | |
| lkcl | kylel, looks great | 15:00 |
| lkcl | the number of times that function is called is so ridiculously high, if "python3 simple/test/test_issuer_svp64.py >& /tmp/f" works you can feel confident committing it | 15:02 |
| lkcl | see how the API is "emerging" rather than being "defined in advance"? | 15:02 |
| lkcl | it's a strange but quite fascinating process whereby you leverage the existing code to ensure functional correctness | 15:03 |
| lkcl | and go through a series of "annealing" code-morphs to *end up* with the target design | 15:04 |
| lkcl | paradoxically, the smaller those changes, the better. | 15:04 |
| lkcl | you'd think counter-intuitively that to create massive end-results you'd have to "do work" in massive monolithic commits | 15:05 |
| lkcl | Las[m], for sheer hilarity i'm just running the coriolis2-schroot script *inside* another schroot :) | 15:14 |
| lkcl | Las[m], did you also get, "E: Couldn't download packages: libdns-export1104 | 16:19 |
| lkcl | " | 16:19 |
| lkcl | which tends to suggest a severe problem with debian debootstrap | 16:20 |
| lkcl | first i've ever encountered something that serious | 16:20 |
| * lkcl re-running to make sure it's not just a transient error | 16:23 | |
| lkcl | ok, it's a transient error | 16:32 |
| lkcl | (symbiflow)root@fizzy:/home/lkcl/src/dev-env-setup# schroot -c coriolis | 16:34 |
| lkcl | E: std::bad_alloc | 16:34 |
| lkcl | wtf?? | 16:34 |
| lkcl | what the heeeelll is going on | 16:36 |
| lkcl | some f*****g bullshit systemd directory | 16:38 |
| lkcl | which if it doesn't exist causes schroot to segfault | 16:38 |
| lkcl | that's a severe policy violation for a debian package | 16:38 |
| lkcl | category 4. important. | 16:40 |
| Las[m] | At least the error is reproducible | 17:25 |
| lkcl | yes. | 17:29 |
| lkcl | mkdir /run/systemd/userdb on the *host* machine (!!) | 17:29 |
| lkcl | not the chroot | 17:30 |
| lkcl | the *host* | 17:30 |
| Las[m] | Thanks, I'll try doing that | 17:39 |
| lkcl | Las[m], there's a lot more i need to sort out | 18:39 |
| lkcl | this was all "fine" when using the bare-metal chroot type | 18:39 |
| lkcl | but it required MANUAL bind-mounting of /run /tmp /dev etc. | 18:40 |
| lkcl | which was intrusive for the host | 18:40 |
| lkcl | i'm also adding apt-acquire-retries "10" to the configs, i'm getting fed up with my ISP | 18:46 |
| * lkcl pushed the current mods to coriolis2-chroot script | 19:03 | |
| lkcl | Las[m], am re-running it again (for the 4th^W6th time) | 19:03 |
| lkcl | :) | 19:03 |
| lkcl | apt-acquire-retry seems to be helping, i've seen at least one warning "error retrying" already (!) | 19:04 |
| lkcl | dang had *another* ISP hissy-fit, had to add something to compensate for "bad proxy" | 20:52 |
| kylel | i was fed up with mine yesterday | 21:32 |
| kylel | yield is frustrating me in the test class | 21:34 |
| programmerjake | meeting in 22min | 21:38 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!