| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.33.115) | 01:00 | |
| + Ar|stote|is (~linx@149.210.33.115) | 01:00 | |
| - mtm (QUIT: Ping timeout: 252 seconds) (~textual@47.202.75.129) | 01:04 | |
| + mtm (~textual@47.202.75.129) | 01:06 | |
| - Gooberpatrol_66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 01:16 | |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 01:17 | |
| - wytch (QUIT: Quit: Leaving.) (~wytch@user/wytch) | 02:01 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~textual@47.202.75.129) | 02:07 | |
| + mtm (~textual@47.202.75.129) | 02:09 | |
| + ZylonMaster (~hjcs@syn-098-015-248-249.res.spectrum.com) | 02:16 | |
| - ZylonMaster (QUIT: Client Quit) (~hjcs@syn-098-015-248-249.res.spectrum.com) | 02:19 | |
| - mjw (QUIT: Ping timeout: 248 seconds) (~mjw@gnu.wildebeest.org) | 02:23 | |
| - paperManu (QUIT: Ping timeout: 252 seconds) (~paperManu@107.159.71.33) | 02:25 | |
| + Guest28 (~Guest28@38.145.3.251) | 02:25 | |
| - Guest28 (QUIT: Client Quit) (~Guest28@38.145.3.251) | 02:28 | |
| - Ar|stote|is (QUIT: Ping timeout: 260 seconds) (~linx@149.210.33.115) | 02:49 | |
| - kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 02:57 | |
| + kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 02:58 | |
| - kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 03:04 | |
| + kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 03:05 | |
| - kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 03:06 | |
| + kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 03:06 | |
| - nsc (QUIT: Ping timeout: 252 seconds) (~nicolas@74-97-142-46.pool.kielnet.net) | 03:55 | |
| + nsc (~nicolas@85-97-142-46.pool.kielnet.net) | 03:56 | |
| + colinsane (~colinunin@97-113-133-123.tukw.qwest.net) | 05:05 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 05:06 | |
| - Gooberpatrol_66 (QUIT: Ping timeout: 252 seconds) (~Gooberpat@user/gooberpatrol66) | 05:07 | |
| + Ar|stote|is (~linx@149.210.33.10) | 07:48 | |
| - exark (QUIT: Read error: Connection reset by peer) (~exark@user/exark) | 07:55 | |
| + exark_ (~exark@user/exark) | 07:55 | |
| josch | currently (due to how the scripts developed organically) we have two places that list the packages that belong into a reform system image: in reform-system-image itself and in reform-check which prints a message if one of the required packages is not installed. This is suboptimal as there are two places with the same list. I'm looking for ideas of how to improve this. | 07:56 |
|---|---|---|
| josch | My current idea is to add a meta-package called task-reform-desktop to the MNT and reform.d.n repos. That package comes in versions 12, 13 and 13~bpo12 to account for the different package sets in stable, stable-backports and stable+1 and each repo would carry the respective version. | 07:58 |
| josch | with such a setup, reform-check could drop its package list and just check whether task-reform-desktop is installed or not | 07:59 |
| josch | the package list would then live in the reform-debian-packages repository instead of in the reform-system-image repo | 07:59 |
| josch | also add a task-reform-desktop-minimal meta-package to support the --quick option of mkimage.sh | 07:59 |
| josch | ch: do you have any creative input maybe? :) | 08:00 |
| ch | from a quick thought i think the package versions as indicated might become a problem and i'd avoid that | 08:02 |
| ch | (cant articulate the envisioned problem atm) | 08:02 |
| ch | moving the list into a package Depends: line makes sense to me | 08:03 |
| ch | iirc d-i wants task-* packages to delegate everything to another metapackage, so maybe dont call it task-* | 08:03 |
| + aloo_shu (~aloo_shu@84.78.242.52) | 08:11 | |
| - aloo_shu (QUIT: Remote host closed the connection) (~aloo_shu@84.78.242.52) | 08:12 | |
| josch | ch: remember that i also maintain a d-i patch to install a reform system so my medium-term goal was actually exactly that: add another task to the installer. What do you mean by "delegate everything to another metapackage"? Do you remember more details? | 09:23 |
| josch | ch: about the version numbers you probably mean that they are missing a debian revision or a minor version? | 09:37 |
| ch | i'll try to find the bugs | 09:37 |
| josch | thank you! | 09:38 |
| ch | https://salsa.debian.org/installer-team/tasksel/-/merge_requests/38#note_552924 | 09:40 |
| ch | https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1088614 | 09:40 |
| ch | basically: add another layer of indirection | 09:41 |
| josch | ooooh that makes a lot of sense | 09:43 |
| josch | ch: thank you for spotting this potential faux pas | 09:44 |
| josch | then lets call them reform-desktop-minimal (which depends on the basic stuff) and reform-desktop-full (which depends on reform-desktop-minimal and some more extras) | 09:44 |
| ch | wild thought: reform-desktop-wayfire, reform-desktop-sway, (+ a set of basics packages) | 09:45 |
| josch | yes, i thought about that as well but then the difference between those two is not that large and i wonder if the few packages that do not get installed for each variant is worth an extra meta package | 09:46 |
| josch | also thinking about reform-desktop-gnome which is supposed to replace wayfire in the future | 09:46 |
| ch | yeah thats a question of what do you want | 10:03 |
| ch | should users get these changes, or should they keep their setup | 10:04 |
| josch | this is not going to change existing setups | 10:10 |
| josch | the meta packages will be installed by reform-system-image for new setups | 10:10 |
| ch | but 'old' setups from new reform-system-image-s will also have it, no? | 10:13 |
| josch | what are old setups from new reform-system-image? | 10:19 |
| ch | well if you add the meta package, say, reform-desktop to a reform-system-image, i install from that, and in three months gnome replaces wayfire, should i get that change? | 10:22 |
| - buckket (QUIT: Quit: buckket) (~buckket@vps.buckket.org) | 10:31 | |
| + buckket (~buckket@vps.buckket.org) | 10:32 | |
| josch | yes, because at the point where gnome replaces wayfire, those who have reform-desktop installed will upgrade their version of reform-desktop which will pull in gnome and consequently apt will offer to auto-remove wayfire | 10:32 |
| minute | sounds sophisticated | 10:34 |
| ch | yeah, just asking if this is really the way to go | 10:34 |
| ch | dotfiles are also part of this picture | 10:34 |
| [tj] | is wayfire dead? | 10:35 |
| BoostisBetter | I sure hope not. I think wayfire is better than Sway because it gives you free form windows super easily tiling. You can use wayfire like Sway but not have any weird windowing issues since it is not tiling windows by default. | 10:37 |
| josch | ch: i'm still thinking about dotfiles as well but i first would like to see the changes necessary for a default gnome setup for MNT as that might require features from dot files that are not yet possible and for which we need to create new things (this is the past discussion where you brought up the idea of jinja2 templating which i still like the most) | 10:39 |
| minute | [tj]: https://community.mnt.re/t/gnome-integration-notes-issue-tracking/3168 | 10:39 |
| minute | BoostisBetter: https://community.mnt.re/t/gnome-integration-notes-issue-tracking/3168 | 10:39 |
| josch | BoostisBetter: this is not a "wayfire is dead" but a "wayfire is more dead than gnome and MNT might want to focus on the less dead desktop" | 10:40 |
| minute | BoostisBetter: [tj]: would you prefer wayfire over gnome? | 10:40 |
| [tj] | I've only used it for the 5 days before my battery board got sad | 10:41 |
| BoostisBetter | minute: 100% but I was having some issues and ended up using Sway | 10:41 |
| BoostisBetter | It was wayfire based though. I forget what the issue was now | 10:41 |
| minute | BoostisBetter: hmm i can't follow. | 10:42 |
| [tj] | I'll always lean to whatever is the minimum maintenance burden for you | 10:43 |
| minute | anyways i can tell you that a gnome-shell desktop with keyboard shortcuts is much more stable and usable than any customized wayfire we could come up with | 10:43 |
| BoostisBetter | minute: what I mean is that I prefer wayfire. However, it seems like development on wayfire has slowed significantly and as a result an issue I was seeing on the pocket with wayfire wasn't going to be getting fixed anytime soon. I can't remember what the bug was. I can try to dig it up tonight though. So as a result of that I am using Sway currently on the Pocket. But only because I value stabili | 10:43 |
| BoostisBetter | ty above everything else on the Pocket. I prefer wayfire. | 10:43 |
| BoostisBetter | minute: what I mean is that I prefer wayfire. However, it seems like development on wayfire has slowed significantly and as a result an issue I was seeing on the pocket with wayfire wasn't going to be getting fixed anytime soon. I can't remember what the bug was. I can try to dig it up tonight though. So as a result of that I am using Sway currently on the Pocket. But only because I value stabili | 10:44 |
| BoostisBetter | ty above everything else on the Pocket. But in general I prefer wayfire. | 10:44 |
| minute | BoostisBetter: ok. if stability is your main thing you'll see that you'll be much happier with the gnome setup than with our wayfire setup | 10:44 |
| BoostisBetter | minute: any chance you could try plasma 5 or 6? | 10:45 |
| minute | and also ease of use and configuration | 10:45 |
| minute | BoostisBetter: not right now. kde is much bigger and more complex than gnome | 10:45 |
| BoostisBetter | minute: yeah I have used Gnome for a long time. I know it quite well. I do think it is a little top heavy though. I like how lightweight Sway and Wayfire are. | 10:45 |
| josch | (kde plasma 6 works without problems on imx8mq) | 10:46 |
| BoostisBetter | josch: that is cool to hear. | 10:46 |
| BoostisBetter | I am using Plasma 5 on my main laptop | 10:46 |
| [tj] | I'm on plasma 5 too | 10:46 |
| [tj] | maybe I'll use kde on the pocket | 10:46 |
| [tj] | when I moved to kde from i3 a few years back it took a couple of weeks of customization to make it behave in a sensible way | 10:47 |
| minute | we can look at plasma 6 again a bit later. one of the most important thing about the gnome move is getting the login manager right. i think installing kde on top of that for individual users will be easier than on the current setup. | 10:47 |
| minute | [tj]: but you don't like sway? | 10:48 |
| [tj] | I think a decade of tiling window managers was enough for me and I'd like to have windows | 10:48 |
| minute | ah | 10:48 |
| [tj] | I found it wasted a ton of space on a high resolution monitor which I could use to look at the grass in my wall paper | 10:49 |
| [tj] | I started making i3 do windowing and realised how stupid that was | 10:49 |
| + mjw (~mjw@gnu.wildebeest.org) | 10:49 | |
| BoostisBetter | minute: I responded to that thread you linked with what I have said here just to have it captured there in the hopes you find it useful. | 10:52 |
| - GNUmoon (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 10:53 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 10:53 | |
| BoostisBetter | minute: also sorry for the politics on the forums. I agree that it is best to leave them out. | 10:55 |
| + xandyv (xandyv@user/xandyv) | 11:31 | |
| grimmware | I've been considering Niri for the pocket (I just haven't gotten round to doing anything about it yet) because I tend to find tiling more of an annoyance on such a small screen and my brain tends toward directional navigation and remembering that kind of layout better than it can remember which numbered workspace things are on | 11:37 |
| grimmware | well, I say that, it's just that browsers and electron apps have opinions about how much you should be able to resize a window before they start doing dumb shit and wasting space | 11:38 |
| BoostisBetter | agreed. Electron is great for making applications work across several platforms but I also hate how they've kind of become the default design choice. | 11:43 |
| + paperManu (~paperManu@107.159.71.33) | 11:45 | |
| josch | is anything happening in the area of accessebility in electron? I.e. you have no or bad sight and need to use a screenreader or don't or cannot use a mouse cursor to click on things? | 11:47 |
| grimmware | Honestly you'd think that the whole setup of "web technologies in a local app" would give itself over well to letting you access the backend separately and programatically so that people can just implement their own frontends but that would require the world to not be awful. | 11:56 |
| minute | BoostisBetter: thanks for understanding. i'm following US politics very closely every day. it's just too much of a distraction in the forum, as one could argue indefinite amounts about it. | 12:09 |
| grimmware | I'm glad that stuff's being discouraged on the forum. I'm struggling pretty hard with my mental health right now and it's good to have a place to talk about the thing we actually *do* have a lot of agency over (our Reforms) instead of the things we don't. | 12:24 |
| - Ar|stote|is (QUIT: Ping timeout: 252 seconds) (~linx@149.210.33.10) | 12:42 | |
| - mjw (QUIT: Ping timeout: 272 seconds) (~mjw@gnu.wildebeest.org) | 12:46 | |
| + Ar|stote|is (~linx@149.210.32.164) | 12:47 | |
| * Guest3390 -> mjw | 12:50 | |
| - mtm (QUIT: Ping timeout: 245 seconds) (~textual@47.202.75.129) | 13:02 | |
| + gustav28 (~gustav@c-78-82-55-198.bbcust.telenor.se) | 13:02 | |
| + mtm (~textual@47-202-75-129.fdr01.sprg.fl.ip.frontiernet.net) | 13:05 | |
| hramrach | josch: the problem with a metapackage is that it's binary: either it's installed or not. On the other hand, a list of packages is something that can be checked, and the user can determine iof they care about that particular piece on their own. | 13:15 |
| hramrach | it's not a bad idea to create a metapackage out of the package list but the package list should still exist separately from the metapackage I think | 13:16 |
| [tj] | minute: I got the replacement charger board and a surprise new rear display panel! Thank you! | 13:21 |
| hramrach | minute: When 'ease of configuration' is broght up with GNOME it kind of sounds to me like configuration becomes easy because there is nothing to configure. Also ease of use is disputable. For any GNOME version above 2 I found it very janky when I tried to use it for whatever reason. Like the way I approach commercial desktops: with enough workarounds it can be used but there is no way to get | 13:24 |
| hramrach | the problems fixed, and workarounds removed, ever. And more breakage can happen arbitrarily and irrevocably as well. | 13:24 |
| hramrach | GNOME project are the people who know better than you how you should use your comouter, and the answer to many problem reports is basically 'you should not want that' | 13:31 |
| minute | hramrach: did you read my post? | 14:00 |
| mjw | btw. my experience is the opposite, before gnome3 I felt I was supposed to keep tweaking it because stuff just didn't work out of the box. Since I haven't felt the need to tweak anything really. | 14:00 |
| minute | hramrach: also, gnome is just an option to replace wayfire which i also heavily customized (even had to fork and extend a window decoration plugin) | 14:01 |
| mjw | That said, I use sway on my reform atm | 14:01 |
| minute | hramrach: i think most opinionated people will continue using sway or install their own compositor/DE of choice. this is more to have a good experience for less advanced users | 14:02 |
| minute | and also i don't wanna continue customizing wayfire if gnome is just better with less maintenance effort for me... it just makes more sense | 14:03 |
| minute | also i'm working with gnome devs to iron out any performance bugs and bugs related to the pocket's small screen size. there's still work going into this, we don't just take something thrown over a wall | 14:03 |
| minute | grimmware: hmm without that sleep in disable it looks like i'm getting the flicker/oscillation in my v1 display as well on pocket after having the internal screen turned off in gnome for a while, then switching back to it. so we should revisit this soonish. an option would be to make sure that the wake from sleep is done symmetrically (i.e. in some enable function). but i also got snow | 14:25 |
| minute | sometimes re-enabling the display, so i'd be willing to set aside a few hours/days to tweak the delays etc more to make this really solid | 14:25 |
| grimmware | Oof I never got snow. | 14:27 |
| grimmware | I’m really keen to get this sorted. I also worked a little on i2c over spi a little the other day on a long train journey, I’m thinking about sharing the WIP so I can get some thoughts on the approach. | 14:28 |
| grimmware | That would enable me to use upstream sysctl and reform2_lpc and do all the logic in userland | 14:30 |
| hramrach | minute: if you want option for less advanced users thre is KDE/MATE/XFCE/... anything that gives the common DE semantics people are used to if they ever used a PC before | 14:55 |
| hramrach | there is a lot of money poured into GNOME development so it does not have some of the problems the other DEs have due to lack of development bandwidth. It has one major problem though: it does not make any sense. | 15:00 |
| Zaba | that is a lot of words to say that you don’t like gnome. I don’t think any default is going to be universally liked though | 15:36 |
| hramrach | there are things that are a matter of like and dislike, and some that are bad design | 15:48 |
| minute | haha i knew this could end in a flamewar like in the olden days | 15:54 |
| minute | hramrach: you know that there are extensions and many config options through gnome tweaks yes? i also disagree with some gnome defaults, but it is quite customizable | 15:55 |
| hramrach | and you know that there is no stable exension API and they break all the time? | 15:56 |
| minute | yes | 15:56 |
| minute | same for wayfire | 15:56 |
| minute | but worse for wayfire because extensions have to be packaged as binaries and take down the compositor | 15:56 |
| minute | hramrach: i forgot, are you using wayfire on the (pocket) reform? | 15:57 |
| hramrach | I am not arguing for wayfire, though. I am arguing that if you want 'desktop environment for less advnaced users' gnome is not a good choice, for reasons of bad design choices on their part. | 15:58 |
| minute | hramrach: so you're using it as the main environment? | 15:58 |
| hramrach | I am using sway | 15:58 |
| minute | hramrach: so why are we having this discussion | 15:58 |
| minute | why do you care about it if you're not gonna use it | 15:59 |
| hramrach | In my view if something like MATE or XFCE was available as one of the defaults that would make Reform truly usable for 'lesss advanced users' out of thebox, unlike GNOME | 16:00 |
| minute | hramrach: so _you_ know what's better for other people but if gnome makes design decisions they're the baddies? | 16:00 |
| minute | hramrach: in any case, your input is noted. i already said that kde can and should be a future option. xfce is not wayland ready yet. i'm not going back to x11. people who want x11 can install it themselves. | 16:03 |
| minute | shit, now i missed the bus because i was too caught up in IRC | 16:04 |
| BoostisBetter | minute: I agree. X11 is a done deal and while I don't actively support it, I am not against it. Gnome has a bunch of organizational issues right now that have we worried about future development but outside of this I haven't had an issue with Gnome. I always took most Linux distros using it to be a sign of its competence. In fact, the Reform running Sway was the first time I even tried another DE | 16:17 |
| BoostisBetter | . | 16:17 |
| BoostisBetter | All that to say that I really appreciate the effort to give us options. | 16:18 |
| + patsavoura (~linx@5.203.168.244) | 16:27 | |
| - Ar|stote|is (QUIT: Ping timeout: 268 seconds) (~linx@149.210.32.164) | 16:29 | |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-133-123.tukw.qwest.net) | 17:14 | |
| jfred | Hmm, I just tried plugging my Pocket Reform into my solar panel with a type-C output and I just get the letter T flashing on the top left hand side of of the OLED display with the battery screen open. Anyone know what that means? | 17:15 |
| + colinsane (~colinunin@97-113-133-123.tukw.qwest.net) | 17:18 | |
| BoostisBetter | Means the controller timed out waitingon the system controller. Usually you can ignore it. | 17:19 |
| jfred | ah interesting. it also doesn't seem to charge from the solar panel I should note | 17:23 |
| jfred | I know PD compatibility has been a struggle | 17:24 |
| - colinsane (QUIT: Ping timeout: 245 seconds) (~colinunin@97-113-133-123.tukw.qwest.net) | 17:26 | |
| + LainExperiments (~LainExper@user/LainExperiments) | 17:27 | |
| frickler | jfred: do you know if the panel really provides PD or "only" normal 5V? you could also look at debugging output from the sysctl, I'm not sure if that's properly documented somewhere | 17:31 |
| - patsavoura (QUIT: Read error: Connection reset by peer) (~linx@5.203.168.244) | 17:31 | |
| jfred | It claims to provide PD at least (I don't have my charging monitor handy but my Fairphone 5 said it was "charging rapidly" when plugged into it) | 17:33 |
| jfred | would be glad to poke at some debug info if you have any pointers for that | 17:33 |
| frickler | jfred: so essentially you would need to compile your own sysctl fw with https://source.mnt.re/reform/pocket-reform/-/blob/main/pocket-reform-sysctl-fw/inc/sysctl.h?ref_type=heads#L21 commented out and flash that | 17:39 |
| + colinsane (~colinunin@97-113-83-26.tukw.qwest.net) | 17:40 | |
| frickler | the script for flashing I think is still broken with the new USB ids, so it needs some manual fiddling | 17:40 |
| - colinsane (QUIT: Client Quit) (~colinunin@97-113-83-26.tukw.qwest.net) | 17:40 | |
| frickler | and in the end then you can run something like "sudo tio -t -L /dev/ttyACM0" and get details about the charger | 17:41 |
| + Ar|stote|is (~linx@5.203.176.236) | 17:45 | |
| jfred | frickler: ah excellent, love a nice little serial debug interface :) and for what it's worth I just flashed the current sysctl fw last night and didn't have trouble with USB IDs (but maybe I will now that I have a newer fw version installed?) | 17:51 |
| - Ar|stote|is (QUIT: Read error: Connection reset by peer) (~linx@5.203.176.236) | 17:53 | |
| frickler | jfred: hmm, how did you flash it? https://source.mnt.re/reform/pocket-reform/-/blob/main/scripts/update-sysctl-firmware.sh or using something different? | 17:55 |
| - LainExperiments (QUIT: Ping timeout: 240 seconds) (~LainExper@user/LainExperiments) | 17:55 | |
| jfred | frickler: oh, no, https://source.mnt.re/reform/pocket-reform/-/blob/main/pocket-reform-sysctl-fw/flash.sh?ref_type=heads - which I realize now looking at the script you posted was maybe slightly dangerous :D | 17:56 |
| hramrach | minute: If you want to discuss UI design I think we should do this on another platform, IRC is not very suitable with limited message lenght and no threading. The 'less advanced users' I have talked with have previous experience with PC computers which look and behave somewhat like Windows in the range 95~7, and DEs that are close to that are understandable to them while DEs that are far from | 18:03 |
| hramrach | that aren't. | 18:03 |
| + Ar|stote|is (~linx@149.210.12.242) | 18:14 | |
| minute | frickler: wait what why | 18:32 |
| minute | ah sorry. for acm debugging | 18:32 |
| minute | jfred: to get newer firmware into bootloader mode you can use reform-mcu-tool | 18:33 |
| minute | root@pocki3000:/home/minute# lsusb | grep -i pocket | 18:34 |
| minute | Bus 001 Device 003: ID 1209:6d06 Generic Pocket Reform Input 1.0 | 18:34 |
| minute | Bus 003 Device 002: ID 1209:6d07 Generic Pocket Reform System Controller 1.0 | 18:34 |
| minute | if you have these kind of IDs | 18:34 |
| minute | root@pocki3000:/home/minute# reform-mcu-tool list | 18:34 |
| minute | Target pocket-input-1.0 ID 1209:6d06 Serial# DE62549203135632 USB bus 1 address 3 | 18:34 |
| minute | Target pocket-sysctl-1.0 ID 1209:6d07 Serial# DE63544193622134 USB bus 3 address 2 | 18:34 |
| minute | so you can do: reform-mcu-tool bootsel pocket-sysctl-1.0 | 18:35 |
| jfred | minute: thanks :) | 18:47 |
| jfred | (pocki3000 is a fun name for a pocket reform. I'm imagining the littlest evil computer with that red eye on the back) | 18:48 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 18:49 | |
| jfred | wait, I had the wrong number. was thinking of HAL-9000, haha | 18:50 |
| josch | hramrach: my goal was to have one machine readable declarative way to store the list of packages that should go into the system image. I think it's a meta package, then I can parse its Depends: and/or Recommends: even if it's not installed to obtain that list so that it for example can be used in reform-help. So since the package meta data is machine readable I don't think I need to maintain yet another | 18:58 |
| josch | list with the same content somewhere? | 18:58 |
| hramrach | josch: if the metatada can be obtained without the package installed that's fine. Not sure how much information about the package is available with online repository access and offline | 19:02 |
| hramrach | https://chimera-linux.org/news/2025/03/new-riscv-server.html | 19:02 |
| hramrach | ^^^ "not dropping RISC-V repos for the time being" | 19:03 |
| josch | hramrach: unless you purge the package, dpkg will remember the dependencies even without being online | 19:06 |
| hramrach | for which version? The last installed? What if it was never installed? | 19:06 |
| hramrach | generally it sounds awkward and fragile to rely on this mechanism | 19:07 |
| josch | i'd assume that a system that was installed in the past was also online at some point and did an "apt update" and then the information is there as well | 19:08 |
| josch | yes, there are situation where the information will not be there but it's only a "nice to have" anyways | 19:08 |
| josch | reform-check has an --offline mode which will print less stuff for systems without internet | 19:09 |
| josch | the alternative is to maintain two lists as it is today | 19:10 |
| josch | and the drawback of that is, that the lists are rarely in sync | 19:10 |
| hramrach | I wonder if the list could be synced at build time | 19:11 |
| hramrach | rather than maintaining it twice | 19:11 |
| josch | yes, i was thinking about that too -- i'm still looking for ideas :) | 19:11 |
| - Gooberpatrol66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 19:12 | |
| hramrach | I think using apt data is better than dpkg if the package has to be used, apt is more likely to be up-to-date | 19:12 |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 19:13 | |
| hramrach | Can this metapackage be produced as part of reform-tools build perhaps? Then only one place for the data is needed. | 19:14 |
| hramrach | Is there much need for updating these separately? | 19:15 |
| josch | yes, reform-tools could include the list in some format and then both reform-help could use it and the meta-package could be built from it. I am a bit unsure about sticking it into reform-tools because the utility of this package is for the reform system images and I'm not so sure those meta packages have much use in the Debian archive. | 19:18 |
| hramrach | if people want to use upstream Debian installer or such they could be useful perhaps | 19:21 |
| josch | yes, though i'm not getting my hopes up. Including meta packages for blends like debian edu has been a todo item for over a decade... | 19:24 |
| hramrach | what's blocking it in that case? | 19:24 |
| + mark_ (~mjw@gnu.wildebeest.org) | 19:27 | |
| * mjw -> Guest4214 | 19:27 | |
| - Guest4214 (QUIT: Killed (zinc.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2486:a800:7602:5eff:dc71:a72c) | 19:27 | |
| * mark_ -> mjw | 19:27 | |
| + Guest4214 (~mjw@2001:1c06:2486:a800:7602:5eff:dc71:a72c) | 19:27 | |
| hramrach | except for the sheer number of packages, some with potentially conflicting names | 19:27 |
| ch | i think i saw some movement there recently | 19:30 |
| ch | hramrach: why is everybody making a big deal out of what chimera linux thinks? | 19:30 |
| hramrach | ch: because the previous article also posted here nicely sums up the frustration with the RiscV architecture | 19:32 |
| hramrach | this one is a followup so I think it should be included for completness | 19:33 |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 19:54 | |
| - Gooberpatrol66 (QUIT: Ping timeout: 276 seconds) (~Gooberpat@user/gooberpatrol66) | 19:55 | |
| - Gooberpatrol_66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 20:11 | |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 20:12 | |
| - mlarkin (QUIT: Ping timeout: 252 seconds) (~mlarkin@47.158.172.62) | 21:01 | |
| + mlarkin (~mlarkin@47.158.255.25) | 21:23 | |
| - hairu (QUIT: Remote host closed the connection) (m-uotkmd@user/hairu) | 21:24 | |
| + hairu (m-uotkmd@user/hairu) | 21:25 | |
| - hairu (QUIT: Remote host closed the connection) (m-uotkmd@user/hairu) | 21:35 | |
| + hairu (m-uotkmd@user/hairu) | 21:36 | |
| - hairu (QUIT: Remote host closed the connection) (m-uotkmd@user/hairu) | 21:37 | |
| + hairu (m-uotkmd@user/hairu) | 21:39 | |
| - hairu (QUIT: Remote host closed the connection) (m-uotkmd@user/hairu) | 21:40 | |
| + hairu (m-uotkmd@user/hairu) | 21:43 | |
| - xandyv (QUIT: Ping timeout: 244 seconds) (xandyv@user/xandyv) | 21:45 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 21:46 | |
| - Gooberpatrol_66 (QUIT: Ping timeout: 260 seconds) (~Gooberpat@user/gooberpatrol66) | 21:47 | |
| + xandyv (xandyv@user/xandyv) | 21:48 | |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-78-82-55-198.bbcust.telenor.se) | 22:15 | |
| - xandyv (QUIT: Ping timeout: 268 seconds) (xandyv@user/xandyv) | 22:54 | |
| + xandyv (xandyv@user/xandyv) | 22:55 | |
| - xandyv (QUIT: Ping timeout: 265 seconds) (xandyv@user/xandyv) | 23:09 | |
| - mtm (QUIT: Quit: Textual IRC Client: www.textualapp.com) (~textual@47-202-75-129.fdr01.sprg.fl.ip.frontiernet.net) | 23:42 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!