Replying to
@Jourei @ironicbadger Move fast and break things!
Remote
Marius Gedminas
@mgedmin@floss.social
mastodon 4.7.3Maintainer of several useless FLOSS projects.
181 Followers
272 Following
20 Posts
Joined November 12, 2022
Website:
GitHub:
Pronouns:
he/him
Ko-fi:
Replying to
Still with cryptsetup present in the initramfs I managed to unlock and mount my root partition in the unitramfs prompt, and when I re-ran update-initramfs -u after booting, it no longer complained, so I'm hopeful.
Replying to
@QuietMisdreavus Flowx for Android graphs the temperature for the next several days (configurable). The shaded areas compare with the same time 24 hours prior, so you can see if it's getting warmer or colder. There's also info about cloud cover, rain/snow, wind speed in the same timeline.
Replying to
@verdre@mastodon.social @karolherbst@chaos.social There's also https://github.com/Tookmund/Swapspace packaged in Debian as swapspace, also not installed by default.
It worked well enough to completely hide an xdg-desktop-portal-gnome memory leak from me, until I accidentally noticed I had 17 GB of swap files for some reason after running htop.
Replying to
@david_chisnall@infosec.exchange @nblr@chaos.social
Not complaining loudly to the logs when cron detects a malformed crontab is a crime if still true.
I looked at the source, and apparently Debian added a patch to report truncated crontabs back in either 2002 or 2010: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=79037 (AFAICT the original 2002 fix was incorrect and it was fixed again, properly in 2010).
I am sad that this was done in a distro patch and not pushed upstream. Is there an active upstream for cron?
Replying to
@eevee @glyph @dotstdy These days snaps of GUI apps refuse to be updated while running. You get annoying notifications instead, asking to close the app so it could then be updated. Snapd doesn't even pre-download the update!
There's supposedly a hard limit of something like 14 days after which a snap will get force-updated even if you never close the app, but my brain won't let me ignore a pending update for so long.
Replying to
@bws @bremner Slack video calls didn't work on Firefox, so I had to use Chromium too. I did some research and apparently there were two APIs for webrtc: Chrome's and W3C standard. Chromium implemented both. Slack refused to use the W3C standard API. Firefox refused to implement the Chrome API. Stalemate.
Replying to
Also, choosing to do this today, when most of Ubuntu's online presence is down due to a DDoS attack, was a decision that might not have been the best, in retrospect.
(an apt upgrade of thunderbird fails if it can't talk to the snap store)
Replying to
All fixed now, hopefully.
I'm not brave enough to try rebooting to see if it would come up again without repeated manual intervention at the initramfs prompt.
cryptsetup-initramfs got removed during the upgrade, and when I reinstalled it in a chroot in a live session, it didn't like something about LUKS device names not matching my /etc/crypttab.
Replying to
@eevee @glyph @dotstdy Chrome had better UX back in the day. Firefox had jank like rounding errors during text rendering making individual lines shift up/down by one pixel as you scrolled the page or selected text.
When I saw the writing on the wall and decided to switch back to Firefox, I had to set up a hg clone and try to submit patches for basic stuff like rearranging tab order with ctrl+shift+pageup/pagedown, because Firefox didn't have it.
whoever is DDoSing Ubuntu servers please stop, I'm trying to debug a new script for fetching PPA repository signing keys
Anyway my FOSS project dashboard is all green again.
I miss my laptop (still waiting for a replacement battery at the repair shop).
Does anyone know why pypy3.11 is so utterly broken on windows-latest on GitHub Actions?
(broken == pip install fails with an OSError? something about an incomplete read? while reading from a socket? while checking for newer pip versions in PyPI?)
Hating Github workflows so much right now.
Replying to on infosec.exchange
@xaetacore@neondystopia.world @0xabad1dea@infosec.exchange If you press Ctrl-X Ctrl-O (in insert mode) while editing a Python file, Vim will execute code from that file (and any __init__.py in a sibling/parent directory). This is called 'omnicomplete' and probably relies on you having a 'filetype plugin on' in your .vimrc, but it's a feature shipped by default.
Replying to
@david_chisnall@infosec.exchange @nblr@chaos.social i've seen this justified as 'the file could be truncated, and it could lead to badness', if e.g., a command like
0 0 * * * rm -r /tmp/daily-debug-logs/
got truncated after the first / (ignoring the modern --preserve-root behaviour of rm for simplicity).
Not complaining loudly to the logs when cron detects a malformed crontab is a crime if still true.
Replying to
@spmatich@ioc.exchange @april@macaw.social I have vague memories of traceroute being declared deprecated with tracepath as the replacement. No idea if that one's installed by default or not.
Replying to
@jwz@mastodon.social @grimmy@mastodon.social Is it the space between the ] and drivers? Because that would be hilarious.
(I got it to work for gtimelog a long time ago, but have no memory of any details. My POTFILES.in has no spaces after the ].)