Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
Dantrell's GNOME Without Systemd Project (Part 2)
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3 ... 5, 6, 7 ... 14, 15, 16  Next  
This topic is locked: you cannot edit posts or make replies.    Gentoo Forums Forum Index Unsupported Software
View previous topic :: View next topic  
Author Message
Shibotto
Apprentice
Apprentice


Joined: 19 Jun 2015
Posts: 156
Location: CET/CEST

PostPosted: Tue Oct 02, 2018 11:36 am    Post subject: Reply with quote

The thumbnailers in /usr/share/thumbnailers/ seem to work fine when manually invoked and both eog and totem are able to generate thumbnails, so the no-sandbox patch still works. It seems only Nautilus is failing (as it generates a file in ~/.cache/thumbnails/fail/), vanilla or not. I have no idea how to get a decent output on stdout or some log file for Nautilus, so I'm unable to track the problem as of now.
Back to top
View user's profile Send private message
saboya
Guru
Guru


Joined: 28 Nov 2006
Posts: 552
Location: Brazil

PostPosted: Tue Oct 02, 2018 2:22 pm    Post subject: Reply with quote

gnome-settings-daemon pulls wacom stuff, which can also be optional:

https://gist.github.com/saboya/d3aa42205952440b87b00242ca6670dd

This compiles fine, but still needs some runtime testing.
Back to top
View user's profile Send private message
Sp1der
n00b
n00b


Joined: 02 Oct 2018
Posts: 1
Location: Ulm, Germany

PostPosted: Tue Oct 02, 2018 5:58 pm    Post subject: librsvg merge fails Reply with quote

Hello,

I'm on the 3.28 extended profile and I can't successfully merge librsvg. This is my build.log. The compiling seems to go fine but ld throws errors. I tried emerging the rust compiler with abi_x86_32 but that didn't help. I don't know if this is a bug or an error on my side, any ideas to fix it?
Back to top
View user's profile Send private message
Elleni
Veteran
Veteran


Joined: 23 May 2006
Posts: 1270

PostPosted: Tue Oct 02, 2018 6:29 pm    Post subject: Reply with quote

Is that normal, that in gnome terminal there is no possibility anymore to open new tab, just new terminal?
Back to top
View user's profile Send private message
runningnak3d
n00b
n00b


Joined: 05 Sep 2018
Posts: 62

PostPosted: Tue Oct 02, 2018 7:06 pm    Post subject: Reply with quote

Elleni wrote:
Is that normal, that in gnome terminal there is no possibility anymore to open new tab, just new terminal?


I am not sure what the behavior was in 3.28 and below (I use Terminator), but in 3.30 if you go to: Edit -> Preferences -> General
you can choose to open new terminals in a tab or window -- it defaults to window. Of course shift+ctrl+t works as well.

-- Brian
Back to top
View user's profile Send private message
dantrell
l33t
l33t


Joined: 01 Jun 2007
Posts: 915
Location: Earth

PostPosted: Tue Oct 02, 2018 7:48 pm    Post subject: Reply with quote

saboya wrote:
I had a couple warnings on gnome-control-center [...]

Quite a few packages (dozens and dozens) have -Wimplicit-function-declaration warnings.

They are harmless and I can't be bothered. :roll:

dantrell wrote:
BuBuaBu wrote:
I guess the networkmanager use flag should not exist.

runningnak3d wrote:
* Add networkmanager use flag for gnome-settings-daemon in package.use

I missed the "(not optional on Linux platforms)" portion when I was checking for Meson configure options.

Edit: I have corrected this for now but I'll be looking into making it optional later.

saboya wrote:
gnome-settings-daemon pulls wacom stuff, which can also be optional [...]

The following packages are now optional dependencies of GNOME Settings Daemon:
  • GUdev
  • NetworkManager
  • Wacom
  • colord
Although, GNOME Settings Daemon will probably halt GNOME's startup without colord but hey, don't let that stop you. :twisted:

Elleni wrote:
It does 8O

I still can't reproduce this issue though. :(

dantrell wrote:
runningnak3d wrote:
Well, there is still the TZ issue, but there is a workaround for that.

This is next on the list.

I also reviewed a few packages and considerable code changes but got nowhere with this. It seems that it isn't due to GNOME Control Center or GNOME Shell but that doesn't make much sense.

It's definitely related to a package in GNOME 3.30 though. I would apply the TZ workaround globally but there isn't a way to make deployment easy and it won't automatically update when the timezone is manually changed so finding an actual fix it my priority.

Shibotto wrote:
The thumbnailers in /usr/share/thumbnailers/ seem to work fine when manually invoked and both eog and totem are able to generate thumbnails, so the no-sandbox patch still works. It seems only Nautilus is failing (as it generates a file in ~/.cache/thumbnails/fail/), vanilla or not. I have no idea how to get a decent output on stdout or some log file for Nautilus, so I'm unable to track the problem as of now.

Update @world with USE="-vanilla-thumbnailer" (~3 packages should be affected).

Delete or move ~/.cache/thumbnails (it might be enough to just delete or move ~/.cache/thumbnails/fail but I can't recall).

Kill any thumbnail processes and re-login.

Sp1der wrote:
I'm on the 3.28 extended profile and I can't successfully merge librsvg. This is my build.log. The compiling seems to go fine but ld throws errors. I tried emerging the rust compiler with abi_x86_32 but that didn't help. I don't know if this is a bug or an error on my side, any ideas to fix it?

I wasn't able to reproduce this issue and if it's multilib related I wouldn't be able to test it.

I'll investigate and see what I find but I don't think anything depends on that particular version so you should be able to downgrade to the 2.42 or 2.40 series for the time being.
_________________
Dantrell B.
Back to top
View user's profile Send private message
saboya
Guru
Guru


Joined: 28 Nov 2006
Posts: 552
Location: Brazil

PostPosted: Tue Oct 02, 2018 11:05 pm    Post subject: Re: librsvg merge fails Reply with quote

Sp1der wrote:
Hello,

I'm on the 3.28 extended profile and I can't successfully merge librsvg. This is my build.log. The compiling seems to go fine but ld throws errors. I tried emerging the rust compiler with abi_x86_32 but that didn't help. I don't know if this is a bug or an error on my side, any ideas to fix it?

If you're on a multilib profile you should set ABI_X86="64 32" on /etc/portage/make.conf instead of manually setting abi_x86_32 on packages.

If you're not on a multilib profile, you shouldn't be messing with that at all.
Back to top
View user's profile Send private message
Shibotto
Apprentice
Apprentice


Joined: 19 Jun 2015
Posts: 156
Location: CET/CEST

PostPosted: Wed Oct 03, 2018 10:47 am    Post subject: Reply with quote

dantrell wrote:

Update @world with USE="-vanilla-thumbnailer" (~3 packages should be affected).

Delete or move ~/.cache/thumbnails (it might be enough to just delete or move ~/.cache/thumbnails/fail but I can't recall).

Kill any thumbnail processes and re-login.


I told you man, I tried this already to no avail :D

Since compiling Nautilus is something even my outdated hardware can do in about a minute, I bisected 3.28 - 3.30 and found the problem. gnome-desktop code for generating thumbnails was included in Nautilus before 3.30, so we need the no-sandbox patch in Nautilus as well. Fortunately the code is still the same, so changing the path of the source file is the only modification needed (src/gnome-desktop/gnome-desktop-thumbnail-script.c).

I've been testing this and it works fine :D
Back to top
View user's profile Send private message
dantrell
l33t
l33t


Joined: 01 Jun 2007
Posts: 915
Location: Earth

PostPosted: Wed Oct 03, 2018 3:56 pm    Post subject: Reply with quote

Shibotto wrote:
I told you man, I tried this already to no avail :D

I was just making sure. :P

Shibotto wrote:
Since compiling Nautilus is something even my outdated hardware can do in about a minute, I bisected 3.28 - 3.30 and found the problem. gnome-desktop code for generating thumbnails was included in Nautilus before 3.30, so we need the no-sandbox patch in Nautilus as well. Fortunately the code is still the same, so changing the path of the source file is the only modification needed (src/gnome-desktop/gnome-desktop-thumbnail-script.c).

I could swear up and down that I tested this and it was fine. I would even swear left and right that I tested it twice. But somehow, after I tested it a third time I was able to reproduce your issue. 8O

Completely baffled, I reached and confirmed the same resolution as you. Nice work!

P.S. For the record. :twisted:
_________________
Dantrell B.
Back to top
View user's profile Send private message
saboya
Guru
Guru


Joined: 28 Nov 2006
Posts: 552
Location: Brazil

PostPosted: Wed Oct 03, 2018 4:04 pm    Post subject: Reply with quote

Hey, dantrell

Any chance this could be supported somehow?

https://github.com/bus1/dbus-broker

The main difficulty seems to be the launcher part. Would have to be OpenRC-specific I guess.
Back to top
View user's profile Send private message
Elleni
Veteran
Veteran


Joined: 23 May 2006
Posts: 1270

PostPosted: Wed Oct 03, 2018 4:46 pm    Post subject: Reply with quote

dantrell wrote:

I still can't reproduce this issue though. :(


Starting as root threw some more messages, maybe this helps. I will try with unstable libva-2.2.0 and see if this helps
Code:
gnome-boxes               

(gnome-boxes:11918): Gtk-WARNING **: 18:41:09.288: GtkFlowBox with a model will ignore sort and filter functions

(gnome-boxes:11918): Gtk-WARNING **: 18:41:09.289: GtkListBox with a model will ignore sort and filter functions

(gnome-boxes:11918): Gtk-WARNING **: 18:41:09.311: Theme parsing error: gtk.css:68:35: The style property GtkButton:child-displacement-x is deprecated and shouldn't be used anymore. It will be removed in a future version

(gnome-boxes:11918): Gtk-WARNING **: 18:41:09.311: Theme parsing error: gtk.css:69:35: The style property GtkButton:child-displacement-y is deprecated and shouldn't be used anymore. It will be removed in a future version

(gnome-boxes:11918): Gtk-WARNING **: 18:41:09.311: Theme parsing error: gtk.css:73:46: The style property GtkScrolledWindow:scrollbars-within-bevel is deprecated and shouldn't be used anymore. It will be removed in a future version
Error setting up default broker: Unable to open qemu+unix:///session: Socket-Erstellung zu '/var/run/libvirt/libvirt-sock' fehlgeschlagen: Datei oder Verzeichnis nicht gefunden

(gnome-boxes:11918): Boxes-WARNING **: 18:41:10.610: (src/src@@gnome-boxes@exe/app.c:3352):boxes_app_setup_default_source_co: runtime check failed: (default_connection != null)
libva info: VA-API version 0.39.4
libva info: va_getDriverName() returns -1
libva error: va_getDriverName() failed with unknown libva error,driver_name=(null)
libva info: VA-API version 0.39.4
libva info: va_getDriverName() returns -1
libva error: va_getDriverName() failed with unknown libva error,driver_name=(null)
error: XDG_RUNTIME_DIR not set in the environment.
libva info: VA-API version 0.39.4
libva info: va_getDriverName() returns 0
libva info: Trying to open /usr/lib64/va/drivers/radeonsi_drv_video.so
libva info: Found init function __vaDriverInit_0_39
libva info: va_openDriver() returns 0
Speicherzugriffsfehler


Edit: Did not solve the issue, so I am reverting back to libva-1.7.3

Code:
gnome-boxes[5657]: segfault at 0 ip 00007feea298748f sp 00007ffd4b0dbbe0 error 4 in libosinfo-1.0.so.0.1002.0[7feea2976000+39000]
[   97.761685] Code: 63 f6 64 48 8b 04 25 28 00 00 00 48 89 44 24 08 31 c0 48 8b 47 18 48 8b 00 48 8b 54 24 08 64 48 33 14 25 28 00 00 00 48 8b 00 <48> 8b 04 f0 75 05 48 83 c4 18 c3 e8 21 cc ff ff 90 48 83 ec 18 64
is what I was getting wieth libva-2.2.0.

Another thing, I realize is, that from time to time, screens get black for 1 or 2 seconds then they show up again. Is this a known issue?

runningnak3d wrote:

I am not sure what the behavior was in 3.28 and below (I use Terminator), but in 3.30 if you go to: Edit -> Preferences -> General
you can choose to open new terminals in a tab or window -- it defaults to window. Of course shift+ctrl+t works as well.

-- Brian


Thank you :oops:


Last edited by Elleni on Wed Oct 03, 2018 4:57 pm; edited 2 times in total
Back to top
View user's profile Send private message
newcomer
n00b
n00b


Joined: 19 Nov 2017
Posts: 11

PostPosted: Thu Oct 04, 2018 2:10 am    Post subject: Reply with quote

Thank you for dantrell's great work, I let Gentoo with Gnome without systemd run smoothly on the USB drive :-)

https://github.com/prefercloud/Gentoo-on-USB
Back to top
View user's profile Send private message
divoxx
n00b
n00b


Joined: 23 Feb 2005
Posts: 23
Location: San Francisco, CA

PostPosted: Thu Oct 04, 2018 2:45 am    Post subject: Reply with quote

After updating portage today there is a gom update which is failing to build.

Quote:
$ emerge -pqv '=dev-libs/gom-0.3.3-r2::dantrell-gnome'
[ebuild U ] dev-libs/gom-0.3.3-r2 [0.3.3-r1] USE="introspection -doc%" PYTHON_TARGETS="python3_6 -python3_4 -python3_5 (-python3_7)"


emerge info, build log and meson log are all available here: https://gist.github.com/divoxx/147bb1fc91f0fe3a5911f86c492132ee

The general error is:

Quote:
bindings/python/meson.build:28:2: ERROR: Problem encountered: Failed to determine pygobject overridedir
Back to top
View user's profile Send private message
dantrell
l33t
l33t


Joined: 01 Jun 2007
Posts: 915
Location: Earth

PostPosted: Thu Oct 04, 2018 3:51 am    Post subject: Reply with quote

saboya wrote:
Any chance this could be supported somehow?

https://github.com/bus1/dbus-broker

The main difficulty seems to be the launcher part. Would have to be OpenRC-specific I guess.

I duct taped something together (although the elogind patch can be simplified because I initially wrote it against an older release).

It requires elogind version 230 (or later) which is keyword masked so you should account for that when testing:

Code:
ACCEPT_KEYWORDS="**" emerge -va1 --nodeps =elogind-234.4-r1 dbus-broker

That said, it builds but I haven't checked to see if it actually works. I also didn't provide an OpenRC init script.

Have fun. :twisted:

Elleni wrote:
Starting as root threw some more messages, maybe this helps.

It lead me here which didn't help much I'll tell you that. :lol:

Elleni wrote:
Another thing, I realize is, that from time to time, screens get black for 1 or 2 seconds then they show up again. Is this a known issue?

This... sounds familiar. 8O

A trip down memory lane is leading me towards a non-GNOME related graphics card or docking/external/multi-monitor setup issue? You know what, I think I distinctly recall a multi-monitor issue.

divoxx wrote:
After updating portage today there is a gom update which is failing to build.

I was able to reproduce and correct this issue.

Sync up and you should be good to go.
_________________
Dantrell B.
Back to top
View user's profile Send private message
Shibotto
Apprentice
Apprentice


Joined: 19 Jun 2015
Posts: 156
Location: CET/CEST

PostPosted: Thu Oct 04, 2018 8:38 am    Post subject: Reply with quote

dantrell wrote:
I was just making sure. :P

I'm glad you did, because you made me check again before posting. Quite often I'm sure I did things, then it turns out I forgot to issue the final command :lol: :lol:
Sorry I sounded rude, I'm still very upset with GNOME developers about the smoothness thing, but in the future I'll try to convey this feeling directly to them instead :evil:

---

@Elleni I forgot I had GNOME Boxes installed, so I checked only now and I have the same segfault on libosinfo as you. Using a more recent sys-apps/osinfo-db solved the problem for me.
Back to top
View user's profile Send private message
dantrell
l33t
l33t


Joined: 01 Jun 2007
Posts: 915
Location: Earth

PostPosted: Thu Oct 04, 2018 3:34 pm    Post subject: Reply with quote

Shibotto wrote:
I'm glad you did, because you made me check again before posting. Quite often I'm sure I did things, then it turns out I forgot to issue the final command :lol: :lol:
Sorry I sounded rude, I'm still very upset with GNOME developers about the smoothness thing, but in the future I'll try to convey this feeling directly to them instead :evil:

No problem but you might have to line up in the queues called NOTABUG and WONTFIX. :twisted:

dantrell wrote:
Elleni wrote:
One issue, I see on my box is that boxes is not starting and dmesg shows [...]

I can't reproduce this. Could it because it needs libosinfo-1 as a run-time dependency?

Edit: Actually it looks like that is already allowed through for stable.

Shibotto wrote:
@Elleni I forgot I had GNOME Boxes installed, so I checked only now and I have the same segfault on libosinfo as you. Using a more recent sys-apps/osinfo-db solved the problem for me.

I thought it was this but I forgot that libosinfo was actually 3 packages.

Thanks to your hint I was able to reproduce and correct this issue.

@Elleni, sync up and you should be good to go.

dantrell wrote:
runningnak3d wrote:
Well, there is still the TZ issue, but there is a workaround for that.

This is next on the list.

Now, if a fix for the TZ issue would just fall from the sky. :lol:
_________________
Dantrell B.
Back to top
View user's profile Send private message
runningnak3d
n00b
n00b


Joined: 05 Sep 2018
Posts: 62

PostPosted: Thu Oct 04, 2018 6:45 pm    Post subject: Reply with quote

dantrell wrote:

Now, if a fix for the TZ issue would just fall from the sky. :lol:


I was going to make a joke that maybe Gnome devs made systemd a runtime requirement for determining the TZ.

A quick Google search returned this: https://serverfault.com/questions/666979/why-do-i-need-to-use-systemd-to-set-my-timezone-in-rhel7

What kind of crack is Poettering smoking? And then you have his followers that are just sucking the systemD. :evil:


I am still digging, but I don't think the problem is with Gnome, I think it is with glib.

-- Brian
Back to top
View user's profile Send private message
Elleni
Veteran
Veteran


Joined: 23 May 2006
Posts: 1270

PostPosted: Thu Oct 04, 2018 7:24 pm    Post subject: Reply with quote

dantrell wrote:

A trip down memory lane is leading me towards a non-GNOME related graphics card or docking/external/multi-monitor setup issue? You know what, I think I distinctly recall a multi-monitor issue.

You are right, but I was pleased to see, that finally with this gnome release my monitors are correctly displayed in the corresponging session of gnome settings, and I can configure them, although, I am missing a feature - There is no option to clone/mirror one monitor to another. May this have something to do, that there are more than two monitors attached? I have 3 monitors and my tv attached.

dantrell wrote:

Sync up and you should be good to go.


I will do so, but atm I decided to remove wayland and enlightenment, as those were two things, I installed out of couriosity not too long ago, and I was wondering if some of the related useflags could have lead to instability and segfaults, so I thought, its maybe worth reducing complexity. I will report back, if segfault on boxes is gone as soon as my box is up to date.

Speaking of segfaults, following another one, I came from realzing
Code:
Chrome_~dThread[17075]: segfault at 0 ip 00007f332a34fe73 sp 00007f3327b7eb00 error 6 in libxul.so[7f3329e13000+5958000]


shibotto wrote:
@Elleni I forgot I had GNOME Boxes installed, so I checked only now and I have the same segfault on libosinfo as you. Using a more recent sys-apps/osinfo-db solved the problem for me.


Boxes is starting successfully now

Compile of mutter-3.30.0 failed after I removed wayland useflag:
http://bpaste.net/show/05c60fd6c768

And gnome-settings-daemon:
http://bpaste.net/show/68e424c6c089

Thanks once more, dantrell and shibotto
Back to top
View user's profile Send private message
Shibotto
Apprentice
Apprentice


Joined: 19 Jun 2015
Posts: 156
Location: CET/CEST

PostPosted: Fri Oct 05, 2018 5:00 pm    Post subject: Reply with quote

runningnak3d wrote:
I am still digging, but I don't think the problem is with Gnome, I think it is with glib.

I'm really interested in knowing how you figured that out 8O
Thank you very much.

Code:
8945227743a26a4fe6966baaf082dd6516e8a03c is the first bad commit
commit 8945227743a26a4fe6966baaf082dd6516e8a03c
Author: Philip Withnall <withnall@endlessm.com>
Date:   Wed Apr 11 15:59:29 2018 +0100

    gtimezone: Add g_time_zone_get_identifier() accessor

    This is a non-trivial accessor which gets the identifier string used to
    create the GTimeZone — unless the string passed to g_time_zone_new() was
    invalid, in which case the identifier will be `UTC`.

    Implementing this required reworking how timezone information was loaded
    so that the tz->name is always set at the same time as tz->t_info, so
    they are in sync. Previously, the tz->name was unconditionally set to
    whatever was passed to g_time_zone_new(), and then not updated if the
    tz->t_info was eventually set to the default UTC information.

    This includes tests for the new g_time_zone_get_identifier() API, and
    for the g_date_time_get_timezone() API added in the previous commit.

    Signed-off-by: Philip Withnall <withnall@endlessm.com>

    https://bugzilla.gnome.org/show_bug.cgi?id=795165

:040000 040000 820d11ac6a87b0195bd4d956c92c68b1998724ea 2bc546373929d1731c85c65b7ab057fcabde73ec M  docs
:040000 040000 c0293f2fe55ecb8be62a52d424b6fe05d747e704 7e07b77617399478cd81d1e946652c475e9065a9 M  glib

Have fun guys. Right know I don't even have time to look at what was actually commited, but if no one gets to it in a couple of hours, I probably will.

EDIT: Quick Github link
Back to top
View user's profile Send private message
runningnak3d
n00b
n00b


Joined: 05 Sep 2018
Posts: 62

PostPosted: Fri Oct 05, 2018 5:47 pm    Post subject: Reply with quote

Shibotto wrote:
runningnak3d wrote:
I am still digging, but I don't think the problem is with Gnome, I think it is with glib.

I'm really interested in knowing how you figured that out 8O
Thank you very much.

Code:
8945227743a26a4fe6966baaf082dd6516e8a03c is the first bad commit
commit 8945227743a26a4fe6966baaf082dd6516e8a03c
Author: Philip Withnall <withnall@endlessm.com>
Date:   Wed Apr 11 15:59:29 2018 +0100

    gtimezone: Add g_time_zone_get_identifier() accessor

    This is a non-trivial accessor which gets the identifier string used to
    create the GTimeZone — unless the string passed to g_time_zone_new() was
    invalid, in which case the identifier will be `UTC`.

    Implementing this required reworking how timezone information was loaded
    so that the tz->name is always set at the same time as tz->t_info, so
    they are in sync. Previously, the tz->name was unconditionally set to
    whatever was passed to g_time_zone_new(), and then not updated if the
    tz->t_info was eventually set to the default UTC information.

    This includes tests for the new g_time_zone_get_identifier() API, and
    for the g_date_time_get_timezone() API added in the previous commit.

    Signed-off-by: Philip Withnall <withnall@endlessm.com>

    https://bugzilla.gnome.org/show_bug.cgi?id=795165

:040000 040000 820d11ac6a87b0195bd4d956c92c68b1998724ea 2bc546373929d1731c85c65b7ab057fcabde73ec M  docs
:040000 040000 c0293f2fe55ecb8be62a52d424b6fe05d747e704 7e07b77617399478cd81d1e946652c475e9065a9 M  glib

Have fun guys. Right know I don't even have time to look at what was actually commited, but if no one gets to it in a couple of hours, I probably will.

EDIT: Quick Github link


Thank YOU very much for continuing to dig. I determined that it was a problem with glib by looking at (crap -- I forget the file in Gnome source, and I am not on my build PC -- will edit this later). Anyway, when I saw that Gnome used glib for TZ, I compiled up a little program to display the timezone -- and it was UTC, so that took Gnome out of the picture.

Unfortunately I got busy and didn't get to dig into WHY. So again, thank you very much for that info. I think maybe we have enough info to make a proper fix now. I hate having hacks ;)

-- Brian
Back to top
View user's profile Send private message
dantrell
l33t
l33t


Joined: 01 Jun 2007
Posts: 915
Location: Earth

PostPosted: Fri Oct 05, 2018 7:03 pm    Post subject: Reply with quote

Elleni wrote:
Compile of mutter-3.30.0 failed after I removed wayland useflag:
http://bpaste.net/show/05c60fd6c768

And gnome-settings-daemon:
http://bpaste.net/show/68e424c6c089

Stop breaking things. :twisted:

Shibotto wrote:
runningnak3d wrote:
I am still digging, but I don't think the problem is with Gnome, I think it is with glib.

I'm really interested in knowing how you figured that out 8O

I ruled out GNOME Control Center and GNOME Shell (and a few other packages) but even I didn't think to look at GLib.

So last night, after I got that hint, I pulled up the code for GLib and sure enough there it was.

Shibotto wrote:
runningnak3d wrote:
Have fun guys. Right know I don't even have time to look at what was actually commited, but if no one gets to it in a couple of hours, I probably will.

EDIT: Quick Github link


Thank YOU very much for continuing to dig. I determined that it was a problem with glib by looking at (crap -- I forget the file in Gnome source, and I am not on my build PC -- will edit this later). Anyway, when I saw that Gnome used glib for TZ, I compiled up a little program to display the timezone -- and it was UTC, so that took Gnome out of the picture.

Unfortunately I got busy and didn't get to dig into WHY. So again, thank you very much for that info.

I direct your attention to GNOME bug #795165.

Ting-Wei Lan mentioned that GNOME Shell defaulted to UTC on FreeBSD and explained why it happened:

Ting-Wei Lanl wrote:
It seems that that the patch expects /etc/localtime to be a symlink, but the assumption is seldom true on FreeBSD. FreeBSD installer copies the timezone file from /usr/share/zoneinfo to /etc/localtime when setting system timezone, so /etc/localtime is a regular file.

Color me surprised when the first response from a GNOME developer was:

Philip Withnall wrote:
That’s a pain. How are you supposed to find the current timezone on FreeBSD then? Why don’t they symlink?

How I imagined FreeBSD (and everyone else with a different timezone location).

But to be fair, he did add a fallback for FreeBSD and DragonflyBSD.

runningnak3d wrote:
I think maybe we have enough info to make a proper fix now. I hate having hacks ;)

That, we do.

P.S. I think I can promote GNOME 3.30 from Paper™ to Aluminum now. :roll:
_________________
Dantrell B.
Back to top
View user's profile Send private message
Elleni
Veteran
Veteran


Joined: 23 May 2006
Posts: 1270

PostPosted: Sat Oct 06, 2018 12:43 pm    Post subject: Reply with quote

dantrell wrote:

Stop breaking things. :twisted:

ok, ok, I readded wayland useflags and recompiled. :oops:

I will try to get rid of wayland, when someone mentiones why mutter does not build without wayland support.
Back to top
View user's profile Send private message
dantrell
l33t
l33t


Joined: 01 Jun 2007
Posts: 915
Location: Earth

PostPosted: Sat Oct 06, 2018 4:56 pm    Post subject: Reply with quote

Elleni wrote:
Is that normal, that in gnome terminal there is no possibility anymore to open new tab, just new terminal?

runningnak3d wrote:
I am not sure what the behavior was in 3.28 and below (I use Terminator), but in 3.30 if you go to: Edit -> Preferences -> General
you can choose to open new terminals in a tab or window -- it defaults to window. Of course shift+ctrl+t works as well.

Having to use that particular setting didn't sound right to me so I spun up a few test installs and GNOME 3.14 showed:

Code:
Open Terminal

GNOME 3.26:

Code:
Open Terminal
Open Tab

GNOME 3.28:

Code:
New Window (Re-positioned in the Context Menu)
New Tab (Re-positioned in the Context Menu)

So naturally I found out what changed with GNOME Terminal 3.30 and reverted back to the GNOME Terminal 3.28 default.

As per usual, you can manage this through the vanilla* USE flags.

And when things cool down, I'll probably be doing some context menu re-ordering and backporting the ability to open terminals in a new tab but that's a trivial issue.

Elleni wrote:
dantrell wrote:

Stop breaking things. :twisted:

ok, ok, I readded wayland useflags and recompiled. :oops:

I will try to get rid of wayland, when someone mentiones why mutter does not build without wayland support.

I was only teasing. :lol:

I wasn't immediately able to reproduce these build failures but I'll be digging deeper.
_________________
Dantrell B.
Back to top
View user's profile Send private message
Elleni
Veteran
Veteran


Joined: 23 May 2006
Posts: 1270

PostPosted: Sat Oct 06, 2018 5:08 pm    Post subject: Reply with quote

I know, thanks for looking into it :wink:
Back to top
View user's profile Send private message
99ninetynine
n00b
n00b


Joined: 28 Sep 2018
Posts: 1

PostPosted: Sat Oct 06, 2018 6:31 pm    Post subject: Reply with quote

Hi Dantrell and thank you for your work.

Is it possible to add iio-sensor-proxy without systemd?
I have x220 tablet and have no idea how to enable autosreen rotation.

Thanks.
Back to top
View user's profile Send private message
Display posts from previous:   
This topic is locked: you cannot edit posts or make replies.    Gentoo Forums Forum Index Unsupported Software All times are GMT
Goto page Previous  1, 2, 3 ... 5, 6, 7 ... 14, 15, 16  Next
Page 6 of 16

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum