Author: Paolo Redaelli

The lamentations of colord and gdm3

Posted on by Paolo Redaelli

I recently installed a machine dedicated to transmit live Holy Masses from my parish and I discovered a subtle malfunctioning of the hardware that sometimes makes it unusable: the Solid State Disk becomes read-only. As you can imagine, almost everything stopped working. Also logging, of course. As I had to hunt it out and catch it, I redirected it on an external USB key, discovering tons of logs like

nov 20 07:51:27 carloacutis colord[426039]: failed to get session [pid 1500]: Nessun dato disponibile (i.e. no data available)

and entire descriptions of the video modes available in the form of something like

nov 14 09:44:52 carloacutis /usr/lib/gdm3/gdm-x-session[2830]: (II) AMDGPU(0): #0: hsize: 1920  vsize 1080  refresh: 60  vid: 49361
 nov 14 09:44:52 carloacutis /usr/lib/gdm3/gdm-x-session[2830]: (II) AMDGPU(0): #1: hsize: 1680  vsize 1050  refresh: 60  vid: 179
 nov 14 09:44:52 carloacutis /usr/lib/gdm3/gdm-x-session[2830]: (II) AMDGPU(0): #2: hsize: 1600  vsize 1200  refresh: 60  vid: 16553
 nov 14 09:44:52 carloacutis /usr/lib/gdm3/gdm-x-session[2830]: (II) AMDGPU(0): #3: hsize: 1600  vsize 900  refresh: 60  vid: 49321

It happens every time the screen blanks.

In a machine currently used by a human, all those tons of logs – something in the range of a hundred of megabytes per day (in Debian bug #813334 it is labelled as “syslog spam”) are not produces. But the screen of this machine remain almost always blank as I’ve automated the transmissions.

I somehow hacked gdm3 with this two onliner patches:

--- gdm3-3.34.1.orig/daemon/gdm-server.c
+++ gdm3-3.34.1/daemon/gdm-server.c
@@ -242,7 +242,7 @@ gdm_server_init_command (GdmServer *serv
         if (debug)
             verbosity = "7";
         else
-            verbosity = "3";
+            verbosity = "0";
 #endif
 
         if (g_access (SYSTEMD_X_SERVER, X_OK) < 0) {
--- gdm3-3.34.1.orig/daemon/gdm-x-session.c
+++ gdm3-3.34.1/daemon/gdm-x-session.c
@@ -282,7 +282,7 @@ spawn_x_server (State        *state,
         if (state->debug_enabled) {
                 g_ptr_array_add (arguments, "7");
         } else {
-                g_ptr_array_add (arguments, "3");
+                g_ptr_array_add (arguments, "0");
         }
 
         if (state->debug_enabled) {

I should push them if I was sure it’s not harmful to anyone else.

I still have not found how to tell those colord daemon that it can stay quiet.

Time has come

Posted on by Paolo Redaelli

What a change from the original “Don’t-be-evil” Google! The shiny knight is slowly but constantly turning into an harsh evil tyrant. See what Yeupou writes at the beginning of Fetching mails from gmail.com with lieer/notmuch instead of fetchmail – # cd /scratch

Google is gradually making traditional IMAPS access to gmail.com impossible, in it’s usual opaque way.

So the time has come to leave Google. It will be a tiring effort, it will be difficoult, but it has to be done.

Che sciatteria!

Posted on by Paolo Redaelli

Da qualche anno racconto agli studenti di architettura come “metter le mani” su qualche calcolo di strutture in acciaio ed in calcestruzzo.

Mi piace far prendere loro contatto con le Norme Tecniche per le Costruzioni e la relativa Circolare interpretativa che in Italia hanno valenza di legge a tutti gli effetti, specialmente le prime. E dove indirizzarli se non sul sito del Consiglio Superiore dei Lavori Pubblici che dovrebbe pubblicare questi documenti in originale?

Bene, guardate cosa mi son trovato a leggere oggi quando ho provato a scaricare la circolare dal collegamento diretto che avevo messo sulla pagina del corso dell’anno scorso:

E va già bene che non han messo un laconico errore 404!

Ma che sciatteria! Un sito governativo non può e non deve permettersi di lasciare carenze informative di questo genere. Ciò denota una grossolana mancanza di professionalità che non deve essere tollerata.