| maxb | Hello, would anyone be able to help me figure out the right place to report this bug? Something appears to be wiping out the 'Synaptics Soft Button Areas' setting - despite the line "(**) Option "SoftButtonAreas" "50% 0 82% 0 0 0 0 0"" actually showing up in my Xorg.log | 16:34 |
|---|---|---|
| maxb | Any hints as to what might be responsible so I can target a report correctly would be appreciated. | 16:35 |
| jbicha | maxb: which version of Ubuntu are you using? | 16:37 |
| maxb | bionic | 16:41 |
| maxb | Sorry, should have said, this didn't happen on bionic around March, but does now | 16:41 |
| maxb | I'm not sure quite when it was introduced though | 16:42 |
| maxb | When I inspect it with xinput list-props, or synclient, all values are set to zero | 16:43 |
| jbicha | run ubuntu-bug mutter to report that issue please | 16:51 |
| jbicha | robert_ancell: good morning, I filed a few more gnome-initial-setup bugs today | 22:02 |
| jbicha | and GunnarHJ left a comment on bug 1761578 | 22:02 |
| ubot5` | bug 1761578 in gnome-initial-setup (Ubuntu) "Get more apps offers to install Firefox snap which is already installed" [Medium,Fix committed] https://launchpad.net/bugs/1761578 | 22:02 |
| maxb | jbicha: I think I found the underlying problem: it appears to be a somewhat dubious gsettings change. Setting org.gnome.desktop.peripherals.touchpad click-method to 'default' appears to rectify the problem | 22:20 |
| maxb | Confusingly, 'default' isn't the default, and the default appears to have been changed in a way to disable behaviour without it appearing to have a configuration UI | 22:20 |
| GunnarHj | jbicha: Do you know anything about that extra translation domain in g-c-c I stumbled upon? | 22:26 |
| maxb | :q | 22:29 |
| maxb | oops | 22:29 |
| tsimonq2 | :wq | 22:33 |
| maxb | Argh. Somehow in fixing that problem, some mystery software component has now started to trample on the Synaptics Click Action property to disable three-fingers to middle-button | 22:37 |
| jbicha | GunnarHj: it's strange https://paste.debian.net/1019344/ but there's no POTFILES.in so that sort of makes sense | 22:41 |
| jbicha | see the readme at https://gitlab.gnome.org/GNOME/gnome-control-center/tree/master/panels/datetime/po-timezones | 22:41 |
| jbicha | we apparently never bothered trying to translate them before in LP | 22:41 |
| jbicha | it's rare for a package to have more than one gettext domain so I'm more ok with handling that in the packages debian/rules | 22:43 |
| GunnarHj | jbicha: The timezone locations are in Swedish for me anyway, so I'm not sure there is even a need to translate it. Maybe it's simply obsolete code. | 22:44 |
| jbicha | I'm sure Debian GNOME will appreciate us not needing to have a dozen extra lines of boilerplate in all our main packages at least :) | 22:44 |
| GunnarHj | yeah.. | 22:45 |
| jbicha | GunnarHj: I assume it's in Swedish *because* it's translated upstream | 22:45 |
| GunnarHj | jbicha: But aren't the upstream translations stripped? | 22:46 |
| jbicha | maybe dh_translations doesn't know about that unusual usage | 22:46 |
| jbicha | you're welcome to make a patch to delete that sv.po and see what happens when you run it :) | 22:47 |
| GunnarHj | jbicha: Which sv.po? | 22:49 |
| GunnarHj | $ dpkg -L gnome-control-center-data | grep .mo$ | 22:49 |
| GunnarHj | $ | 22:49 |
| jbicha | the one in the directory I pointed to above | 22:50 |
| GunnarHj | Yeah, I realized that once I had asked. But there are no .mo files shipped with the packages AFAICT. | 22:50 |
| jbicha | 🤷 | 22:51 |
| GunnarHj | jbicha: And not in the langpacks. So it's a bit confusing. | 22:53 |
| jbicha | strange. Debian does have those files https://buildd.debian.org/status/fetch.php?pkg=gnome-control-center&arch=all&ver=1%3A3.28.0-1&stamp=1521423013&raw=0 | 22:54 |
| GunnarHj | jbicha: That's not so strange. We strip them when building, they don't. | 22:56 |
| GunnarHj | Can it be that kind of translations which are used to create translations in static files? | 22:56 |
| GunnarHj | jbicha: Let's ask someone tomorrow. There is possibly a simple explanation. | 23:00 |
| jbicha | GunnarHj: I checked on Debian and deleted that .mo. The timezones still appear translated | 23:01 |
| jbicha | so I'm going with the guess that those translations are somehow built into the binary | 23:01 |
| GunnarHj | jbicha: Me too. | 23:02 |
| jbicha | if that's true, there's still a bug in that distros don't need to install those .mo files | 23:02 |
| GunnarHj | Right. In that case, they shouldn't need to be built in the first place. | 23:03 |
| maxb | (filed bug 1762261 about one half of earlier reported touchpad issues) | 23:03 |
| ubot5` | bug 1762261 in mutter (Ubuntu) "Bug in Ubuntu patch for synaptics support prevents me configuring my touchpad same as previous Ubuntu releases" [Undecided,New] https://launchpad.net/bugs/1762261 | 23:03 |
| jbicha | maxb: oh, that sounds different than it sounded like you were talking about. Both the GNOME 3.28 and Ubuntu 18.04 release notes mention the change in touchpad right-click behavior | 23:39 |
| jbicha | and recommand that you use GNOME Tweaks if you want the old behavior | 23:39 |
| maxb | OK, so part of this is on me for not thoroughly reading the release notes, but even if you assume that everyone will do so, the situation remains that previously you *could* configure the gnome desktop to get out of the way and let you set custom stuff, and now you no longer can | 23:46 |
| maxb | It looks like the Tweaks UI doesn't permit setting click-method='default' | 23:51 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!