whitequark[cis] changed the topic of #glasgow to: https://glasgow-embedded.org · digital interface explorer · https://www.crowdsupply.com/1bitsquared/glasgow · meetings Saturday 2200 UTC · code https://github.com/GlasgowEmbedded/glasgow · logs https://libera.irclog.whitequark.org/glasgow · matrix #glasgow-interface-explorer:matrix.org · discord https://1bitsquared.com/pages/chat
ewenmcneill[m] has joined #glasgow
<ewenmcneill[m]> supersat (@_discord_342359737444204544:catircservices.org) for your cable question, if you have them cheap male to female "du pont" cables might be a suitable sacrificial adapter between the valuable Glasgow fly leads and a DB25 pin (which is noticeably larger, but cheap "du pont" cables tend to be somewhat loose). Ie, treat the "du pont" cable like an "extension lead".
<ewenmcneill[m]> Re shipping: I got a DHL shipment notification a full day before the CrowdSupply shipping notification came. And my "ordered day 2" (non-early bird) cases haven't shipped yet. So I suspect Crowd Supply are shipping new batches of orders each time someone finds/opens another shipping box of them and inventories it....
<ewenmcneill[m]> (Also when ordering, and only when ordering, multiple items via Crowd Supply it's possible to choose "ship together" or "ship items as available". In the past I've definitely had "ship items as available" work for part shipments. So I'm hoping my non-early-bird cases will ship from the first case batch, in the next few weeks.)
Wanda[cis] has quit [Quit: Idle timeout reached: 172800s]
joerg has quit [Ping timeout: 255 seconds]
joerg has joined #glasgow
dos has quit [Ping timeout: 255 seconds]
dos has joined #glasgow
ar has quit [Ping timeout: 252 seconds]
q3k has quit [Ping timeout: 245 seconds]
ar has joined #glasgow
q3k has joined #glasgow
arachnist has joined #glasgow
ar has quit [Quit: leaving]
arachnist is now known as ar
<shiz> whitequark[cis]: yeah, it was a mostly-joking jab at my own incompetence at writing gateware :p
joerg is now known as Guest5085
Guest5085 has quit [Killed (cadmium.libera.chat (Nickname regained by services))]
joerg has joined #glasgow
q3k has quit [Changing host]
q3k has joined #glasgow
Foxyloxy has quit [Quit: Textual IRC Client: www.textualapp.com]
GNUmoon has quit [Ping timeout: 252 seconds]
Foxyloxy has joined #glasgow
GNUmoon has joined #glasgow
Foxyloxy has quit [Quit: Textual IRC Client: www.textualapp.com]
GNUmoon has quit [Ping timeout: 252 seconds]
GNUmoon has joined #glasgow
brolin has joined #glasgow
Foxyloxy has joined #glasgow
brolin has quit [Ping timeout: 240 seconds]
_whitelogger has joined #glasgow
ar-jan has joined #glasgow
<_whitenotifier-f> [glasgow] whitequark opened pull request #442: Add `udevadm` commands to the manual to remove the need for plugging out/in - https://github.com/GlasgowEmbedded/glasgow/pull/442
<whitequark[cis]> can arch/rhel/centos people chime in on the permissions to use for the USB device?
ar-jan has quit [Ping timeout: 255 seconds]
ar-jan has joined #glasgow
<_whitenotifier-f> [glasgow] whitequark commented on issue #441: udev rules on Arch(-ish) systems - https://github.com/GlasgowEmbedded/glasgow/issues/441#issuecomment-1758126974
<_whitenotifier-f> [glasgow] whitequark closed issue #441: udev rules on Arch(-ish) systems - https://github.com/GlasgowEmbedded/glasgow/issues/441
<_whitenotifier-f> [glasgow] whitequark commented on issue #441: udev rules on Arch(-ish) systems - https://github.com/GlasgowEmbedded/glasgow/issues/441#issuecomment-1758134130
bvernoux has joined #glasgow
jslcom[m] has joined #glasgow
<jslcom[m]> crw-rw---- 1 root dialout 4, 64 Oct 4 18:51 /dev/ttyS0
<jslcom[m]> Make sure your user account is a member of the dialout group.
<whitequark[cis]> yeah but glasgow isn't a modem
<whitequark[cis]> (well, usually. sometimes it is a modem. but not in a unix sense)
<jslcom[m]> crw-rw---- 1 root dialout 188, 0 Oct 4 18:50 /dev/ttyUSB0
<whitequark[cis]> right but that's technically still a modem (it's a CDC-ACM device, USB serial modem)
<whitequark[cis]> and glasgow is a general USB device, it lives under /dev/bus/usb
<jslcom[m]> If you want a non-root account to access the device, you need something similar. You could create a new group and use that instead.
<whitequark[cis]> right, that's an obvious solution, though a heavyweight one; you would think by 2023 we would have a way to handle pluggable devices relevant to the physical terminal session...
<jslcom[m]> You could also create a udev rule to change permissions/ownership on instantiation.
<whitequark[cis]> we have one already
<whitequark[cis]> it uses GROUP="plugdev", MODE="0660", which doesn't work on systems without a plugdev group
<whitequark[cis]> it could use TAG+="uaccess", TAG+="udev-acl", which is what I'm currently messing around with; it appears to be broken in some confusing way, as systemd stuff usually is
<jslcom[m]> There are variations between distributions. The only way to ensure it works across them all is to create a new group.
<_whitenotifier-f> [glasgow] whitequark commented on issue #441: udev rules on Arch(-ish) systems - https://github.com/GlasgowEmbedded/glasgow/issues/441#issuecomment-1758146846
<jslcom[m]> Some distros (such as duvian) don't have systemd.
<jslcom[m]> s/duvian/devuan/
<whitequark[cis]> then they would not have systemd-udevd either, making the question of which udev rules to use irrelevant
<whitequark[cis]> (it's been many years since udev became a systemd component)
<jslcom[m]> You can choose which distros you want to claim to support. The folks running the oddball ones can probably figure it out themselves.
<whitequark[cis]> I'm aware of this, as well as other basic Linux distribution related facts
<whitequark[cis]> the question boils down to: what is the biggest share of Linux distributions we can support without doing something weird?
<jslcom[m]> Most people doing development are using debian-based distros because they have support for newer toolchains.
duderonomy has joined #glasgow
<_whitenotifier-f> [glasgow] whitequark synchronize pull request #442: Add `udevadm` commands to the manual to remove the need for plugging out/in - https://github.com/GlasgowEmbedded/glasgow/pull/442
<_whitenotifier-f> [glasgow] whitequark commented on issue #441: udev rules on Arch(-ish) systems - https://github.com/GlasgowEmbedded/glasgow/issues/441#issuecomment-1758155196
<_whitenotifier-f> [glasgow] whitequark commented on issue #441: udev rules on Arch(-ish) systems - https://github.com/GlasgowEmbedded/glasgow/issues/441#issuecomment-1758159344
<_whitenotifier-f> [glasgow] whitequark reopened issue #441: udev rules on Arch(-ish) systems - https://github.com/GlasgowEmbedded/glasgow/issues/441
<whitequark[cis]> actually I figured it out
<whitequark[cis]> SUBSYSTEM=="usb", ATTRS{idVendor}=="04b4", ATTRS{idProduct}=="8613", TAG+="uaccess" in a file called 70-glasgow.rules (and not 99-glasgow-rules) seems to work
<whitequark[cis]> * SUBSYSTEM=="usb", ATTRS{idVendor}=="04b4", ATTRS{idProduct}=="8613", TAG+="uaccess" in a file called 70-glasgow.rules (and not 99-glasgow.rules) seems to work
<jslcom[m]> So it was the instantiaon order?
<whitequark[cis]> yeah
<_whitenotifier-f> [glasgow] whitequark commented on issue #441: udev rules on Arch(-ish) systems - https://github.com/GlasgowEmbedded/glasgow/issues/441#issuecomment-1758170501
<Wanda[cis]> oh. is that why this wasn't working?
Wanda[cis] has joined #glasgow
<whitequark[cis]> yep.
<whitequark[cis]> you love it don't you?
<whitequark[cis]> (I do not in fact love it)
<Wanda[cis]> I.
<Wanda[cis]> was wondering about this.
<Wanda[cis]> but came to the conclusion that invocation order cannot possibly matter when it's the only thing in my /etc/udev/rules.d/
<Wanda[cis]> it indeed fixes the problem on my laptop
<Wanda[cis]> and I can now get rid of the satanic workaround
<Wanda[cis]> absolutely lovely
<whitequark[cis]> $ ls /lib/udev/rules.d/
<whitequark[cis]> (yes, they are all merged)
<whitequark[cis]> there is a 70-uaccess.rules there but it's not where it's implemented; that's just some default uaccess rules
<whitequark[cis]> I think it's 71-seat.rules that applies it: TAG=="uaccess", SUBSYSTEM!="sound", TAG+="seat"
<Wanda[cis]> yeah, it seems so
<whitequark[cis]> now what the fuck does that do? good question, never heard of it before
<whitequark[cis]> (from 73-seat-late.rules) none of these words are in the bible
<whitequark[cis]> oh I guess it's the RUN{builtin}+="uaccess" does that?
<whitequark[cis]> I wish this was written in a reasonable language like C++
<whitequark[cis]> you know, something easy to understand and modify
<Wanda[cis]> ... yeah
<whitequark[cis]> at least C++ has structured control flow and inheritance.
<whitequark[cis]> lmfaosob
<whitequark[cis]> in the meantime github broke
<_whitenotifier-f> [glasgow] whitequark opened pull request #443: Python 3.12 is released, 3.13-dev is the next development snapshot - https://github.com/GlasgowEmbedded/glasgow/pull/443
<Wanda[cis]> oh right. aiohttp. lolsob.
<Wanda[cis]> at least they have a prerelease now?
<whitequark[cis]> YEP. STILL BROKEN
<whitequark[cis]> oh i see
dne has quit [Remote host closed the connection]
<_whitenotifier-f> [glasgow] whitequark commented on pull request #443: Python 3.12 is released, 3.13-dev is the next development snapshot - https://github.com/GlasgowEmbedded/glasgow/pull/443#issuecomment-1758190371
dne has joined #glasgow
<shiz> late addendum: eudev is, afaics, still a thing :P
<whitequark[cis]> well
<whitequark[cis]> > eudev is masked and scheduled for removal from Gentoo. It's recommended to use udev instead. The transition should be uneventful, however the corresponding News Item must be read and followed to avoid any issues.
<whitequark[cis]> are you sure?
<shiz> at least upstream
<shiz> dunno what gentoo is doing
<whitequark[cis]> was ... it not a gentoo project?
<whitequark[cis]> is gentoo not literally the eudev upstream?
<shiz> originally, not for a while anymore
<whitequark[cis]> wtf
<shiz> "On 2021-08-20 Gentoo decided to abandon eudev and a new project was established on 2021-09-14 by Alpine, Devuan and Gentoo contributors (alphabetical order)."
<whitequark[cis]> that doesn't support uaccess, right
<shiz> doesn't seem so
* shiz feels guilty now for bringing it up
<whitequark[cis]> so what do eudev distros do now for pluggable devices?
<whitequark[cis]> actually hold on
<shiz> my /gues/ is they all have a plugdev group, or they didn't think about it
<whitequark[cis]> it's no longer a builtin in eudev
<shiz> fancy..
<whitequark[cis]> with a truly inspiring amount of #if 0
<whitequark[cis]> shiz: well the old builtin did something similar
<whitequark[cis]> this is arguably a better implementation than in systemd-udevd
<whitequark[cis]> as in
<shiz> :P
<shiz> okay, then +uaccess should still work
<shiz> right?
<whitequark[cis]> having the same code as a separate helper is probably better than stuffing seat management in udev
<shiz> presuming elogind is there
<shiz> yeah, i agree
<whitequark[cis]> but it's literally the same code afaict
<whitequark[cis]> shiz: yeah
<whitequark[cis]> and if you don't have elogind your desktop probably doesn't work or you're a hardcore oldschool sysadmin using startx and in neither case it's our problem
<shiz> :)
<whitequark[cis]> I guess the #if 0 are to delineate code copied from systemd-udevd so that's reasonable
<shiz> i think so, yeah
<_whitenotifier-f> [glasgow] whitequark synchronize pull request #442: Update udev rules and manual to remove non-portable (Debian-specific) use of the plugdev group and 0660 mode - https://github.com/GlasgowEmbedded/glasgow/pull/442
<_whitenotifier-f> [glasgow] github-merge-queue[bot] created branch gh-readonly-queue/main/pr-443-f3e3be347b3155bf5facde0612be1570ee3cdaa0 - https://github.com/GlasgowEmbedded/glasgow
<_whitenotifier-f> [glasgow] whitequark synchronize pull request #442: Update udev rules and manual to remove non-portable (Debian-specific) use of the plugdev group and 0660 mode - https://github.com/GlasgowEmbedded/glasgow/pull/442
<_whitenotifier-f> [GlasgowEmbedded/glasgow] github-merge-queue[bot] pushed 1 commit to main [+0/-0/±1] https://github.com/GlasgowEmbedded/glasgow/compare/f3e3be347b31...851c4affd9d7
<_whitenotifier-f> [GlasgowEmbedded/glasgow] whitequark 851c4af - CI: Python 3.12 is released, 3.13-dev is the next development snapshot.
<_whitenotifier-f> [glasgow] github-merge-queue[bot] deleted branch gh-readonly-queue/main/pr-443-f3e3be347b3155bf5facde0612be1570ee3cdaa0 - https://github.com/GlasgowEmbedded/glasgow
<_whitenotifier-f> [glasgow] whitequark closed pull request #443: Python 3.12 is released, 3.13-dev is the next development snapshot - https://github.com/GlasgowEmbedded/glasgow/pull/443
<whitequark[cis]> so, who wants to talk about package management and lockfiles?
<whitequark[cis]> I'm thinking we need to check a lockfile in, to make sure our past commits can be reproducibly tested
<whitequark[cis]> (bisection will eventually become a nightmare without this)
<whitequark[cis]> also the lockfile should probably have the least version of each dependency that's permitted by the pyproject.toml
<whitequark[cis]> but obviously we still need to check everything with the latest versions of dependencies too
<whitequark[cis]> how should this be organized on CI?
<whitequark[cis]> probably the easiest way I can think of is expanding our test matrix to twice as many items as it currently has
<whitequark[cis]> github test matrix has this lovely unhinged feature called include: https://docs.github.com/en/actions/using-jobs/using-a-matrix-for-your-jobs#expanding-or-adding-matrix-configurations
nemanjan00[m] has quit [Quit: Idle timeout reached: 172800s]
_whitelogger has joined #glasgow
<_whitenotifier-f> [glasgow] whitequark opened pull request #444: Include tests with a checked-in lockfile in the matrix on CI - https://github.com/GlasgowEmbedded/glasgow/pull/444
<_whitenotifier-f> [glasgow] whitequark synchronize pull request #444: Include tests with a checked-in lockfile in the matrix on CI - https://github.com/GlasgowEmbedded/glasgow/pull/444
<_whitenotifier-f> [glasgow] whitequark synchronize pull request #444: Include tests with a checked-in lockfile in the matrix on CI - https://github.com/GlasgowEmbedded/glasgow/pull/444
<_whitenotifier-f> [glasgow] whitequark synchronize pull request #444: Include tests with a checked-in lockfile in the matrix on CI - https://github.com/GlasgowEmbedded/glasgow/pull/444
<whitequark[cis]> wh... where did aiohttp-remote go?!
<whitequark[cis]> oh it's called aiohttp-remotes. wonder why nothing before has triggered an error on this? incomprehensible
<SnoopJ> are you running the pre-release?
<whitequark[cis]> no
<whitequark[cis]> not related to 3.12 at all
<SnoopJ> oh, a misnomer
<_whitenotifier-f> [glasgow] whitequark synchronize pull request #444: Include tests with a checked-in lockfile in the matrix on CI - https://github.com/GlasgowEmbedded/glasgow/pull/444
<_whitenotifier-f> [glasgow] whitequark synchronize pull request #444: Include tests with a checked-in lockfile in the matrix on CI - https://github.com/GlasgowEmbedded/glasgow/pull/444
<_whitenotifier-f> [glasgow] whitequark synchronize pull request #444: Include tests with a checked-in lockfile in the matrix on CI - https://github.com/GlasgowEmbedded/glasgow/pull/444
<whitequark[cis]> okay i think i'm done with this PR
bvernoux has quit [Quit: Leaving]
brolin has joined #glasgow
ar-jan has quit [Ping timeout: 260 seconds]
Eli2_ has joined #glasgow
Eli2| has quit [Ping timeout: 258 seconds]
brolin has quit [Ping timeout: 240 seconds]
brolin has joined #glasgow
brolin has quit [Ping timeout: 246 seconds]
<_whitenotifier-f> [glasgow] ewenmcneill commented on pull request #442: Update udev rules and manual to remove non-portable (Debian-specific) use of the plugdev group and 0660 mode - https://github.com/GlasgowEmbedded/glasgow/pull/442#issuecomment-1758594624
brolin has joined #glasgow
<_whitenotifier-f> [glasgow] bobtwinkles reviewed pull request #442 commit - https://github.com/GlasgowEmbedded/glasgow/pull/442#discussion_r1355836121
brolin has quit [Ping timeout: 272 seconds]
duderonomy has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
duderonomy has joined #glasgow
<_whitenotifier-f> [glasgow] whitequark synchronize pull request #442: Update udev rules and manual to remove non-portable (Debian-specific) use of the plugdev group and 0660 mode - https://github.com/GlasgowEmbedded/glasgow/pull/442
<_whitenotifier-f> [glasgow] whitequark reviewed pull request #442 commit - https://github.com/GlasgowEmbedded/glasgow/pull/442#discussion_r1355852476
brolin has joined #glasgow
brolin has quit [Ping timeout: 264 seconds]
brolin has joined #glasgow
duderonomy has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
brolin has quit [Ping timeout: 272 seconds]
brolin has joined #glasgow
duderonomy has joined #glasgow
<_whitenotifier-f> [glasgow] bobtwinkles reviewed pull request #442 commit - https://github.com/GlasgowEmbedded/glasgow/pull/442#discussion_r1355895305
<_whitenotifier-f> [glasgow] whitequark reviewed pull request #442 commit - https://github.com/GlasgowEmbedded/glasgow/pull/442#discussion_r1355895914