**** BEGIN LOGGING AT Sun May 13 03:00:07 2018 May 13 03:00:36 no, i don't think it's going to sleep, but i believe that this is a known issue -- one can for instance ssh into it after it is supposed to be sleeping. May 13 03:03:50 what shell are you using depscribe[m] ? May 13 03:16:36 sorry, went away for a moment May 13 03:16:45 I found it because I'm VNC'd onto a linux machine May 13 03:17:01 and whatever window I have there suddenly goes fullscreen whenever I shut the lid May 13 03:20:54 interesting May 13 03:21:13 I do wonder if it was intentional that closing the lid would generate a keypress May 13 03:21:31 i'm also noticing that sometimes a different application shows up once I open it back up May 13 03:22:10 i'm probably unique, but this is actually causing some usability issues May 13 03:22:30 notably, there are no defined f-keys on it, and sending an f11 gets really cumbersome May 13 03:22:52 (to restore the application back to normal size) May 13 03:37:13 Yeah, ESC [23~ May 13 03:37:38 I think it must be intended to hook to for sleeping May 13 03:39:05 Beiriannydd[w] (IRC): bash. May 13 03:40:15 depscribe[m]: does the environment variable TMOUT have a value? May 13 03:47:49 wait.. that's a key combo for f11? May 13 03:48:51 sheepbat: that's the ANSI sequence that is sent by closing the lid, so yes May 13 04:02:36 I guess you could generate another F11 by closing the lid again :) May 13 04:09:36 Beiriannydd[w] (IRC): will have to check tomorrow -- need to sleep noww. May 13 05:58:52 Beiriannydd[w], beware that my sources are for Linux, not Android May 13 05:59:11 it will break keyboard handling on Android, for example May 13 06:00:32 https://github.com/gemian/gemini-linux-kernel-3.18/commit/60bb1bd77ef6f9127dbfe11ac47c8967963d6b0f - this should probably answer night discussion about F11 on LED close May 13 06:01:03 (Chinese ODM has actually implemented LED close event as F11 key which is/should be trapped on Android) May 13 06:01:31 if you try to connect BT/USB keyboard on Android and press F11, it will lock the device May 13 06:12:49 TheKit: good job I'm trying it with Linux right now ;) May 13 06:15:09 TheKit: I opened a PR against your tree May 13 06:19:10 I commented on it, maybe you could check https://blogs.msdn.microsoft.com/commandline/2018/02/28/per-directory-case-sensitivity-and-wsl/? May 13 06:19:54 I'm running in docker though :D May 13 06:20:00 hmm May 13 06:24:14 I wonder if Docker for Windows is compatible May 13 06:24:20 one way to find out May 13 06:26:50 do you mount Windows filesystem inside or something? May 13 06:27:28 yeah, there's an open issue... that's sad May 13 06:28:56 I guess I could do the O= into the container itself May 13 06:35:22 I backed out the 2nd commit May 13 06:39:24 TheKit: I backed out the second commit May 13 06:43:27 Looks like that might have fixed the weird problem I was having with config_data.gz too May 13 06:45:49 I'll document that you can't do the build target on the windows mount May 13 08:24:55 depscribe[m]: you need to add xss-lock gemian-lock as your screensaver for sleep to work, also turn wifi/bt off if you want it to give any kind of length of stand by time (45-50% drop in 12h WiFi on, to 3-5% drop with WiFi off), also need to be logged in or screen saver will not be active, still some work todo for sleep on login screen May 13 08:26:45 sheepbat: if your getting F11/12 in linux then you need to update your kernel (see commit from TheKit) May 13 10:21:07 i'm getting it in android, but it does sound like that could be the source May 13 10:21:09 thanks May 13 10:21:29 well its an intrisic part of how android is built May 13 10:22:21 try pluggin in a usb/bt keyboard and press F11 then you can do nothing on the keyboard at all, eg the on(esc) button wont work until you press F12 May 13 14:42:04 hmmm.. nope May 13 14:42:49 just tried it, and my (other) android phone appears to be completely functional after pressing f11 May 13 14:43:00 i can try with the gemini though May 13 14:45:17 ..hunh. it actually does act that way with the gemini May 13 14:45:24 fair enough May 13 14:51:08 yea the linux kernel started life as the android kernel May 13 14:52:59 ..I think you have that backwards? May 13 14:53:50 well that too I guess May 13 14:54:36 I just mean that our branch which started with the F11/12 keys is what is used on the android side May 13 14:54:48 then there are the linux only changes May 13 14:55:12 of course the android kernel will be getting other changes too May 13 14:55:18 just we don't get visibility May 13 14:55:32 Hi all, how does the process of order locking and shipping work exactly - my order has been locked since 4 days, does that mean it has been shipped yet? May 13 14:55:56 locking is manual (means they are about to print your label) May 13 14:56:19 then shipping happens later, but there is no way of knowing how long between May 13 14:57:18 ok, thanks, I don't mind for a few more days, but since I didn't receive any tracking numbers I'd just like to make sure that's normal May 13 14:59:57 tracking info came from the delivery company, so as they are using new companies this time that might be different May 13 15:03:10 adam_b: turns out that even with wifi and bluetooth on, a full charge keeps the gadget running (with, admittedly, very little use, so in standby) a full 24 hours. what worries me, though, is that during the course of this it logs me out every few hours. i thought i'd set the screensaver per the wiki, but maybe not. i'll hook up a mouse and look; mouse needed for rmb. May 13 15:04:19 also, fully charges -- full target on lxqt panel -- in about 90 minutes. May 13 15:04:28 by log out you mean the lock screen? May 13 15:04:44 or back to the proper login screen with the dropdown for the keyboard May 13 15:05:56 if you've configured gemian-lock then you should always have to login again for any lenght of time of closed device (the screen shows a map of the world with yellow blobs for logerithemized backer numbers (a bit like the planet one) May 13 15:08:00 adam_b: negative. when i close the cover i'm logged in as [user] and a little while later, cover closed, i'm presented with the initial window manager login screen. May 13 15:08:26 thats new May 13 15:09:06 nor do i have any map of the world -- just plain tan background with little icons with user names and place for password. May 13 15:10:50 yea sounds like you've got something odd going on, if you do a ps x | grep lock do you see the xss-lock gemian-lock? May 13 15:14:42 pts/21 S+ 0:00 grep lock May 13 15:14:47 so no. May 13 15:15:08 then try the screensaver config again, and remove whatever you had tried before May 13 15:22:49 in lxqt session settings > autostart > global autostart xss-lock-gemian-lock is checked. and the command seems right, but i deleted it and tried again, so we'll see what happens. May 13 16:08:53 also, an unrelated observation: if you haven't disabled the blue LED, then to boot linux you do, as you've noted, get off the esc key as soon as it vibrates, but also get off the silver key as soon as the blue LED comes on. May 13 16:09:50 btw -- am i right in the belief that the planet partition tool requires a reflash and also makes it impossible to upgrade the kernel without reflashing? May 13 16:13:31 gemian lock process now shows up. May 13 16:14:02 depscribe[m]: with the linux tool you can individually flash partitions so long as you keep the same scatter file (so the sizes are all the same) May 13 16:14:30 its only the boot partition that won't flash with dd the linux_boot one is fine May 13 16:14:57 so its only if you swap to have linux as the default boot that you get problems with the new things May 13 16:17:20 I am <- this close -> to having a linux kernel that will run Docker. May 13 16:17:56 Beiriannydd[w]: I've been playing with lxc containers, mostly having problems with network connections May 13 16:18:21 does this include your custom scatter files? i wondered about the tool that lets one select the partition sizes (which i can't find at the moment) and the default boot. i didn't play with it but it implies that it's possible to devote the whole device to linux. but wouldn't that break some things in linux? doesn't our version require at least some android install? May 13 16:19:51 depscribe[m]: the partitionTool is on the planet site, also linked from the wiki in various places, it only goes down to 3G minimum for Android so no they don't yet support linux only May 13 16:21:33 so not much change from your 50-5 scatter file but for setting the default OS. does it keep us from updating the kernel by normal means (dist-upgrade)? May 13 16:22:27 depscribe[m]: no point in you reflashing May 13 16:22:49 believe me, i was not eager to! May 13 16:23:16 its mainly been released so that people who want android can still keep the eSIM function whilst having dual boot (the new downloads have x25/x27 seperate) May 13 16:23:20 are there lines beyond the debian ones i need to add to sources.list? May 13 16:23:26 no May 13 16:23:52 well you've currently got two lines one for debian and one for gemian (though it still has the old name gka) May 13 16:26:07 okay, yup, good. thanks. May 13 16:28:30 we briefly discussed last night the idea of there being something like "alt-tap" to bring a persistent (until some action is taken or another alt-tap cancels it) right mouse button action. this would pretty much eliminate the necessity of a mouse. is this something it would be difficult to map? May 13 16:30:04 should probably have some wishlist wiki page or raise issues in github or something to track requests and keep discussions in one place etc May 13 16:51:01 the OESF wiki, maybe? it doesn't look as if anything is happening there. May 13 16:51:37 reason i asked about sources.list is i was wondering as to kernel changes and whether they're sent along automagically. May 13 16:53:09 no kernel updates are manual for the moment, you remember the 'dd' command when you were setting it up May 13 17:10:31 are they put into a compiled kernel or do we need to compile it locally? May 13 17:18:32 and where would we find it? May 13 17:41:11 adam_b: same problem as you from docker - networking issues in the container May 13 17:43:01 depscribe[m]: the command is of the form of wget newkernel.img, dd newkernel.img to kernelpartition, I suspect you've already updated to the newest kernel May 13 17:43:30 but yes for the moment you have to monitor diffs on the DebianTP page to see what you have todo to keep uptodate May 13 17:44:30 the one i have is the march 26 one. May 13 17:44:32 Beiriannydd[w]: yea its all todo with adding the right groups for permissions etc May 13 17:44:40 yep nothing new since then May 13 17:44:52 oh of course! May 13 17:44:55 well experimental things todo with getting lxc/docker going May 13 17:45:04 but we've not released that yet as they are not working May 13 17:45:43 Beiriannydd[w]: thought I suspect there is more to it as that didn't help much May 13 17:53:17 I'll poke at it. May 13 17:59:16 have we a way to dump out of X entirely (as alt-ctrl-f1 in desktop linux)? it would be useful sometimes; often it's a choice of session at the WM's login screen, but not here and we don't have an f1, best i can tell. sometimes useful to, say, start up mc to find a file and make a quick edit. May 13 18:03:45 depscribe[m]: you've noticed how on a regular debian when it takes forever to reboot/shutdown you get a text only screen with a timeout countdown message saying some part of the system is refusing to close, but on the gemini we have no text only version so you just get the last x screen (or just the mouse pointer on a black screen) showing until it times out May 13 18:05:40 right. so there's no full-screen text-only console available other than by sshing in? May 13 18:05:42 oh and F1-10 are mapped to fn+shift+1-0 May 13 18:06:06 depscribe[m]: not unless someone makes the appropriate drivers May 13 18:08:59 adam_b: got it. i'm a big fan of mc and have it installed on the gemini, and lxqt was kind enough to give it a menu item, but it comes up unbelievably tiny and maximizing the window is no help. if there are settings in lxqt to make it a little friendlier, i haven't found them yet. May 13 18:09:52 if you run it in QTerminal you can change the font size May 13 18:10:11 Ctrl +/- May 13 18:16:20 OK I see what you mean, so if you set your default lxqt terminal to QTerminal that will make it more usable - From settings (fn-del) - Session Settings - Default Applications - Terminal Emulation - enter/select 'qterminal', logout and in again May 13 18:20:55 thought some people don't much like QTerminal - https://github.com/lxqt/lxqt/issues/433 May 13 18:21:59 the terminal you can actually see is always better than the terminal you can't! May 13 18:24:09 well QTerminal covers you for that May 13 18:39:21 adam_b (IRC): yes it is. May 13 18:39:22 here's a little observation that i make only because it might be useful in another context. i thought i would see if i could watch streamed video of the volcani in hawaii on gemini. in a browser player it was deadly bad, dropping all but one frame every four or five seconds. but when i made the player full screen, it played perfectly (or if it dropped any frames it wasn't noticeable). i was surprised. (though when May 13 18:39:23 the screen timed out and went black while the video was still going and i hit esc, it both brought back the screen and de-maximized the video.) not sure whether any of this is useful, but in case it is . . . May 13 18:42:02 depscribe: was that in firefox or chromium? I would have thought they were clever enough to fend of the screensaver, but I guess not May 13 19:03:58 firefox May 13 19:04:55 gets blanked in ios, too, but there's no esc to hit in ios. May 13 19:11:23 what surprised me wasn't so much the blanking, which seems like a fairly trivial adjustment, as the fact that the full-screen video was so good. it was pretty amazing. May 13 19:22:18 btw -- setting the default to qterminal makes all the difference. thanks again, adam. May 13 21:37:52 any advantages/disadvantages to using xfwm4 instead of kwin? it seems generally more lightweight; wondering what, if anything, i'd get by switching to it, and if i could dig back out fairly easily if i don't like it. **** ENDING LOGGING AT Mon May 14 03:00:01 2018