Article 5DM7Q multiple problems with slackware-current (Jan2021) and xfce-4.16

multiple problems with slackware-current (Jan2021) and xfce-4.16

by
Rod3775
from LinuxQuestions.org on (#5DM7Q)
Since late Dec, Slackware-current has used xfce-4.16. This has introduced a number of problems. xfce4-panel-4.16 changed its API and dropped support for GTK-2 plugins. I run it on both x86_64 machines and an old i586 tower. The x86_64 version and Patrick-supported plugins work as distributed in binary. Several of the i586 pieces do not. Systray and Launcher, which are "core" parts of Xfce-panel will not run.

However, if I download distributed SOURCE from a Slackware mirror, and compile it on my i586 machine with Patrick's latest gcc suite, it works fine. Even though Launcher is part of xfce4-panel, the distributed binary thinks it has GTK2 parts. Ldd shows no GTK2 libraries, and Thunar can run it just fine.

sensors-plugin, netload-plugin, and mailwatch-plugin, which are on SlackBuilds, required downloading the latest source from Xfce.org, building a source package for each of them and creating a Slackware package from that. This not a -current problem, but will persist for a while because the SlackBuilds maintainers opted to only support 14.2 until 15.0 comes out as a stable "release". This isn't a bug, its a feature.

On my laptop, I run nm-applet and make Wifi connections through NetworkManager, which displays in the (now working) Systray. However,
NetworkManager uses repeated calls to glibc to check a compile-time(!?!) setting, and generates almost 15,000 lines of log entries PER DAY in .xsession-errors, even though the program runs just fine. Here's a patch for that:
-- /tmp/NetworkManager-1.28.0/libnm-core/nm-setting-connection.c2020-12-06 05:47:55.000000000 -0800
+++ /usr/local/slackbuilds/NetworkManager-1.28.0/libnm-core/nm-setting-connection.c2021-01-31 06:54:13.482428263 -0800
@@ -19,6 +19,15 @@
#include "nm-setting-vlan.h"
#include "systemd/nm-sd-utils-shared.h"

+/*
+ Routines in this file are called frequently (10x/minute) and flood the .xsession-error
+ log with "critical" messages. The code works fine anyway.
+*/
+#undef g_return_if_fail
+#undef g_return_val_if_fail
+#define g_return_if_fail(a) if(!(a))return
+#define g_return_val_if_fail(a,b) if(!(a))return b
+
/**
* SECTION:nm-setting-connection
* @short_description: Describes general connection properties

As a final issue, the -current version of the nouveau package on my i586 tower interacts with xfwm4 to generate a rapid flood of syslog messages from the 5.10.(10|11) kernel that fills the disk and crashes the system (although graphics seem to work fine until the disk fills!). I blacklisted nouveau on that machine. It doesn't happen on my x86_64 tower, and my x86_64 laptop doesn't have NVidia hardware, so no issue there.latest?d=yIl2AUoC8zA latest?i=atl__roiWf0:tb2MRqCvFkI:F7zBnMy latest?i=atl__roiWf0:tb2MRqCvFkI:V_sGLiP latest?d=qj6IDK7rITs latest?i=atl__roiWf0:tb2MRqCvFkI:gIN9vFwatl__roiWf0
External Content
Source RSS or Atom Feed
Feed Location https://feeds.feedburner.com/linuxquestions/latest
Feed Title LinuxQuestions.org
Feed Link https://www.linuxquestions.org/questions/
Reply 0 comments