| josch | minute: https://source.mnt.re/reform/reform-debian-packages/-/merge_requests/29 | 00:39 |
|---|---|---|
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 00:47 | |
| - mtm (QUIT: Ping timeout: 246 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 01:02 | |
| minute | josch: thank you very much! my brain wasn't working well enough anymore ^^ this looks good! | 01:32 |
| + sevan (~sevan@2001:470:1f1d:1d6:5a55:caff:fe24:ed4) | 01:45 | |
| - jacobk (QUIT: Quit: http://quassel-irc.org - Chat comfortably. Anywhere.) (~quassel@utdpat242017.utdallas.edu) | 02:24 | |
| + jacobk (~quassel@utdpat242017.utdallas.edu) | 03:05 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 03:08 | |
| - jacobk (QUIT: Ping timeout: 268 seconds) (~quassel@utdpat242017.utdallas.edu) | 03:11 | |
| - skipwich (QUIT: Quit: DISCONNECT) (~skipwich@user/skipwich) | 04:02 | |
| + skipwich (~skipwich@user/skipwich) | 04:03 | |
| + jacobk (~quassel@64.189.201.150) | 05:46 | |
| - sevan (QUIT: Ping timeout: 256 seconds) (~sevan@2001:470:1f1d:1d6:5a55:caff:fe24:ed4) | 06:07 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:20) | 06:35 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 06:45 | |
| - GNUmoon (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 06:47 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 06:48 | |
| - GNUmoon (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 07:00 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 07:00 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 07:16 | |
| - anzu (QUIT: Quit: WeeChat 3.5) (~anzu@melkki.cs.helsinki.fi) | 08:40 | |
| + chomwitt (~chomwitt@2a02:587:7a11:6600:1ac0:4dff:fedb:a3f1) | 09:18 | |
| - Nixkernal (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~quassel@2a02:1210:1613:e600:e064:d3b8:33d1:5081) | 09:27 | |
| - chomwitt (QUIT: Ping timeout: 260 seconds) (~chomwitt@2a02:587:7a11:6600:1ac0:4dff:fedb:a3f1) | 10:29 | |
| hramrach | josch: the gitlab web ui may show garbage, don't trust it | 10:59 |
| josch | oh? o0 | 11:00 |
| hramrach | Seems to be triggered by having a PR open while it's updated | 11:30 |
| hramrach | I probably don't have the latest gitlab, maybe they fixed it. Probably related to some in-browser cached data in local storage. When it comes to debugging in-browser JavaScript applications I had 0 success so far. | 11:34 |
| - GNUmoon (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 11:34 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 11:35 | |
| - mjw (QUIT: Killed (iridium.libera.chat (Nickname regained by services))) (~mjw@gnu.wildebeest.org) | 11:49 | |
| * Guest3510 -> mjw | 11:49 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 12:18 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 12:18 | |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@64.189.201.150) | 12:18 | |
| - Gooberpatrol_66 (QUIT: Ping timeout: 260 seconds) (~Gooberpat@user/gooberpatrol66) | 12:18 | |
| + jacobk (~quassel@64.189.201.150) | 12:18 | |
| minute | josch: now testing !29 on a311d pocket | 12:23 |
| AbortRetryFail | upgrade time let's roll :) | 12:26 |
| minute | josch: your !29 of reform-debian-packages probably doesn't reference the staging branch of reform-tools, right? | 12:26 |
| minute | josch: ah i found a little error "linux/arch/arm64/boot/dts/freescale/meson-g12b-bananapi-cm4-mnt-pocket-reform.dts" this should be amlogic, and also the dtb is not added to the linux makefile | 12:29 |
| minute | josch: but good news, with my dtb manually put in place, the kernel from your branch works nicely on a311d pocket and display comes on | 12:40 |
| josch | minute: a few hours ago i started a reform-debian-packages job for the rk3588-reform2-josch branch (that's MR 29) with the staging branch of reform-tools: https://source.mnt.re/reform/reform-debian-packages/-/pipelines/1790 | 12:53 |
| minute | josch: niiice | 12:54 |
| josch | minute: when that had finished an hour ago i also started a reform-system-image pipeline run pulling in the packages from that branch and the results are here: https://source.mnt.re/reform/reform-system-image/-/jobs/3599/artifacts/browse | 12:54 |
| minute | josch: great. so we only need to fix the dtb thingie. and i have one more issue i wanted to discuss. on pocket reform, we need to add the kernel parameter "fbcon=rotate:3". but currently we have no way to detect in u-boot hook that we're on pocket | 13:05 |
| josch | minute: like this? https://source.mnt.re/reform/reform-debian-packages/-/merge_requests/29/diffs?commit_id=c8fc1515b4e3b52bd979ae3dc4c0d3f54b8da970 | 13:06 |
| josch | minute: but we are building a different u-boot for the pocket, right? | 13:07 |
| josch | and that u-boot can then set fbcon=rotate:3 in its ${bootargs} which will be picked up by boot.scr or extlinux.conf | 13:07 |
| minute | josch: yes, thanks for that (dtb) | 13:07 |
| + chomwitt (~chomwitt@2a02:587:7a11:6600:1ac0:4dff:fedb:a3f1) | 13:07 | |
| minute | josch: sure, that would be possible @ u-boot. it will add another 2 uboot builds (for now). | 13:10 |
| minute | but there's another issue here | 13:10 |
| minute | if someone buys rcm4 with bpi, we don't know if they will put it in pocket reform or reform | 13:10 |
| josch | i see what you mean but the example is not a good one because the tools disallow flashing u-boot to the a311d :) | 13:11 |
| minute | exactly | 13:11 |
| josch | but it would happen with imx8mp i guess | 13:11 |
| minute | so we can only fix the console for that by adjusting boot.scr | 13:12 |
| minute | i mean, it could be kind of like the dtb switcheroo for imx8mq with dual display vs single? | 13:13 |
| minute | or the system images for pocket override the dtb in boot.scr which in turn should make the machine id sticky | 13:14 |
| josch | ooooh you are talking about the configured dtb not the bootargs | 13:15 |
| minute | hmm that, too | 13:15 |
| minute | both | 13:15 |
| minute | :D | 13:15 |
| josch | but bootargs should be handled transparently these days | 13:15 |
| minute | maybe we could use "chosen" in the dtb to fix the rotation | 13:15 |
| josch | we only added workarounds for platforms where old u-boot versions didn't populate ${bootargs} correctly in flash-kernel/ubootenv.d/00reform2_ubootenv | 13:15 |
| minute | i will test if "chosen" is respected, then we only have to make sure the right dtb is on the device | 13:16 |
| josch | about choosing the right dtb, i don't think it is too much to ask for users who want to plug in ther module into a different device to run flash-kernel with a different device name before switching it over, no? | 13:17 |
| minute | josch: yeah that's ok, or they would download the system image for pocket/a311d | 13:18 |
| josch | for flash-kernel, it's the FK_MACHINE environment variable which allows it to populate /boot for a different machine than the one it is currently running on | 13:20 |
| josch | minute: one lesson learned from flash-kernel/ubootenv.d/00reform2_ubootenv is, that it's important to make sure that the created u-boot binary populates ${board} with something meaningful. For example for a311d, the value is just "u200" which is not very helpful. :) | 13:23 |
| josch | so maybe now that you are creating new u-boot binaries, make sure that ${board} is set to a value that specifies both the SoM as well as whether it's for the big reform or the pocket? | 13:24 |
| minute | josch: so far we are not even deploying the a311d uboot anywhere | 13:36 |
| minute | josch: we just ship the vendor uboot | 13:36 |
| minute | ok, chosen { bootargs } in dtb is not respected ;/ | 13:44 |
| minute | josch: can we adjust ubootenv.d/00reform2_ubootenv somehow depending on /etc/flash-kernel/machine? | 13:47 |
| josch | minute: maybe i don't understand the problem yet. You are currently shipping a a311d u-boot for the big reform. What stops you from building one for the pocket reform as well and let that u-boot export different ${bootargs}? | 13:52 |
| josch | Also, when you are building an a311d u-boot for the pocket, it can set a different ${board} which can then used by ubootenv.d/00reform2_ubootenv | 13:52 |
| josch | but since the boot.scr created via ubootenv.d/00reform2_ubootenv already respects ${bootargs} that should not be necessary | 13:52 |
| minute | josch: ah, it's a lapse in my thinking. i forgot that we are _not_ booting from emmc on a311d, but we're indeed using our own uboot | 13:53 |
| josch | minute: hence my "your example is bad" from above :) | 13:54 |
| minute | yeah | 13:54 |
| minute | sorry, i was confused by all the m * n variants | 13:54 |
| josch | but you are booting from emmc for imx8mp so you might have a problem there again | 13:54 |
| josch | yeah, i mean that's why you wanted to keep the matrix small in the first place XD | 13:54 |
| minute | yep, but i plan to reflash that uboot for shipping, and imx8mp for reform will be rare | 13:54 |
| josch | it becomes confusing very fast :D | 13:54 |
| josch | okay, cool! | 13:54 |
| minute | josch: ok, then my TODO is to split the a311d u-boot build for reform + pocket? | 13:56 |
| josch | yes | 13:58 |
| josch | and provide imx8mp u-boot for reform + pocket | 13:58 |
| josch | and provide rk3588 u-boot | 13:58 |
| josch | as far as i'm concerned, my further work on reform-system-image to support more boards is blocked on all the u-boots being available :) | 13:58 |
| minute | ok, written all of these down | 13:58 |
| minute | aye aye | 13:59 |
| josch | essentially, there are all the elements of the matrix you want to have supported: https://source.mnt.re/reform/reform-tools/-/tree/staging/modules?ref_type=heads | 14:00 |
| minute | great | 14:00 |
| josch | and when you go over the list, you see that we only have u-boot for ls1028a-reform, a311d-reform, imx8mq-reform and imx8mq+hdmi-reform | 14:00 |
| minute | ok i'll have some food and then i'll make all those uboots happen | 14:01 |
| josch | so we have four and we are missing four | 14:01 |
| josch | wow, awesome! :D | 14:01 |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 14:22 | |
| josch | whoops, I forgot to add support for MNT Pocket Reform with BPI-CM4 Module and MNT Reform 2 with i.MX8MP Module to flash-kernel -- fixing... | 14:22 |
| - jacobk (QUIT: Ping timeout: 268 seconds) (~quassel@64.189.201.150) | 14:23 | |
| josch | minute: should I add http://dump.mntmn.com/imx8mp-mnt-reform2.dts to reform-debian-packages MR 29? | 14:24 |
| minute | josch: oh that would be nice yes | 14:26 |
| + jacobk (~quassel@64.189.201.150) | 14:29 | |
| minute | interesting, noticed only now that emacs' diff mode supports messing with the contents of the patch file | 14:31 |
| minute | i.e. it adjusts @@ -213,6 +213,7 to @@ -213,6 +213,8 if i add a line | 14:31 |
| + mark_ (~mjw@gnu.wildebeest.org) | 14:35 | |
| josch | that is quite the killer feature! | 14:38 |
| josch | turns out, rk3588-mnt-reform2.dts also wasn't built yet XD | 14:38 |
| josch | (dts is in the git but not copied into the build tree and no patching of the Makefile) | 14:39 |
| minute | :0 | 14:40 |
| minute | josch: https://source.mnt.re/reform/reform-a311d-uboot/-/merge_requests/1 (not sure yet if it builds/works) | 14:47 |
| minute | josch: about "u200" i remember i tried to change it once but the build didn't work then | 14:49 |
| minute | i probably did it wrong | 14:49 |
| josch | yeah, i remember having a similar experience | 14:50 |
| josch | other than that, looks good! | 14:50 |
| josch | minute: we have a build failure after adding imx8mp-mnt-reform2.dts: | 14:55 |
| josch | /<<PKGBUILDDIR>>/arch/arm64/boot/dts/freescale/imx8mp.dtsi:1847.34-1849.9: ERROR (phandle_references): /soc@0/bus@32c00000/dsi@32e60000/ports/port@0/endpoint: Reference to non-existent node or label "lcdif_out" | 14:55 |
| josch | also defined at /<<PKGBUILDDIR>>/arch/arm64/boot/dts/freescale/imx8mp-mnt-reform2.dts:810.22-812.6 | 14:55 |
| minute | oh huh | 14:56 |
| minute | looking into it | 14:56 |
| josch | i pushed all my changes to the reform-tools staging branch and the reform-debian-packages rk3588-reform2-josch branch -- i'm now afk until later tonight :) | 14:57 |
| - mark_ (QUIT: Quit: Leaving) (~mjw@gnu.wildebeest.org) | 14:57 | |
| - chomwitt (QUIT: Ping timeout: 272 seconds) (~chomwitt@2a02:587:7a11:6600:1ac0:4dff:fedb:a3f1) | 14:57 | |
| minute | josch: ok for imx8mp, port@0 is already in the dtsi | 14:58 |
| minute | (of mipi_dsi) | 14:58 |
| josch | minute: feel free to push into the rk3588-reform2-josch branch to fix things :) | 14:59 |
| minute | ok :0 | 15:00 |
| minute | josch: initial u-boot for imx8mp setup https://source.mnt.re/reform/reform-imx8mp-uboot | 15:23 |
| minute | (pocket and reform2) | 15:23 |
| - S0rin (QUIT: Quit: WeeChat 3.8) (~S0rin@user/s0rin) | 16:24 | |
| + chomwitt (~chomwitt@2a02:587:7a11:6600:1ac0:4dff:fedb:a3f1) | 16:32 | |
| minute | josch: https://source.mnt.re/reform/reform-a311d-uboot/-/jobs/3629/artifacts/browse | 16:39 |
| josch | minute: nice! | 16:40 |
| josch | minute: Can we make the naming a bit more homogenous? Currently we have the device tree model name, the device tree dtb path name and the reform-system-image name which all refer to a combination of SoM and platform. | 16:41 |
| josch | Instead of adding a new group of names for the u-boot images, maybe some of the existing names could be copied | 16:42 |
| minute | josch: yes, i also noticed this. i'll gladly copy the dtb names | 16:42 |
| minute | josch: "flash" at the end or at the beginning of the string? | 16:42 |
| josch | uff :D | 16:42 |
| josch | my gut feeling is to put it at the end | 16:43 |
| josch | but i have no objective reasoning which makes me prefer that | 16:43 |
| minute | ok | 16:45 |
| minute | imx8mp-mnt-reform2-flash.bin | 16:45 |
| minute | etc | 16:45 |
| minute | meson-g12b-bananapi-cm4-mnt-pocket-reform-flash.bin | 16:47 |
| minute | etc | 16:47 |
| josch | minute: also, currently flash.bin is a real file -- what happens if you make it a symlink? Does it then show up in the artifact list as a symlink to X? That would make it obvious what this flash.bin is compared to the other files. | 16:51 |
| minute | josch: optimally i want to get rid of that flash.bin | 16:52 |
| minute | josch: i just didn't know yet if that breaks anything for you | 16:52 |
| josch | i don't think it breaks something to not have it | 16:52 |
| josch | because the existing scripts, download from existing tags | 16:52 |
| minute | ok, gonna remove it | 16:52 |
| josch | so if somebody has old versions of their stuff, they will just download flash.bin from an old tag and it will still exist there | 16:53 |
| josch | minute: while we are at naming stuff... now that you are about to have eight different configurations supported, what about naming the system image after the dtb name as well? | 16:54 |
| josch | imx8mp-mnt-reform2-system.img | 16:54 |
| josch | meson-g12b-bananapi-cm4-mnt-reform2-system.img | 16:54 |
| josch | doing so will probably break something in the links you have pointing to the artifacts from the main websites though | 16:55 |
| minute | this will break many things yeah, also the new handbook | 16:55 |
| minute | i think it's better to have these user-facing paths less unwieldly. nobody knows what a meson g12b is | 16:56 |
| minute | the pocket reform system images shall just be pocket-reform-system-image-a311d.img and pocket-reform-system-image-imx8mp.img | 16:57 |
| minute | lol without the -image | 16:57 |
| minute | pocket-reform-system-a311d.img | 16:57 |
| minute | etc | 16:57 |
| josch | minute: https://paste.debian.net/1310807/ | 16:58 |
| josch | i'm currently refactoring mkimage.sh to accomodate for this, hence my asking | 16:59 |
| minute | josch: great, this looks nice and simple to me | 17:00 |
| erle | naming different things differently is always a plus for cross-platform work | 17:07 |
| josch | erle: this was more a question of whether the same things should be named the same or not :) | 17:08 |
| josch | but "a311d" is probably a bit more approachable than "meson-g12b-bananapi-cm4" | 17:09 |
| josch | for what it's worth, the mapping between device tree model name, system image name, dtb name and now u-boot name is stored in a machine readable way in the files that will be shipped in /usr/share/reform-tools/modules/*.conf | 17:10 |
| minute | yeah that's nice | 17:11 |
| erle | josch do these three things have different naming schemes for other platforms? | 17:12 |
| erle | four things | 17:13 |
| josch | three things. dtb and u-boot images are named the same | 17:14 |
| - Gooberpatrol66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 17:14 | |
| josch | dtb names and device tree model names are the same elsewhere (theoretically, a lot not upstreamed yet) | 17:14 |
| josch | the only "made up" naming is the one of the system images | 17:15 |
| minute | aha > Image 'main-section' has faked external blobs and is non-functional: lpddr4_pmu_train_1d_imem_202006.bin lpddr4_pmu_train_1d_dmem_202006.bin lpddr4_pmu_train_2d_imem_202006.bin lpddr4_pmu_train_2d_dmem_202006.bin | 17:25 |
| minute | it's blobbin' time | 17:25 |
| mjw | what is so special about memory training that it needs to be a proprietary blob? :{ | 17:27 |
| minute | mjw: nothing, just chip industry things | 17:28 |
| mjw | sigh | 17:28 |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 17:29 | |
| mjw | Is there a project or a person who is looking at this? Would it be solvable by throwing money at the right people? | 17:32 |
| mjw | ACTION realizes that sounded disrespectful. But I mean, it isn't really something I could work on without needing to learn a lot of new things. But if someone already has that knowledge maybe I could do some cheerleading by adding a little money in the pot. | 17:33 |
| minute | mjw: by the time someone has figured that out, the soc will be obsolete | 17:35 |
| mjw | grrr | 17:35 |
| hramrach | it's been done for some SoCs before being completely obsolete | 17:36 |
| minute | mjw: sorry. it's just that this stuff is very thankless to RE because it's a strange ISA with vendor extensions and the registers are unknown | 17:37 |
| minute | it's basically the same blob in the imx8mq, imx8mp, and a311d | 17:37 |
| mjw | yeah, I understand. Just a little disappointing. | 17:38 |
| minute | because they all bought the same synopsys ddrc IP | 17:38 |
| minute | ls1028a does not have this issue | 17:38 |
| mjw | ACTION shakes fist at "chip industry" | 17:38 |
| mjw | And I just got one of those! :) | 17:38 |
| minute | ls1028a needs a blob for eDP though. i have some ideas for a display controller for it but i can't spend time on it because it would tank us financially | 17:38 |
| minute | (if i spent all my time on that and it would still be a niche product) | 17:39 |
| mjw | ACTION is afraid he doesn't have that amount of cash to throw at it | 17:39 |
| mjw | At least I hope your aren't selling your time cheaply. | 17:40 |
| minute | mjw: i am earning significantly less than i would in a normal tech job | 17:42 |
| hramrach | that's basically why it's been done for some SoCs before being completely obsolete - they reused some memory controller IP from now obsolete SoCs | 17:43 |
| minute | i earn around 2680 eur/mo net and my apartment rent is around 1200 eur/mo (for around 35qm) | 17:45 |
| mjw | And that is also a little infuriating. Free Software and Open Hardware are unaffordable :{ | 17:45 |
| mjw | [Working on]... | 17:46 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 17:56 | |
| minute | josch: ok finally, imx8mp flash bins https://source.mnt.re/reform/reform-imx8mp-uboot/-/jobs/3642/artifacts/browse | 18:02 |
| - chomwitt (QUIT: Ping timeout: 260 seconds) (~chomwitt@2a02:587:7a11:6600:1ac0:4dff:fedb:a3f1) | 18:36 | |
| + mark_ (~mjw@82-217-174-174.cable.dynamic.v4.ziggo.nl) | 18:47 | |
| * mjw -> Guest7107 | 18:48 | |
| - Guest7107 (QUIT: Killed (silver.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 18:48 | |
| * mark_ -> mjw | 18:48 | |
| + Guest7107 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 18:48 | |
| - mjw (QUIT: Ping timeout: 268 seconds) (~mjw@82-217-174-174.cable.dynamic.v4.ziggo.nl) | 18:54 | |
| + mjw (~mjw@gnu.wildebeest.org) | 19:03 | |
| - Guest7107 (QUIT: Ping timeout: 255 seconds) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:42 | |
| + sevan (~sevan@2001:470:1f1d:1d6:5a55:caff:fe24:ed4) | 20:08 | |
| minute | josch: here's my new prototype for setting up reform-setup and tuigreet in the system image, looking for your advice to make this more debian conformant | 20:22 |
| minute | josch: https://source.mnt.re/reform/reform-system-image/-/commit/82e059cf7c0044171b3778b5f20d77936693c3f9 | 20:23 |
| minute | josch: better: https://source.mnt.re/reform/reform-system-image/-/merge_requests/92 | 20:29 |
| - jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@64.189.201.150) | 20:29 | |
| - sevan (QUIT: Changing host) (~sevan@2001:470:1f1d:1d6:5a55:caff:fe24:ed4) | 20:35 | |
| + sevan (~sevan@user/venture37) | 20:35 | |
| + jacobk (~quassel@utdpat241106.utdallas.edu) | 20:47 | |
| josch | minute: now playing Stardew Valley in multiplay on my reform, then reviewing your MR. Looking forward! :) | 21:13 |
| josch | *multiplayer | 21:14 |
| minute | josch: niiice | 21:18 |
| minute | i'm also now in feierabend but first trying to import a midi file into my DAW format | 21:18 |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@utdpat241106.utdallas.edu) | 21:37 | |
| + chomwitt (~chomwitt@2a02:587:7a11:6600:1ac0:4dff:fedb:a3f1) | 22:16 | |
| + athena (~athena@i577B8B57.versanet.de) | 23:46 | |
| * athena -> athenametro | 23:47 | |
| - chomwitt (QUIT: Ping timeout: 260 seconds) (~chomwitt@2a02:587:7a11:6600:1ac0:4dff:fedb:a3f1) | 23:47 | |
| athenametro | Will people with an MNT Reform be at the easterhegg in Regensburg at the end of this month? I'd love to see one in person | 23:49 |
| + jacobk (~quassel@64.189.201.150) | 23:52 | |
| vkoskiv | I'll be in Germany, in Saarbrücken | 23:57 |
| vkoskiv | s/in/but in/ | 23:58 |
| vkoskiv | That sounded silly without the but :D | 23:58 |
| vkoskiv | I was too focused on the ü | 23:58 |
| erle | maybe | 23:59 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!