**** BEGIN LOGGING AT Fri May 25 03:00:04 2018 May 25 07:42:52 for the linux UI work we should clearly distinguish between touch-usable ui and mouse-usable desktop.. both have their uses and look quite different May 25 07:44:56 Oh look, a thing i have an opinion on! ;) May 25 07:45:33 did someone say opinion? May 25 07:45:34 Let me link you a video which shows why i think you are more correct than you might think i think you are ;) https://vimeo.com/112050358 May 25 07:48:12 Also, a slightly more detailed and wordy version of what that video shows in a more abstract fashion: https://dot.kde.org/2014/11/21/calligra-gemini-added-calligra-suite May 25 07:50:21 it's not even an opinion, it's a fact that the touch and mouse ui's really need to be different. you can of course use touch ui with mouse, but it wastes awful lot of screen space and can't use right click context menus. May 25 07:51:11 they really have to be completely different which is why all convergence attempts to far have been total failures May 25 07:51:23 you can't just make everything bigger May 25 07:51:28 That hasn't stopped people from pouring gigantic sums of money into it, though... witness Microsoft's attempt at doing convertible-friendly UX May 25 07:51:40 yes, or unity 8 May 25 07:51:47 So, as much as i certainly agree it's stupid, doesn't mean people don't still try and do it ;) May 25 07:52:52 the ubuntu touch convergence design was particularly terrible - the desktop version of their apps just had the mobile side tray thing opened all the time. it was basically "tablet view" that you get on android, with a mouse May 25 07:54:07 the window manager stuff wasn't too bad May 25 07:54:26 would be a poor choice for gemini tho, as it relies on edge swipes to edges you can't really reach due to the keyboard May 25 07:55:54 sailfish works quite nicely with swipes, although the bottom to top swipe is a bit difficult to do. but not a showstopper issue. May 25 07:59:37 There's a reason why apple has not put a touchscreen on their laptops. May 25 07:59:51 And why ipad does not support mouse. May 25 10:26:08 the git diff scan is about 32% done... May 25 10:26:20 it's going to take another couple of days i guess May 25 10:27:53 there are the files which don't match any git revision so far: http://paste.ubuntu.com/p/47qNNj5kwq/ May 25 15:24:50 Anyone else have issues booting to linux? I seem to keep getting it into TWRP or android. May 25 15:30:47 you are holding the power button too long May 25 15:31:50 hmm May 25 15:32:03 there's a function called openLedSpeed May 25 15:32:40 also - the voice assistant button action is hardcoded into the framework to execute a specific service in a specific package, so there is no easy way to override it May 25 15:34:27 Thank you. The instructions I was following, as I understood them said to hold both. May 25 15:34:54 you have to hold the power button to turn it on May 25 15:35:08 but then you have to let go of the power button and keep holding the voice button to boot linux May 25 15:35:23 they really could have made it simpler :/ May 25 15:39:50 yeah. got it working though. Now I'm seeing a DBUS error when trying to connect to wifi that I haven't seen before. May 25 15:46:30 net.connman.Error.NotSupported May 25 15:49:47 so afaict there is no way to override the voice button on the standard firmware May 25 15:50:42 you'd either have to patch the firmware, or make an app with id "com.pripla.gemini.voiceassistant" May 25 15:50:57 if you do the latter you wont be able to install it on stock firmware, because it is already installed May 25 17:08:20 Captain_Harlock: for booting to linux, the ques that work for me are to push the esc key; when the haptic feedback occurs, instantly push the silver button and let up on the esc key. when the blue led comes on, let up on the silver button. May 25 17:09:07 i've also encountered the dbus error -- but to apparently no effect, because it seems it connects to wifi anyway. May 25 17:26:14 I have received this message on a separate room on Matrix, called "System Alerts": May 25 17:26:22 "To continue to access the matrix.org homeserver you must agree to our new and much improved usage and privacy policies. We will shortly be blocking users who haven’t accepted them, so please click here to review and accept the terms as soon as possible!" May 25 17:27:02 it seems that those Matrix users who don't sign it will not be able to use Matrix at all... bummer May 25 17:33:05 if its decentralized how can they block? May 25 17:50:40 almost certainly GDPR related May 25 17:52:16 varti[m]: that should only apply to the matrix.org server rather than all Matrix servers May 25 18:03:08 nobody understands GDPR May 25 18:03:20 all the emails i get are wrong May 25 18:03:27 "click here to opt out" - wrong May 25 18:03:36 "accept or we delete your account" - wrong May 25 18:04:24 it's pretty amazing, isn't it. you'd think they just heard about the law this week. May 25 18:05:06 i'm not really surprised May 25 18:06:53 the register was having a good time with it today. May 25 18:16:06 i've set up a filter in gmail :) May 25 18:19:07 thatls like hiding your gunpowder in a burning building. May 25 18:21:18 gmail is the only mail service that can handle my mailbox May 25 18:23:03 after my isp signed up with the unholy alliamce of the corpses of yahoo and aol, bought by verizon, and announced they were going through my mail for anything they could use, i switched to protonmail. May 25 18:23:34 i enjoy targeted advertising May 25 18:24:10 it's creepy May 25 18:25:01 why? May 25 18:25:05 speaking of which, can we readily change the dns nameserver on gemini to, say, 1.1.1.1 instead of 8.8.8.8? May 25 18:25:17 on which OS? May 25 18:25:55 linux. i've heard it changes back, that it's hardcoded someplace, but i haven't really poked around in it yet. May 25 18:27:37 check /etc/resolv.conf May 25 18:33:44 slowly in the process of switching to protonmail as well May 25 18:33:50 resolv.conf is a link, so the change needs to be made upstream of there. May 25 18:34:10 what is it a link to? May 25 18:34:26 i'd run my own mail server again but the amount of work to be done to not get blocked as spam is prohibitive May 25 18:34:36 ../run/systemd/resolve/stub-resolv.conf? May 25 18:35:03 it's /run/connman/resolv.conf May 25 18:35:11 ah May 25 18:35:17 configure it in connman then :) May 25 18:35:42 or switch to systemd-networkd, or NetworkManager May 25 18:36:23 i prefer s-n for headless and n-m for desktops May 25 18:37:42 i just looked in connman and no joy. apparently this has been something of an issue in debian since at least jessie May 25 18:40:03 i'm tempted to just make my own /etc/resolv.conf and lock it, but i surely will have forgotten it when down the road it breaks something else. May 25 18:43:53 does anyone know how you actually get an esim to register? May 25 18:44:06 like how do you make it work if there's nothing to plug in? May 25 18:44:36 does it just exist to make it easier to lock phones to a network? May 25 18:44:44 does anyone actually have esim service? May 25 18:44:57 i dunno, that's what i'm asking i guess? May 25 18:45:46 i thought it was kind of a placeholder, like the proprietary expansion slots for devices that went out of business before anybody made anything for the expansion slot. May 25 18:45:59 placeholder? May 25 18:46:10 that esims are planned but not real yet. May 25 18:46:35 esim is like a virtual sim. there's no physical card, it's all done with crypto modules in the phone May 25 18:46:50 so somehow it has to register itself over the network May 25 18:47:16 right. but much was made of gemini's ability to use these things that don't exist yet except in a specification. May 25 18:48:04 "we can do this thing that doesn't exist." May 25 19:17:24 heh... is there a flashing tool that's not an opaque blob? May 25 19:17:33 because I got an interesting problem May 25 19:19:40 see, I have a nice laptop here. 64-bit Linux, as the docs want. So nice it doesn't even need qemu to execute Gemini binaries :) May 25 19:20:35 the source is on github May 25 19:20:48 https://github.com/dguidipc/SP-Flash-Tool-src May 25 19:22:09 ooh great, awesome. Won't be around a noisy box until much later today :p May 25 19:22:41 https://www.oesf.org/forum/index.php?s=&showtopic=34921&view=findpost&p=285989 May 25 19:23:26 btw, the correct command to reload udev rules is "sudo udevadm control --reload" May 25 19:23:56 you also need to blacklist the modem device if you dont want a brick May 25 19:24:22 https://github.com/ali1234/gemini-stuff May 25 19:27:07 pity it's qt4 only -- qt4 is on its way out and won't be in Buster, otherwise it'd be good to package this thing to people wouldn't have to repeat the work May 25 19:27:33 make a flatpak May 25 19:54:00 "you also need to blacklist the modem device if you dont want a brick" May 25 19:54:08 dafuq? O_o May 25 19:54:20 it is explained on my github May 25 19:56:22 ali1234: I think you may be misusing the word "brick" there May 25 19:56:45 maybe. who knows what can happen if you feed the download agent garbage? May 25 20:34:59 kilobyte, Pine64? May 25 20:36:32 SP FlashTool sources are pretty useless, unless someone tries to reverse the protocol, as they keept parts that does actual flashing as precompiled blobs May 25 20:36:39 *that do May 25 20:37:26 (they = MediaTek) May 25 20:41:34 Hello all May 25 20:41:50 Had my gemini for a week and just found time to flash linux May 25 20:42:15 2~Seeing the option to boot linux by default I of course went for that May 25 20:42:26 But now it looks as though I can't update the kernel easily that way May 25 20:42:54 What's the reason there? Is there a work-around? May 25 20:43:01 TheKit: very close. Pine64 is a dev board, Pinebook is a laptop. May 25 20:43:49 munchausen, so linux is in boot partition? May 25 20:45:02 TheKit I guess so yeah May 25 20:46:19 flash it to recovery partition, reboot into "recovery" mode (by holding Esc), flash boot partition... May 25 20:46:33 or flash TWRP as recovery and use it to flash boot.img May 25 20:47:14 Hmm seems like a pain. How come you can't just dd it? May 25 20:47:34 I don't know anything about the partition scheme (or bootloader) May 25 20:47:53 the source of issue is https://github.com/dguidipc/gemini-lk/blob/master/lk/platform/mt6797/write_protect.c#L137 May 25 20:48:29 Oh May 25 20:48:42 The boot partition is write protected? May 25 20:48:56 That's the bootloader source? May 25 20:49:56 yes, it's write protected in case of normal boot May 25 20:50:02 and not for recovery May 25 20:51:03 I see May 25 20:52:36 Thanks. So now to investigate building/updating the bootloader... May 25 20:58:59 building it out of Android tree is going to be interesting May 25 21:47:37 TheKit I guess the other option is sticking a tiny second stage bootloader in the boot partition that just jumps to the kernel in another partition May 25 23:56:40 curious: are things like phone and cellular data and accelerated video the kind of thing we would reasonably expect in the next technology preview? May 26 02:30:35 Hello from the Gemini via a freedompop sim. :) **** ENDING LOGGING AT Sat May 26 03:00:02 2018