Izzy's questions

A couple of weeks ago, when I tried switching to the Contacts tab of Evolution, the entire application froze. Well, that sometimes happens with Evolution – so I simply called killev, also took care all its sub-processes were terminated, and restarted it. Same thing still: As soon as I want to switch to contacts, Evolution freezes – and killev reports No response from Evolution -- killing the process.

Now I thought this might be due to several library and kernel updates without reboot, there might be some mismatch – and the "Windows repair method" would fix it. Having finished the other open tasks, today I finally rebooted the machine. To my surprise, this did not fix the issue – behaviour is still the same: as soon as I try to access contacts, Evolution freezes.

System: Ubuntu 12.04 (and no, updating to 14.04 or 16.04 is not an option for this machine)
Evolution: 3.4.2-1ubuntu3.1~stracciatellappa+precise1 (same version for -common, dataserver and plugins); evolution-webcal: 2.32.0-2ubuntu2.1~stracciatellappa+precise1; evolution-indicator: 0.2.20-0ubuntu7

In my ~/.xsession-errors I see a couple of warnings related to Evolution:

(evolution:4096): LIBDBUSMENU-GLIB-WARNING **: Trying to remove a child that doesn't believe we're it's parent.
(evolution:4096): GLib-GObject-WARNING **: cannot register existing type `EABContactDisplay'

Immediately following one of the latest Ubuntu system updates¹, notifications about newly arrived mails stopped working. The corresponding plugin² is still enabled, and I'm not aware of having changed any related configuration. Just to make sure, I've also checked the accounts' settings³ – but couldn't find anything related here. Neither can it be the "global notification daemon", as notify-send "Test" "Dummy notification" brings up the notification as expected.

I'm out of ideas where else I could check (some notification service that needs to be started explicitly but somehow failed? Notifications for other apps like Skype still pop up, and notify-send works as expected). I've already checked syslog for errors, but none are shown.

I'm running Evolution 3.4.2-1ubuntu3.1~stracciatellappa+precise1 on 12.04 here. I've checked for the mail-notification-evolution package, it's not installed⁴ – but as notifications worked before the last reboot, this cannot be the reason.

In case it matters: I'm using LXDE as desktop environment, but Gnome desktop is installed as well (though not selected as default). Again, as it worked in the past (before that last update/reboot), all required components should be available.

Update: I've just figured evolution-indicator should be responsible for notifications here. But there were no changes to that package either within the last year.

Any ideas what I could check/adjust to make notifications from Evolution work again?

^{1: I'm not sure which update exactly it was (there were no updates for Evolution itself); the issue started after the latest reboot. I assume it were the last kernel updates (3.2.0-94), but cannot tell for sure. Walking through the list of latest updates/installs in /var/log/apt/history.log doesn't reveal anything I'd recognize as related.}

^{2: Edit › Plugins › Notifications for new mails}

^{3: Edit › Settings › (Account_Name)}

^{4: according to dpkg -l mail-notification-evolution it never has been installed – it doesn't even have an install candidate in the stracciatella PPA I've used to install Evolution 3.4.2 from quite some time ago; the package in the "standard repo" depends on Evolution < 3.3 and thus cannot be(en) used anyway}

During the last months, the "levels" editor (Color→Levels) was starting up slower and slower – but only when called up via the menus. Called up a second time by clicking the image (when it was the last tool used) made it show up immediately. But even when called up a second time via the menus, the slowness persisted. Recently, it reached the threshold of being unbearable, taking more than 10s to show up.

I'm not aware of any configuration changes or software installs which could have caused this as a side-effect, so I'm clueless of what the culprit might be.

I know I can remove visited pages from the history via History→Show all History, and simply mark the unwanted entries plus hit the Delete key. But this seems not to remove the URLs I've typed into the address bar when calling up those pages. So they are still used to predict the URLs I want to type in later. Especially with sites one heavily frequents (like here on the StackExchange network), this clutters up the list of suggested URLs.

Is there any way to remove the "typed URLs" from the browser history completely? Preferable with some easy clicks for the end-user -- but I wouldn't mind doing so via command-line if necessary (e.g. modifying Firefox's sqlite databases directly).

Edit: Thanks to Dan's answer a part of my issue is explained -- and I found out I missed to fully describe it. Until reading Dan's explanations, I didn't see some of the details.

So let's say I've visited http://readwrite.com/somearticle and http://readwrite.com/otherarticle. And of course a lot of other pages having an "r" somewhere in their URL. Now, on an empty urlbar, I type an "r". Immediately, auto-complete fills that up with "readwrite.com" (not one of the full URLs), and the drop-down shows a bunch of other pages (but none from readwrite.com).

• Dan's first method doesn't work, as the auto-completed string is not contained in the dropdown
• Dan's second suggestion (switching auto-complete off altogether) is no option for me, I just want to get rid of some specific "suggestions"
• I have no matching bookmark (to be more specific: none for readwrite.com)

So how to get rid of that "readwrite.com" suggestion? I've checked the entire history and made sure there's no entry left for it. I've also checked my bookmarks, no traces there either.

I use PHP's virtual() directive quite a lot on one of my sites, including central elements. This worked fine for the last ~10 years -- but after upgrading to 12.04 it somehow got broken.

Example setup (simplified)

To make it easier to understand, I simplify some things (contents). So say I need a HTML fragment like <P>For further instructions, please look <A HREF='foobar'>here</P> in multiple pages. 10 years ago, I used SSI for that, so it is put into a file in a central place -- so if e.g. the targeted URL changes, I only need to update it in one place. To serve multiple languages, I have Apache's MultiViews enabled -- and at $DOCUMENT_ROOT/central/ there are the files:

• foobar.html (English variant, and the default)
• foobar.html.de (German variant).

Now in the PHP code, I simply placed:

<? virtual("/central/foobar"); ?>

and let Apache take care to deliver the correct language variant.

The problem

As said, this worked fine for about 10 years: German visitors got the German variant, all others the English (depending on their preferred language). But after upgrading to Ubuntu 12.04, it no longer worked: Either nothing was delivered from the virtual() command, or (in connection with framesets) it even ended up in binary gibberish.

Trying to figure out what happens, I played with a lot of things. I first thought MultiViews was (somehow) not available anymore -- but calling http://<server>/central/foobar showed the right variant, depending on the configured language preferences. This also proved there was nothing wrong with file permissions. The error.log gave no clues either (no error message thrown).

Finally, just as a "last ressort", I changed the PHP command to <? virtual("central/foobar.html"); ?> -- and that very same file was in fact included. But the language dependend stuff obviously did no longer work.

Of course I tried to find some change (most likely in PHP's virtual() command), using Google a lot, and also searching the questions here -- unfortunately to no avail.

Finally: The question

Putting "design questions" aside (surely today I would design things differently -- but at least currently I miss the time to change that for a quite huge amount of pages): What can be done to make it work again? I surely missed something -- but I cannot figure out what...

Addition (EDIT)

My remaining problem now is that as soon as the document needs to be negotiated (e.g. has different language variants), virtual() always adds the content at the very beginning of the generated document (i.e. in front of the <HTML> tag) -- regardless where it was placed. It however gets placed correctly when no negotiation is involved.

Any idea what's going on there, and how to fix it?

In KDE 3.5 there was the powermanager plugin for the kicker panel, where one could easily switch between different profiles. So normally, I want the display to go to sleep after, say, 10 minutes of user inactivity. But when I watch a video, this should not happen.

I know I could go via the KDE launcher, settings, powermanager, and change the setting. But this is too complicated for daily use.

I also know there are those activities. But I don't want to change background etc., just the display timeout. Furthermore, when I tried this, many of the open apps simply crashed and got restarted. I simply don't like this approach.

Preferably, it would be something to plug-in to the panel, which opens a menu to select the profile -- as it was with KDE 3.5. Is there a similar solution I missed?

I often have to deal with documents using multiple images -- as e.g. Howtos, short documentations, or even books with many examples. Always when I start a new document with LibreOffice Writer and insert the first pictures, they are embedded by default. I need them to be linked by default -- but could not find a suitable setting. Can anybody tell me where that's hidden? I know it was possible with StarOffice 8, which I used before upgrading to Ubuntu 12.04.

In other words (to be clear from the beginning): How do I tell LibreOffice to always default to link inserted images instead of embedding them?

I know there's a command to let compiz take over from whatever windows manager is running, using compiz --replace &, same for unity with unity --replace &. What are the similar commands for other window managers (e.g. KDE4's plasma, Gnome, LXDE, XFCE)? And will this take all running application windows along?

On a fresh started system, free reports about 1.5G used RAM (8G RAM alltogether, Ubuntu 12.04 with lightdm and plasma desktop, one konsole window started). Having the apps running I use, it still consumes not more than 2G. However, having the system running for a couple of days, more and more of my free RAM disappears -- without showing up in the list of used apps: while smem --pie=name reports less than 20% used (and 80% being available), everything else says differently. free -m for example reports on about day 7:

             total       used       free     shared    buffers     cached
Mem:          7459       7013        446          0        178        997
-/+ buffers/cache:       5836       1623
Swap:         9536        296       9240

(so you can see, it's not the buffers or the cache). Today this finally ended with the system crashing completely: the windows manager being gone, apps "hanging in the air" (frameless) -- and a popup notifying me about "too many open files". Syslog reports:

kernel: [856738.020829] VFS: file-max limit 752838 reached

So I closed those applications I was able to close, and killed X using Ctrl-Alt-backspace. X tried to come up again after that with failsafeX, but was unable to do so as it could no longer detect its configuration. So I switched to a console using Ctrl-Alt-F2, captured all information I could think of (vmstat, free, smem, proc/meminfo, lsof, ps aux), and finally rebooted. X again came up with failsafeX; this time I told it to "recover from my backed-up configuration", then switched to a console and successfully used startx to bring up the graphical environment.

I have no real clue to what is causing this issue -- though it must have to do either with X itself, or with some user processes running on X -- as after killing X, free -m output looked like this:

             total       used       free     shared    buffers     cached
Mem:          7459       2677       4781          0         62        419
-/+ buffers/cache:       2195       5263
Swap:         9536         59       9477

(~3.5GB being freed) -- to compare with the output after a fresh start:

             total       used       free     shared    buffers     cached
Mem:          7459       1483       5975          0         63        730
-/+ buffers/cache:        689       6769
Swap:         9536          0       9536

Two more helpful outputs are provided by memstat -u. Shortly before the crash:

User     Count     Swap      USS      PSS      RSS
mail         1        0      200      207      616
whoopsie     1      764      740      817     2300
colord       1     3200      836      894     2156
root        62    70404   352996   382260   569920
izzy        80   177508  1465416  1519266  1851840

After having X killed:

User     Count     Swap      USS      PSS      RSS
mail         1        0      184      188      356
izzy         1     1400      708      739     1080
whoopsie     1      848      668      826     1772
colord       1     3204      804      888     1728
root        62    54876   131708   149950   267860

And after a restart, back in X:

User     Count     Swap      USS      PSS      RSS
mail         1        0      212      217      628
whoopsie     1        0     1536     1880     5096
colord       1        0     3740     4217     7936
root        54        0   148668   180911   345132
izzy        47        0   370928   437562   915056

Edit: Just added two graphs from my monitoring system. Interesting to see: everytime when there's a "jump" in memory consumption, CPU peaks as well. Just found this right now -- and it reminds me of another indicator pointing to X itself: Often when returning to my machine and unlocking the screen, I found something doing heavvy work on my CPU. Checking with top, it always turned out to be /usr/bin/X :0 -auth /var/run/lightdm/root/:0 -nolisten tcp vt7 -novtswitch -background none.

So after this long explanation, finally my questions:

1. What could be the possible causes?
2. How can I better identify involved processes/applications?
3. What steps could be taken to avoid this behaviour -- short from rebooting the machine all X days?

I was running 8.04 (Hardy) for about 5 years on my old machine, never having experienced the like (always more than 100 days uptime, before rebooting for e.g. kernel updates). This now is a complete new machine with a fresh install of 12.04. In case it matters, some specs:

AMD A4-3400 APU with Radeon(tm) HD Graphics, using the open-source ati/radeon driver (so no fglrx installed), 8GB RAM, WDC WD1002FAEX-0 hdd (1TB), Asus F1A75-V Evo mainboard. Ubuntu 12.04 64-bit with KDE4/Plasma. Apps usually open more or less permanently include Evolution, Firefox, konsole (with Midnight Commander running inside, about 4 tabs), and LibreOffice -- plus occasionally Calibre, Gimp and Moneyplex (banking software I'm already using for almost 20 years now, in a version which did fine on Hardy).

Edit: Today I found one of the "evil guys": KDE4s plasma-desktop. Used memory was again up to 5GB, when I did a killall plasma-desktop && plasma-desktop. This freed up 1.3GB RAM! ps says:

                             RSS    SIZE   VSZ
plasma usage before restart  120988 526472 1300816
plasma usage after restart   92352  495972 1263632

So where have those 1.3GB been? The diff between those values, if added up, amount to 96MB -- not 1.3GB.

And this can only be one part, as still 3.7GB are in use (should be less than 2GB). I monitored this during the past 6 days using several tools: used memory (not talking about cache and buffers) increases slowly but steadily. Even if I'm not at my desk to run anything...

As to monitoring processes with open files, I currently use the following 1-liner (I love shell and especially bash) to get the top-5:

echo "$(for pid in $(ls -a /proc|egrep '^([0-9])*$'|sort -n 2>/dev/null); do \
if [ -e /proc/$pid/fd ]; then FHC=$(ls -l /proc/$pid/fd|wc -l); \
if [ $FHC -gt 0 ]; then PNAME="$(cat /proc/$pid/comm)"; \
echo "$FHC files opened by $pid ($PNAME)"; fi; fi; done)"|sort -r -n|head -n5

Command here in 4 lines for better readability. Nothing much yet from there -- except that Skype doesn't like have the Internet connection broken. Each disconnect causes a slight increase of its open files, but nothing dramatic. On the other hand it seems like plasma is also responsible for that:

See the drop of file handles at the end? That was the plasma restart.

My provider forces a re-connect all 24h, which breaks my XChat session. Though the connection usually is re-established within seconds, due to the changed IP the "old session" is invalidated. XChat automatically joins the network(s), but then my nick gets refused as it "is already in use". I can manually solve that by "/nickserv ghost MyUser MyPass" -- but I could not find any place to configure XChat to do that automatically. Is that possible, and if so, how?

I'm fighting with Evolution. Basically it's working fine -- but it is very slow to react in certain situations.

Helper questions

1. Could it be that changing away from Bonobo has to do with slowing-down?

2. There might be some trouble with the new engine and "asynchronous actions". What to do about it? Are there e.g. any configuration files?

3. I want to get the previous "working mood" back. How can I speed this thing up?

Different scenarios

• when sending a mail, the composer window hangs there inactive for a couple of seconds, everything grayed out (the bigger the message/attachments, the longer it takes). Though there is a green check mark saying it's sent, I'm not sure a) why it's still blocking everything and b) whether I could simply close it without "breaking"/"losing" anything. In earlier versions, the composer window was closing pretty fast, and one could see the message being stored into the local "outbox" until it was sent, and one could immediately continue with the next task. I prefer that behaviour over the current, where I cannot do anything in the application until the window closes.
• switching between modules. Coming from mail and switching to the address book takes a couple of seconds. Same for switching to the calendar. I read about different "possible causes" and tried a few things:
  • I only have 3 local address books, so no networking should be involved here. To make sure, I switched to offline mode and then tried to access the address book. No noticeable difference.
  • I use 3 Google Calendars. Switching to offline mode made a minor difference, but so minor that it also could be "imagination" since one might have expected this in this case
  • according to some reports, disabling the tasks should help. Well, it didn't in my case, as I don't use them regularly (just two local items stored here)

Maybe I should also mention that I'm using the KDE4 desktop (so no Unity or Gnome, though both is installed on the computer). And I did not have this issue before I updated to 12.04.

Due to a system crash, I had to do a fresh install on a new computer. Luckily my harddisk was not affected -- but unfortunately the "good old 8.04" didn't want to boot from it on the new system...

So with 12.04 installed on a fresh harddisk, I copied all evolution folders from the old disk to the corresponding locations (~/.evolution plus ~/.gconf/apps/evolution -- the old Evo on 8.04 had no ~/.local/share/evolution), and started the app. The migration assistant popped up, and I went through -- but in the end, only one of my "accounts" in each category was migrated and available: local mail, one of 4 IMAP accounts (though data of all 4 accounts where moved to ~/.local/share/evolution), one of 4 address books, and so on.

Not a big deal for the IMAP stuff, as I just had to configure it (data are stored on the server in this case). But how can I get my addressbooks back? As described above, starting the "old installation" to export those data is not an option, as the old system cannot be started anymore. Is there a way to fire the migration assistant for a given dataset (in my case: tell it to just migrate-and-import a single specified addressbook)?