Bug#1032381: aide: Broken aideinit due to _aide user handling
Yep this is a systemd system The only relevant customisation I can think of is adduser.conf set to 'DIR_MODE=0750' (which I reverted to the new 0700 default as part of the upgrade, but not sure if that happened before or after aide was upgraded), and proc mounted with 'hidepid=2' (which I know systemd sometimes isn't happy with) Will get a separate bug written up soon On 02/05/2023 12:41, Marc Haber wrote: Hi Dark, please file a new bug report, that gives more information about your system. Are you using systemd? Greetings Marc
Bug#873073: [Pkg-samba-maint] Bug#873073: Other ways to disable IPv6 ?
On 24/07/19 22:14, Mathieu Parent wrote: > Le mer. 24 juil. 2019 à 00:27, Dark Penguin a écrit : >> >> On 24/07/19 00:15, Mathieu Parent wrote: >>> Version: 2:4.9.5+dfsg-5 >>> >>> I'm closing this bug with the version in buster. >>> >>> Le dim. 21 juil. 2019 à 21:45, Dark Penguin a >>> écrit : >>>> >>>> I tried this just now. The result is, basically, nothing: my kernel does >>>> not have IPv6 support anyway, so restricting IPv6 out on the systemd >>>> level does not change anything. There are still error messages about >>>> being unable to bind to IPv6 upon restarting smbd, however with this, >>>> restarting it also takes a few seconds instead of happening almost >>>> instantly. >>>> >>>> I guess the "proper" solution would be the same: if there are no IPv6 >>>> interfaces in the system, smbd should not try to bind to them. If it was >>>> specifically instructed to bind to a certain interface and it is >>>> unavailable, then output an error message "This interface is requested >>>> but unavailable", instead of "open_socket_in(): socket() call failed: >>>> Address family not supported by protocol". This error message is not >>>> even decipherable without Google's help. >>> >>> The following binds to IPv4 only: >>> >>> interfaces = lo 0.0.0.0 >>> bind to interfaces only = yes >>> >>> Regards >> >> >> Umm... I've actually tried that before. > > Confirmed with: >interfaces = 127.0.0.0/8 0.0.0.0 >bind interfaces only = yes > > > $ sudo ss -lntp | grep smbd > LISTEN0 50 127.0.0.1:445 > 0.0.0.0:*users:(("smbd",pid=9146,fd=32)) > LISTEN0 50 127.0.0.1:139 > 0.0.0.0:*users:(("smbd",pid=9146,fd=33)) We should probably reopen this then?.. -- darkpenguin
Bug#873073: [Pkg-samba-maint] Bug#873073: Still present in Buster
Oh, there was a file like that?.. ^_^ Ooohkay, this has revealed something interesting. In /etc/default/networking, it says: # Which interface to wait for. # If none given, wait for all auto interfaces, or if there are none, # wait for at least one hotplug interface. #WAIT_ONLINE_IFACE= I have two interfaces: "auto lo" and "allow-hotplug eth0". So, in my case, by default it waits for lo, and then the network is assumed to be up. (Apparently "if there are none" includes lo, which makes little sense.) Changing this to "auto eth0" has fixed the smbd problem! I remember having read somewhere that "allow-hotplug implies auto", so I usually change eth0 to allow-hotplug without expecting any consequences of this kind. Turns out that with "allow-hotplug", the system does not wait for the interface to come up at boot, that's why samba starts without it. Changing it back to "auto" fixes this problem, but introduces another one. For some reason, DHCP address acquisition takes 5 to 8 seconds, so the downside is adding 8 seconds to the system boot time. So, to sum it up: - If you are running a samba server, then either run it on an "auto" interface, or restart it after configuring a hotplug interface by adding this line in the interface's definition: up /bin/systemctl restart smbd.service nmbd.service || true This actually makes sense. If you reconfigure a hotplug interface, you'd probably want to restart everything that is supposed to listen on it. (Come to think about it, why doesn't sshd have this problem?.. Maybe because it listens on 0.0.0.0 ?.. Specifying 0.0.0.0/0 as an interface to listen on in smb.conf does not help.) - Still, it would be nice to fix samba so that it does not try to bind to nonexistent interfaces (even though it is not even asked to), and does not output a high-level uninformative error message when it fails to do that. This would allow us to not worry about IPv6 and not see the strange red error message at the same time. It could probe those interfaces if it wants to bind to all of them, but a negative probing result is expected, and is not a high-level error. On 15/07/19 16:36, Mathieu Parent wrote: > Le dim. 14 juil. 2019 à 14:51, Dark Penguin a écrit : >> >> This issue is still present in Buster. >> >> Changing the interface name to its IP address in smb.conf does not make >> sense because we're talking about a situation when your IP address is >> managed by DHCP. >> >> Tweaking the systemd service files does not help according to the OP, >> and even if it did, this is not a solution - network-online.target >> apparently still does not work. >> >> Mathieu, is there a bug report with more details about the >> network-online.target problem? It looks pretty serious since it can >> potentially break all network-based daemons, and is still not fixed in >> Buster. > > Have you considered tuning /etc/default/networking ? > > For example: > WAIT_ONLINE_METHOD="ifup" > WAIT_ONLINE_IFACE="enp2s0" > WAIT_ONLINE_ADDRESS="" > WAIT_ONLINE_TIMEOUT=300 > > or: > WAIT_ONLINE_METHOD="ping" > WAIT_ONLINE_IFACE="" > WAIT_ONLINE_ADDRESS="128.31.0.62" > WAIT_ONLINE_TIMEOUT=300 > > See also /lib/ifupdown/wait-online.sh > > What is the content of /etc/network/interfaces? > > Regards > -- darkpenguin
Bug#873073: [Pkg-samba-maint] Bug#873073: Other ways to disable IPv6 ?
I tried this just now. The result is, basically, nothing: my kernel does not have IPv6 support anyway, so restricting IPv6 out on the systemd level does not change anything. There are still error messages about being unable to bind to IPv6 upon restarting smbd, however with this, restarting it also takes a few seconds instead of happening almost instantly. I guess the "proper" solution would be the same: if there are no IPv6 interfaces in the system, smbd should not try to bind to them. If it was specifically instructed to bind to a certain interface and it is unavailable, then output an error message "This interface is requested but unavailable", instead of "open_socket_in(): socket() call failed: Address family not supported by protocol". This error message is not even decipherable without Google's help. On 15/07/19 16:51, Mathieu Parent wrote: > Le dim. 14 juil. 2019 à 15:09, Dark Penguin a écrit : >> >> This particular case could be less of a problem (at least to me) if >> adding "bind interfaces only = yes" was not the only way to disable IPv6 >> in samba. >> >> Would it make sense to change the priority of an error message on >> startup about missing IPv6 support? If samba sees that IPv6 is not >> supported on this system, shouldn't it deduce that it is therefore not >> required, and output an "info" message like "Not binding to IPv6 - >> protocol not supported" instead of an error? > > Have you tried overriding with: > > cat /etc/systemd/system/smbd.service > [Service] > RestrictAddressFamilies=AF_UNIX AF_INET > > (then systemctl daemon-reload and systemctl restart smbd) > > (Not tested...) > > Regards > -- darkpenguin
Bug#873073: Other ways to disable IPv6 ?
This particular case could be less of a problem (at least to me) if adding "bind interfaces only = yes" was not the only way to disable IPv6 in samba. Would it make sense to change the priority of an error message on startup about missing IPv6 support? If samba sees that IPv6 is not supported on this system, shouldn't it deduce that it is therefore not required, and output an "info" message like "Not binding to IPv6 - protocol not supported" instead of an error? -- darkpenguin
Bug#873073: Still present in Buster
This issue is still present in Buster. Changing the interface name to its IP address in smb.conf does not make sense because we're talking about a situation when your IP address is managed by DHCP. Tweaking the systemd service files does not help according to the OP, and even if it did, this is not a solution - network-online.target apparently still does not work. Mathieu, is there a bug report with more details about the network-online.target problem? It looks pretty serious since it can potentially break all network-based daemons, and is still not fixed in Buster. -- darkpenguin
Bug#929892: vfat mounts are not protected
Package: bilibop-lockfs Version: 0.5.6 In lockfs mode, if there are any vfat mounts to be protected, they are ignored and simply mounted as read-write. Everything else is properly protected with overlayfs. Isn't it kind of dangerous?.. The user is expecting everything to be safe! If it is not possible to protect a vfat filesystem due to overlayfs limitations, then maybe it should be mounted read-only instead of read-write? At least then the user will notice it before breaking something, and add it to the whitelist if they want it read-write, easily replicating the current behaviour. And currently, there is no way to do the opposite - which is, "in lockfs mode, mount everything that can not be properly protected as read-only". This is reproducible on Debian Buster amd64 , and also on Raspbian Stretch with bilibop 0.5.6 from Debian Buster armhf. -- darkpenguin
Bug#929893: Emergency shell due to broken parsing
Package: bilibop-lockfs Version: 0.5.6 Steps to reproduce: # apt-get install bilibop --no-install-recommends Now, add a single line to the bottom of /etc/bilibop/bilibop.conf : BILIBOP_LOCKFS_WHITELIST=/anything It does not matter if this directory exists or not. Now do an "update-initramfs -u -k all", and reboot. You won't be able to boot, because all mounts except "/" will fail, giving you an emergency shell. Attempting to mount one of them manually (by "mount /mydir") gives me the following result: /sbin/mount.lockfs: 1: eval: net.ifnames=0: not found This is apparently a part of my kernel command line. Here it is, just in case: # cat /proc/cmdline BOOT_IMAGE=/boot/vmlinuz-4.19.0-5-amd64 root=UUID=64e67ea6-1627-4258-9c70-8bcc120daf86 ro console=tty1 console=ttyS0 ipv6.disable=1 net.ifnames=0 consoleblank=0 lockfs This is reproducible on Debian Buster amd64 and Raspbian Stretch with bilibop 0.5.6 from Debian Buster armhf. Other lines in bilibop.conf work fine, only the "whitelist" one seems to break things. However, whitelisting other mounts or marking them "noauto" in /etc/fstab seems to help, despite the option itself being the source of the problem. %) -- darkpenguin
Bug#929891: Error message about udev rules not found
Package: bilibop-rules Version: 0.5.6 After installing bilibop, even without enabling lockfs, I see this error message during the boot process: --- Begin: Running /scripts/local-bottom ... Warning: bilibop-lockfs: Nothing to do. ls: /root/etc/udev/rules.d/70-persistent-cd.rules: No such file or directory ls: /root/etc/udev/rules.d/70-persistent-net.rules: No such file or directory done. --- With lockfs enabled, it is still there: --- Begin: Running /scripts/local-bottom ... Warning: bilibop-lockfs: Locking files. Begin: bilibop-lockfs: Setting up overlay branches ... done. Success: bilibop-lockfs: Root filesystem is now locked (soft policy). Warning: bilibop-lockfs: Modifying temporary static filesystem table (fstab). ls: /root/etc/udev/rules.d/70-persistent-cd.rules: No such file or directory ls: /root/etc/udev/rules.d/70-persistent-net.rules: No such file or directory done. --- This is reproducible on Debian Buster amd64, and I saw it on Raspbian Stretch with bilibop 0.5.6 from Debian Buster armhf. I don't see any negative consequence of this, other than making the user scared a little, but it is still an error message meant to be noticed. Maybe it will help find other bugs. :) -- darkpenguin
Bug#929890: Error during installation and uninstallation
Package: bilibop-rules Version: 0.5.6 During installation AND uninstallation, I see the following error message: --- Setting up bilibop-rules (0.5.6) ... dpkg: error: --compare-versions takes three arguments: Type dpkg --help for help about installing and deinstalling packages [*]; Use 'apt' or 'aptitude' for user-friendly package management; Type dpkg -Dhelp for a list of dpkg debug flag values; Type dpkg --force-help for a list of forcing options; Type dpkg-deb --help for help about manipulating *.deb files; Options marked [*] produce a lot of output - pipe it through 'less' or 'more' ! --- I don't see any other negative effect of this, but an error message is there for someone to notice, so I noticed it. :) This is reproducible on Debian Buster amd64 (debootstrapped installation in qemu), and Raspbian Stretch with bilibop 0.5.6 from Debian Buster armhf. -- darkpenguin
Bug#816874: Bug#816779: Remmina doesn't always honor scale ratio lock
Yes, this was a bug in 1.1.2, which was buggy. So I moved to 1.2.0, which was not in the repo at the time, but without this bug. -- darkpenguin
Bug#845445: mutter: illegal instruction error
Package: mutter Version: 3.14.4-1~deb8u1 Severity: important Dear Maintainer, When trying to use mutter I receive a illegal instruction error. Try this with two terminals: Terminal 1: Xephyr -ac -br -noreset -screen 800x600 :1 Terminal 2: mutter -d :1 I expected Mutter functioning within a Xephyr window. Instead I got: "[1] 21548 illegal hardware instruction mutter -d :1". -- System Information: Debian Release: 8.6 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Kernel: Linux 3.16.0-4-amd64 (SMP w/2 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages mutter depends on: ii gnome-themes-standard 3.14.2.2-1 ii gsettings-desktop-schemas 3.14.1-1 ii libc6 2.19-18+deb8u6 ii libclutter-1.0-0 1.20.0-1 ii libglib2.0-0 2.42.1-1+b1 ii libmutter0e 3.14.4-1~deb8u1 ii libx11-6 2:1.6.2-3 ii libxcomposite1 1:0.4.4-1 ii mutter-common 3.14.4-1~deb8u1 ii zenity 3.14.0-1 Versions of packages mutter recommends: ii openbox [x-session-manager] 3.5.2-8 Versions of packages mutter suggests: ii gnome-control-center 1:3.14.2-3 ii xdg-user-dirs 0.15-2
Bug#787950: mdadm/checkarray: broken on current debian systems
The easiest fix for this is changing #!/bin/sh to #!/bin/bash in that file. And I believe it's the "correct" fix, because the shebang line is invented specifically to tell you which interpreter to use, so it you write a script for Bash, you should put Bash there, not Dash or whatever sh could be installed on that system. And I think Bash is installed by default on almost any system, exactly because a lot of scripts require specifically Bash to execute. Changing only two letters in one file. To fix a grave-level bug that has been there for a year. In Debian STABLE. Is it too much to ask?.. By the way, how come this bug report is not "grave", but "normal"?! Going on without periodical scrubs is a direct way to lose all your data, which is, by definition, "grave", correct? And the bug affects 100% of the users. All of them. Except maybe the ones who changed their default shell to Bash, which is hardly a lot. Actually, I have a list of similar bugs in Jessie that are around for over a year, with extremely simple fixes (if you know about the bug in the first place). Because of this tendency to ignore such bugs, Debian Stable is now becoming the most unstable distro out there, effectively failing its only mission: "The goal of the Debian Project is Debian Stable" (I can't find the exact document on the Debian website in which I read this only yesterday). -- darkpenguin
Bug#835840: Bugs in Remmina
Just came here to ask for better Remmina support. Use it every day, and use Wheezy and Devuan exclusively. What's the hold up, biggest blocker / barrier? How can I help? Apparently, it's just the number of issues in FreeRDP 2.0. Actually, I understand the reasoning, but I think this is an exceptional situation specifically described somewhere in the Debian release policy: something like "usually, packages make their way from unstable into testing when the number of bugs is low enough, but sometimes the maintainer can decide that the advantages of having that package outweigh the unstability risks, so they an force it manually". We're not talking about "from unstable into testing" right now, but the idea is the same: anyone who actually tried both versions agrees that the new version that works but suggests an unstable package is better than the old version that does barely work at all. Actually, there are only 18 bugs in Remmina 1.2.0 versus 20 bugs in Remmina 1.1.2 in Jessie, 5 of which are "important" (3 of them being segfaults). I'd say 18 < 20 , really. (I hope you did not count wishlist items, "notabug"-s, "questions", "outofscope"-s and everything else while looking at the number of open issues on Github? You did filter them by "bugs" label, right?.. And we don't necessarily have to package FreeRDP 2.0 - having Remmina alone would be better already.) My experience with Remmina 1.1.2 on Jessie: - Every few minutes, there is a segfault (important) - When changing my layout with Alt+Shift, Alt gets stuck (important) - Scaling doesn't work properly (important) - Shift+Tab doesn't get across (normal, if you don't use it often) - Sometimes the main window opens when connecting from tray (minor) Now, I'm sorry for some off-topic, but the only package that can beat the record of five bugs everyone is affected by is nut, which has four 100% reproducible grave bugs that render the whole critically important package not only useless, but harmful (it's UPS management software - you can damage your hardware and lost your data because of those bugs). Basically, it does not shut down your machine upon power loss (which is the purpose of the package), it does not shut down your UPS after that, and two common workarounds for that do not work. That's partly because of systemd, which is somehow deemed stable enough to be not only present, but default and mandatory in Debian. Of course, all those are fixed in Remmina 1.2.0, and it works flawlessly. -- darkpenguin
Bug#835840: Bugs in Remmina
Have you actually tried either of them? I don't know how many bugs are in Remmina 1.1.2 and how many in Remmina 1.2.0, but my actual experience (I use Remmina a lot every day) shows that Remmina 1.1.2 is utterly broken and barely usable, while 1.2.0 works just fine. Actually, I notice this kind of things more and more lately. Debian Stable is becoming the most unstable distro because nobody is interested in fixing bugs for software this old - and for some reason, even the most obvious bugs have not been fixed before the release. In the newer versions, all those things have been fixed years ago, but those fixes do not make their way into stable. I don't have anything against old software, but not buggy software, and it's called "stable" because it's supposed to actually be stable. I've been using exclusively Debian Stable for many years, but seems like it's time to abandon ship because of this tendency. (Why is it that Remmina is "not yet stable enough", but systemd is?.. That's what most of those problems are with...) By the way, I'm not using FreeRDP - only Remmina, for VNC. And no, Remmina does NOT "depend" on FreeRDP - you can run it without FreeRDP. So maybe it would be an acceptable solution to package Remmina 1.2.0 as a separate package called "remmina-next", and not package the "unstable" FreeRDP at all?.. -- darkpenguin
Bug#835634: [Nut-upsuser] "Halt then reboot" script with systemd
I've asked on the systemd mailing lists whether it's possible with systemd. Maybe this thread could be helpful... https://lists.freedesktop.org/archives/systemd-devel/2016-September/037447.html -- darkpenguin
Bug#835634: nut: POWEROFF_WAIT option has no effect
Package: nut Version: 2.7.2-4 Severity: important The POWEROFF_WAIT option in nut.conf does not work (has no effect). I have a suspicion that it has something to do with systemd handling system shutdowns nowadays, though it's too complex for me to investigate further and find any proof (other than /sbin/shutdown and /sbin/poweroff pointing to /bin/systemctl, which does not prove anything). This bug introduces quite a regression I believe is worth fixing in Jessie (even if it turns out to be fixed in later versions). Also, it becomes especially hilarious together with #83 - nut also doesn't send the command to kill the power; so nut doesn't handle the case when your UPS supports cutting the power, AND it doesn't handle the situation when it does not. Seriously, did no one notice a 100% broken package BEFORE shipping Jessie?.. A package that manages power supply, of all things. Hence the "important" severity. Details: If the command for the UPS to kill the power is sent when AC power is already back, my UPS ignores it. (And it's APC, so I really didn't expect that kind of thing from them.) So, there is a risk to have all my machines require a manual restart after a power failure. And nut has a workaround just for this case... Actions required: - Add "POWEROFF_WAIT=15m" line in /etc/nut/nut.conf Intended behaviour: - After an emergency shutdown triggerd by NUT, the system is not powered down, only halted. After a timeout (15 minutes in this example) it reboots, in case the UPS did not cut the power for some reason. Actual behaviour: - The system is always powered off, ignoring the option added. Minor notes: - man nut.conf has a section about this option which directs me to /usr/share/doc/nut/docs/shutdown.txt.gz , which is not there anymore. - man nut.conf does say that "if you want to specify time in seconds, use the 's' suffix", but it does not say which suffix should I use if I do NOT want to specify time in seconds. According to the example below, using "m" is also possible - apparently, for minutes, but I was wondering if "m" is already deprecated (along with shutdown.txt.gz mentioned above) and I should use no suffix by default... I don't want to do a real power failure test again just to find out. -- System Information: Debian Release: 8.5 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.1.18-aufs (SMP w/2 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=ru_RU.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages nut depends on: ii nut-client 2.7.2-4 ii nut-server 2.7.2-4 nut recommends no packages. nut suggests no packages. -- no debconf information -- darkpenguin
Bug#835555: nut does not kill power after shutdown
Package: nut Version: 2.7.2-4 Severity: important After an emergency shutdown due to power loss, the command to kill the power is never sent to the UPS. And since all the machines are powered down, the load is zero, so this will continue for a pretty long time, during which power is very likely to return. Because of this, not only do the connected machines require a manual power-on, but the UPS and the battery may be damaged (batteries do not like discharging completely, and UPS-es do not like overheating due to operating at low wattages, according to their manuals). I know that NUT has a workaround for this situation for UPS-es that do not support killing the power (POWEROFF_WAIT). However, people KNOW that their UPS DOES support it, so they will not have it turned on. And even if they do a test run (and does everyone do a completely real power failure test run after an upgrade to Jessie from a perfectly working Wheezy installation?), they may not be patient enough to wait until the UPS cuts the power automatically, because the timeout may be set to something pretty long. My investigation revealed that the power kill command is supposed to be sent by /lib/systemd/system-shutdown/nutshutdown (which is really neither obvious nor documented, as far as I could find, but oh well): /sbin/upsmon -K >/dev/null 2>&1 && /sbin/upsdrvctl shutdown However, /sbin/upsmon -K ALWAYS returns false, even if the flag is there (I've tried replacing "/sbin/upsdrvctl shutdown" with a debug command and it never got executed). Changing that line to "if [ -f /etc/killpower ] ; then /sbin/upsdrvctl shutdown ; fi" fixes the problem for me. I do not know why does upsmon -K not work as intended, but there are many possible reasons: - it likes to complain that "the file is there, but I can't read it, so what if there's no magic word?! I won't do it without the magic word!" - it likes to complain that "one UPSMON is already running" (and systemd seems to be particularly good at starting upsmon when it's not needed and not being able to stop it when you try to do that) - maybe some other permission problems between "root" and "nut" (see: https://bugs.launchpad.net/ubuntu/+source/nut/+bug/839490 and http://askubuntu.com/questions/511387/nut-not-shutting-down-ups-does-14-04-server-block-usb-comms-in-runlevel-0 ) -- System Information: Debian Release: 8.5 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.1.18-aufs (SMP w/2 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=ru_RU.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages nut depends on: ii nut-client 2.7.2-4 ii nut-server 2.7.2-4 nut recommends no packages. nut suggests no packages. -- no debconf information -- darkpenguin
Bug#834747: nginx-extras: Feature request: Add 3rd party module graphite-nginx-module.
Please note that there is currently a "Won't Fix" for "No New Modules or Flavors", which may still apply here. This is bug #790623 (https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=790623) -- Thomas On Thu, Aug 18, 2016 at 10:39 AM, Roman V. Nikolaevwrote: > Package: nginx-extras > Version: 1.10.1-1 > Severity: wishlist > > Dear Maintainer, > > Please add new module to nginx-extras: > graphite-nginx-module - an nginx module for collecting location stats into > Graphite. > > Url: https://github.com/mailru/graphite-nginx-module > License: BSD > Depends: lua-nginx-module > > -- System Information: > Debian Release: 8.4 > APT prefers stable > APT policy: (990, 'stable'), (500, 'testing-updates'), (500, > 'testing-proposed-updates'), (500, 'stable-updates'), (500, > 'proposed-updates'), (500, 'unstable'), (500, 'testing'), (500, 'oldstable'), > (1, 'experimental') > Architecture: amd64 (x86_64) > > Kernel: Linux 4.6.0-1-amd64 (SMP w/8 CPU cores) > Locale: LANG=ru_RU.UTF-8, LC_CTYPE=ru_RU.UTF-8 (charmap=UTF-8) > Shell: /bin/sh linked to /bin/dash > Init: systemd (via /run/systemd/system) > > Versions of packages nginx-extras depends on: > ii libc6 2.23-2 > ii libexpat1 2.1.0-6+deb8u2 > ii libgd3 2.2.1-1 > ii libgeoip1 1.6.2-4 > ii liblua5.1-0 5.1.5-7.1 > ii libpam0g1.1.8-3.1+deb8u1+b1 > ii libpcre32:8.35-3.3+deb8u4 > pn libperl5.18 > ii libssl1.0.0 1.0.1k-3+deb8u5 > ii libxml2 2.9.3+dfsg1-1 > ii libxslt1.1 1.1.28-2+b2 > ii nginx-common1.6.2-5+deb8u2 > ii perl5.20.2-3+deb8u5 > pn perlapi-5.18.1 > ii zlib1g 1:1.2.8.dfsg-2+b1 > > nginx-extras recommends no packages. > > Versions of packages nginx-extras suggests: > pn nginx-doc >
Bug#834164: systemd: Services are not killed after a timeout or killed immediately after ExecStop
Package: systemd Version: 215-17+deb8u4 Severity: normal I am trying to set up a service shutdown sequence in a way that the service is asked nicely to quit, and if it doesn't, then it's killed after a timeout. Sounds very simple - this is very basic. According to man systemd.service, TimeoutStopSec= Configures the time to wait for stop. If a service is asked to stop, but does not terminate in the specified time, it will be terminated forcibly via SIGTERM, and after another timeout of equal duration with SIGKILL (see KillMode= in systemd.kill(5)). However, with this setup: ExecStop= KillMode=none TimeoutStopSec=10s the service is NEVER killed - even after 10s timeout. Maybe I don't understand what's written in the manual? I just can't believe such a basic thing could seriously be overlooked up until now... I've also tried the same with KillMode=process (and without KillMode at all - it defaults to control-group), hoping that maybe it will give my service some time to shut down after ExecStop=, but it killed the process immediately (and it went into "failed" mode for not stopping quickly enough). Why do we have ExecStop= at all if the service is killed immediately anyway? This behaviour us the same in a newer version of systemd in Linux Mint 18, so it's an upstream bug. (Either that, or I really don't understand something that simple.) Here is a sample .service file that can be used safely to test this behaviour. = = = = = = = test.service = = = = = = = [Unit] Description=Test service [Service] ExecStart=/usr/bin/screen -h 1024 -dmS test ftp nonExistentServer Type=forking ExecStop=/usr/bin/screen -p 0 -S test -X eval 'stuff \"ServiceGoDie\"\015' # Change the command to "bye" to actually let the "service" quit # Otherwise it is supposed to fail to quit peacefully # and be killed after 10s after "systemctl stop test" KillMode=none TimeoutStopSec=10s [Install] WantedBy=multi-user.target = = = = = = = -- Package-specific info: -- System Information: Debian Release: 8.5 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.1.18-aufs (SMP w/2 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=ru_RU.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages systemd depends on: ii acl 2.2.52-2 ii adduser 3.113+nmu3 ii initscripts 2.88dsf-59 ii libacl1 2.2.52-2 ii libaudit1 1:2.4-1+b1 ii libblkid1 2.25.2-6 ii libc6 2.19-18+deb8u4 ii libcap2 1:2.24-8 ii libcap2-bin 1:2.24-8 ii libcryptsetup4 2:1.6.6-5 ii libgcrypt20 1.6.3-2+deb8u1 ii libkmod218-3 ii liblzma55.1.1alpha+20120614-2+b3 ii libpam0g1.1.8-3.1+deb8u1+b1 ii libselinux1 2.3-2 ii libsystemd0 215-17+deb8u4 ii mount 2.25.2-6 ii sysv-rc 2.88dsf-59 ii udev215-17+deb8u4 ii util-linux 2.25.2-6 Versions of packages systemd recommends: ii dbus1.8.20-0+deb8u1 ii libpam-systemd 215-17+deb8u4 Versions of packages systemd suggests: pn systemd-ui -- no debconf information -- darkpenguin
Bug#826068: X11vnc layouts problem
I did some research. I totally mis-diagnosed the problem earlier; seems like it's a problem with key translation (or how is it called properly?). First of all, I've confirmed that this is indeed a problem with upstream x11vnc. It's present in Linux Mint 18 Cinnamon, it doesn't matter which VNC viewer do I use (the problem persists even with x11vnc's "Java viewer"), and it's not present with other VNC servers. (Do I have to go upstream in this case? Where?) I've read the manpage carefully and tried all kinds of keyboard-related options, and some of them had some effect, but I couldn't get the "correct" behaviour, so I"m trying to find out what the problem is by looking at the "x11vnc -dk" output. And I can't figure it out on my own, as I know very little about how keyboards work and how the keys are supposed to be transferred... (But I see that other VNC servers got it perfectly right, so it should be possible; I just really like x11vnc and don't want to go with the other ones!) The problem is: - I have two layouts on my local machine: "us" and "ru" - Everything works fine when the only layout on the remote machine is "us"; my local layout is honored and all keys work properly. - When I add the "ru" layout on the remote machine, the problems start: everything is fine as long as my local layout is "us", but if I switch it to "ru", then most keys are mangled in a way hard to describe in a brief way. I believe the correct behaviour would be similar to what vino has: the remote layout is completely ignored, and only the local layout is honored. The only thing that helped me partially was disabling modtweak; it only mangled my keys even further. So I'm looking at the debug output without modtweak. (And I'm omitting all the "Shift" presses and all the releases for uncluttered view.) Case #1: - "ru" is NOT added on the remote machine - the local layout is "ru" I press й (located on "q") - I get й : 09/08/2016 05:46:23 # keyboard(down, 0x6ca "Cyrillic_shorti") uip=0 2058.5674 09/08/2016 05:46:23 keyboard(): KeySym 0x6ca "Cyrillic_shorti" -> KeyCode 0x5d 09/08/2016 05:46:23 XTestFakeKeyEvent(dpy, keycode=0x5d "Cyrillic_shorti", down) 09/08/2016 05:46:23 calling XTestFakeKeyEvent(93, 1) 2058.5685 I press " (located on Shift+2) - I get " : 09/08/2016 05:48:26 # keyboard(down, 0x22 "quotedbl") uip=0 2182.3190 09/08/2016 05:48:26 keyboard(): KeySym 0x22 "quotedbl" -> KeyCode 0x30 09/08/2016 05:48:26 XTestFakeKeyEvent(dpy, keycode=0x30 "apostrophe", down) 09/08/2016 05:48:26 calling XTestFakeKeyEvent(48, 1) 2182.3194 I press № (located on Shift+3) - I get № : 09/08/2016 05:50:22 # keyboard(down, 0x6b0 "numerosign") uip=0 2297.9047 09/08/2016 05:50:22 keyboard(): KeySym 0x6b0 "numerosign" -> KeyCode 0x78 09/08/2016 05:50:22 XTestFakeKeyEvent(dpy, keycode=0x78 "numerosign", down) 09/08/2016 05:50:22 calling XTestFakeKeyEvent(120, 1) 2297.9071 Case #2: - adding "ru" to the remote machine, but not enabling it - the local layout is "ru" I press й (located on "q") - but for some reason I get q : 09/08/2016 05:52:17 # keyboard(down, 0x6ca "Cyrillic_shorti") uip=0 2413.0940 09/08/2016 05:52:17 keyboard(): KeySym 0x6ca "Cyrillic_shorti" -> KeyCode 0x18 09/08/2016 05:52:17 XTestFakeKeyEvent(dpy, keycode=0x18 "q", down) 09/08/2016 05:52:17 calling XTestFakeKeyEvent(24, 1) 2413.0950 (Why is it being translated back into English?! Who's doing it and how to disable it?!) I press " (located on Shift+2) - I get " : 09/08/2016 05:53:27 # keyboard(down, 0x22 "quotedbl") uip=0 2483.1056 09/08/2016 05:53:27 keyboard(): KeySym 0x22 "quotedbl" -> KeyCode 0x30 09/08/2016 05:53:27 XTestFakeKeyEvent(dpy, keycode=0x30 "apostrophe", down) 09/08/2016 05:53:27 calling XTestFakeKeyEvent(48, 1) 2483.1066 (Why is this NOT translated into English, then?!) I press ; (located on Shift+3) - I get : : 09/08/2016 05:57:24 # keyboard(down, 0x3b "semicolon") uip=0 2720.3259 09/08/2016 05:57:24 keyboard(): KeySym 0x3b "semicolon" -> KeyCode 0x2f 09/08/2016 05:57:24 XTestFakeKeyEvent(dpy, keycode=0x2f "semicolon", down) 09/08/2016 05:57:24 calling XTestFakeKeyEvent(47, 1) 2720.3265 (It is a SEMI-colon in both "us" and "ru" - why does it turn into a COLON?!) Case #3: - the remote layout is "ru" - the local layout is "ru" I press й (located on "q") - I get й: 09/08/2016 06:01:12 # keyboard(down, 0x6ca "Cyrillic_shorti") uip=0 2948.2130 09/08/2016 06:01:12 keyboard(): KeySym 0x6ca "Cyrillic_shorti" -> KeyCode 0x18 09/08/2016 06:01:12 XTestFakeKeyEvent(dpy, keycode=0x18 "q", down) 09/08/2016 06:01:12 calling XTestFakeKeyEvent(24, 1) 2948.2142 I press " (located on Shift+2) - I get Э (located on the ' and " key ): 09/08/2016 06:02:28 # keyboard(down, 0x22 "quotedbl") uip=0 3023.8559 09/08/2016 06:02:28 keyboard(): KeySym 0x22 "quotedbl" -> KeyCode 0x30 09/08/2016 06:02:28 XTestFakeKeyEvent(dpy, keycode=0x30 "apostrophe", down) 09/08/2016 06:02:28 calling XTestFakeKeyEvent(48, 1) 3023.8568 I
Bug#816779: Remmina doesn't always honor scale ratio lock
This is one of two bugs with Remmina that really annoy me; after a while, I've gathered more data and now I think have the simplest way to reproduce the problem. Also, this time all the machines are running Debian Jessie with MATE. - I connect over VNC to one machine with a 1024x768 screen. A small window appears in the center of my FullHD screen. (This is a good starting point to see all kinds of scaling problems that I have.) - Then I connect to another machine with a FullHD screen. The window is still the same size, but now a FullHD screen is cramped into it (without honoring my "Maintain scale ratio" option, of course.) Now if I resize the window by dragging its border, or click "Maintain scale ratio" twice to disable it and enable it back, then it resizes the image properly; but whatever else I do, the scale ration is not honored. What I've tried: - Being not happy, I try to "Maximize" the window. The image is now stretched in a thin horizontal line with black borders above and below - neither honoring the scale ratio, nor even scaling to fit all the screen. - If I click "Resize the windows to fit in remote resolution", the same happens. - If I click "Toggle fullscreen mode", the same happens. - After any of those, if I toggle the "Scaling mode" off and back on, then it scales to cover all the window, but still does not honor the "Maintain scale ratio". I've also tried this with RDP - connecting to Windows machines. The results are the same. And finally, only in the case when the remote screen is SMALLER than the window, turning "Scaled mode" off and on helps resize the window properly. When the remote screen is LARGER than the window, it does not honor the aspect ratio. -- darkpenguin
Bug#826068: X11vnc layouts problem
Package: x11vnc Version: 0.9.13-1.2+b2 Severity: normal When the local layout is non-English, capital letters (and also "Shift-" keys) come out lower-case (or "without Shift"). Example: - I connect to a Debian Jessie machine running x11vnc over MATE (I've also tried x11vnc on LMDE2 with Cinnamon) using Remmina on Debian Jessie or TightVNC on Windows. Both local and remote layouts are english by default. Everything is fine at this point. - I press Alt+Shift to change layouts. This changes both local and remote layouts (to Russian, in my case, but I assume it would be the same with any non-English layout). Now, all upper-case that I type come out lower-case, and if I try to type '!' , I get '1' on the other side. - If I change the remote layout by clicking it with my mouse, but leave my local layout non-English, the problem persists. However, if I change my local layout back to English, then the problem is gone, whichever layout there is on the remote side. -- System Information: Debian Release: 8.4 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.1.18-aufs (SMP w/2 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages x11vnc depends on: ii libavahi-client3 0.6.31-5 ii libavahi-common3 0.6.31-5 ii libc6 2.19-18+deb8u4 ii libjpeg62-turbo 1:1.3.1-12 ii libssl1.0.0 1.0.1k-3+deb8u5 ii libvncclient0 0.9.9+dfsg2-6.1+deb8u1 ii libvncserver0 0.9.9+dfsg2-6.1+deb8u1 ii libx11-6 2:1.6.2-3 ii libxdamage1 1:1.1.4-2+b1 ii libxext6 2:1.3.3-1 ii libxfixes31:5.0.1-2+b2 ii libxinerama1 2:1.1.3-1+b1 ii libxrandr22:1.4.2-1+b1 ii libxtst6 2:1.2.2-1+b1 ii openssl 1.0.1k-3+deb8u5 ii tk8.6.0+8 ii x11vnc-data 0.9.13-1.2 ii zlib1g1:1.2.8.dfsg-2+b1 x11vnc recommends no packages. x11vnc suggests no packages. -- no debconf information -- darkpenguin
Bug#816875: Update
I've tried it with relatively recent versions of x11vnc and tightvncserver on Debian Jessie and Linux Mint. It works. Apparently, it doesn't work with older versions of various VNC servers on various operating systems... but I don't think this is something that really needs to be fixed. I think we can close this bug now. Sorry for not checking it with recent versions of VNC servers in the beginning - I did not have an opportunity back then... -- darkpenguin
Bug#820897: Update
I've tried it with relatively recent versions of x11vnc and tightvncserver on Debian Jessie and Linux Mint. It works. Apparently, it doesn't work with older versions of various VNC servers on various operating systems - upon closer inspection, it seems that the Alt key tends to get "stuck" if you release it before releasing Shift... but I don't think this is something that really needs to be fixed. I think we can close this bug now. Sorry for not checking it with recent versions of VNC servers in the beginning - I did not have an opportunity back then... -- darkpenguin
Bug#816874: New information
First of all, I mistyped about LMDE: I mean, I'm using MATE on Jessie, of course. Now, I've tried to gather more data. - Click the Remmina tray icon, then quickly (within the double-click timeout) click one of the items in the context menu: the item is activated, but the main Remmina window also pops up. (This is what's most annoying.) - Move the mouse over the tray icon, hold the left mouse button, quickly move the mouse over to some item in the context menu, release the button: the item is activated, but the main Remmina window also pops up. - Left-click the tray icon, then immediately right-click it (just for testing purposes): the main Remmina window pops up. However: - Double-right-click the tray icon: nothing happens. - Right-click the icon, then immediately left-click it: the main window does not pop up. - Click the tray icon (the context menu pops up). Now, double-click the icon (intending to open the main window). It does not pop up - the context menu disappears and then reappears again, but the main window does not pop up. -- darkpenguin
Bug#820897: Remmina - buggy Alt+Shift
Package: remmina Version: 1.1.1-2 I switch my keyboard layout with Alt+Shift on all my computers. I use Remmina on MATE on Debian Jessie to connect to other computers - an old version of tightvncserver on Windows XP, Vino on Debian Squeeze, and tightvncserver on Debian Wheezy (I'm not sure if I've also tried this with X11vnc on Debian Jessie - I can't confirm this at the moment). I believe Remmina is to blame because I don't have this problem with an old TightVNC viewer running on Windows XP/7. Let's say I'm editing an OpenOffice Writer document, for example. Expected behaviour: - I press Alt+Shift - Layout changes (on both local and remote computers) - I can continue typing - the Alt key press does not get across. Observed behaviour: - I press Alt+Shift - Layout changes (on both local and remote computers) - The "File" menu is highlighted - because the Alt key press has gotten across to the remote computer. Now if I continue typing, I'll accidentally launch menu items instead, so I have to remember to press Alt again after every layout change. Here is the reportbug information. -- System Information: Debian Release: 8.3 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.1.18-aufs (SMP w/2 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages remmina depends on: ii dbus-x111.8.20-0+deb8u1 ii libavahi-client30.6.31-5 ii libavahi-common30.6.31-5 ii libavahi-ui-gtk3-0 0.6.31-5 ii libc6 2.19-18+deb8u3 ii libgcrypt20 1.6.3-2+deb8u1 ii libgdk-pixbuf2.0-0 2.31.1-2+deb8u4 ii libglib2.0-02.42.1-1 ii libgtk-3-0 3.14.5-1+deb8u1 ii libssh-40.6.3-4+deb8u2 ii libvte-2.90-9 1:0.36.3-1 ii libx11-62:1.6.2-3 ii remmina-common 1.1.1-2 Versions of packages remmina recommends: ii remmina-plugin-rdp 1.1.1-2 ii remmina-plugin-vnc 1.1.1-2 remmina suggests no packages. -- no debconf information -- darkpenguin
Bug#816875: Remmina does not transfer Shift+Tab over VNC
Hi, you can use reportbug for reporting a bug. Reportbug collect some very useful informations for the maintainer. Please run 'reportbug remmina -N 816779' to give me this infos. And on which operating system and which VNC Server has the error occurred? I'm running the client on MATE on Debian Jessie. And now that I think about it, it may be an issue not with Remmina, but with both tightvncserver and x11vnc, since I can reproduce it from the Windows TightVNC client... Or maybe it's an issue with both Remmina and TightVNC client?.. This needs further investigation... Here is the reportbug output. -- System Information: Debian Release: 8.3 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.1.18-aufs (SMP w/2 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages remmina depends on: ii dbus-x111.8.20-0+deb8u1 ii libavahi-client30.6.31-5 ii libavahi-common30.6.31-5 ii libavahi-ui-gtk3-0 0.6.31-5 ii libc6 2.19-18+deb8u3 ii libgcrypt20 1.6.3-2+deb8u1 ii libgdk-pixbuf2.0-0 2.31.1-2+deb8u4 ii libglib2.0-02.42.1-1 ii libgtk-3-0 3.14.5-1+deb8u1 ii libssh-40.6.3-4+deb8u2 ii libvte-2.90-9 1:0.36.3-1 ii libx11-62:1.6.2-3 ii remmina-common 1.1.1-2 Versions of packages remmina recommends: ii remmina-plugin-rdp 1.1.1-2 ii remmina-plugin-vnc 1.1.1-2 remmina suggests no packages. -- no debconf information -- darkpenguin
Bug#816874: Remina tray icon treats everything like a double-click
> Hi, > > you can use reportbug for reporting a bug. Reportbug collect some very > useful informations for the maintainer. > > Please run 'reportbug remmina -N 816874' to give me this infos. > > And on which desktop environment has the error occurred? I'm using LMDE2 on Debian Jessie. Here is the reportbug output. -- System Information: Debian Release: 8.3 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.1.18-aufs (SMP w/2 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages remmina depends on: ii dbus-x111.8.20-0+deb8u1 ii libavahi-client30.6.31-5 ii libavahi-common30.6.31-5 ii libavahi-ui-gtk3-0 0.6.31-5 ii libc6 2.19-18+deb8u3 ii libgcrypt20 1.6.3-2+deb8u1 ii libgdk-pixbuf2.0-0 2.31.1-2+deb8u4 ii libglib2.0-02.42.1-1 ii libgtk-3-0 3.14.5-1+deb8u1 ii libssh-40.6.3-4+deb8u2 ii libvte-2.90-9 1:0.36.3-1 ii libx11-62:1.6.2-3 ii remmina-common 1.1.1-2 Versions of packages remmina recommends: ii remmina-plugin-rdp 1.1.1-2 ii remmina-plugin-vnc 1.1.1-2 remmina suggests no packages. -- no debconf information -- darkpenguin
Bug#816779: Remmina doesn't always honor scale ratio lock
> you can use reportbug for reporting a bug. Reportbug collect some very > useful informations for the maintainer. > > Please run 'reportbug remmina -N 816779' to give me this infos. > > And on which operating system and which VNC Server has the error > occurred? Hello, Sorry for the delay - I've been kind of on a vacation... I didn't know exactly what information really is useful and what is only littering the report. I'm trying to connect to an old tightvncserver on Windows XP and Vino on Debian Squeeze (but I remember having similar problems no matter where I connect - I've definitely tried something modern, most probably x11vnc on a LMDE2 LiveCD). Now I'm trying to analyze Remmina's behaviour, and I've made some interesting observations. Apparently, it does re-scale the image properly if you resize the window by dragging its border, but sometimes it fails - you drag the window to the new size, and the image is neither scaled correctly nor even stretched to occupy the whole window - instead, it's stretched in a disfigured way somewhere in between. The same happens when you Maximize the window or un-Maximize it. Can you not reproduce this problem on your system, whatever it is?.. Also, you can try the following cases: - I connect with "Scaled Mode" and "Scale Ration Lock" enabled. Instead of maximizing the window and scaling the image maintaining the ratio, the window remembers the last used size, and stretches the image without maintainig the ratio. - I resize the window a little by dragging its border. The image is now scaled correctly. - I disable "Scaled Mode" by clicking a button on the toolbar and enable it again. Now the image is stretched incorrectly again, taking my whole window. I drag its border again to make it scale properly again. I'm running MATE on Debian Jessie. Here is the reportbug output. -- System Information: Debian Release: 8.3 APT prefers stable-updates APT policy: (500, 'stable-updates'), (500, 'stable') Architecture: amd64 (x86_64) Foreign Architectures: i386 Kernel: Linux 4.1.18-aufs (SMP w/2 CPU cores) Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash Init: systemd (via /run/systemd/system) Versions of packages remmina depends on: ii dbus-x111.8.20-0+deb8u1 ii libavahi-client30.6.31-5 ii libavahi-common30.6.31-5 ii libavahi-ui-gtk3-0 0.6.31-5 ii libc6 2.19-18+deb8u3 ii libgcrypt20 1.6.3-2+deb8u1 ii libgdk-pixbuf2.0-0 2.31.1-2+deb8u4 ii libglib2.0-02.42.1-1 ii libgtk-3-0 3.14.5-1+deb8u1 ii libssh-40.6.3-4+deb8u2 ii libvte-2.90-9 1:0.36.3-1 ii libx11-62:1.6.2-3 ii remmina-common 1.1.1-2 Versions of packages remmina recommends: ii remmina-plugin-rdp 1.1.1-2 ii remmina-plugin-vnc 1.1.1-2 remmina suggests no packages. -- no debconf information -- darkpenguin
Bug#816875: Remmina does not transfer Shift+Tab over VNC
Package: remmina Version: 1.1.1-2 In a VNC connection, Shift+Tab combination just does not get across to the remote desktop. It does not have any result at all, with or without "Grab Keyboard" enabled. Shift+Tab is one of the very basic shortcuts for people who mostly use keyboard and not mouse: open some application on the remote screen, enter the login, Tab, enter the password... oh, I made a typo in the login! Shift+Tab... Nothing happens! You have to reach for the mouse to click the login field again... Shift+Tab does work correctly over RDP, though. -- darkpenguin
Bug#816874: Remina tray icon treats everything like a double-click
Package: remmina Version: 1.1.1-2 I have Remmina autostart when I log in, so the Remmina icon sits in the tray. Whenever I want to connect to some host, I click the tray icon and the I choose an item from the list of saved hosts. Now, I only have very few hosts, so those two clicks happen in less than a second. And for some reason, Remmina treats it like a double-click and opens the main program window, even though the first click is on the icon, and the second click is on a menu item. And I have to close the main window every time, or remember to make a long pause after clicking the tray icon and before choosing the host to connect to. Also, instead of two separate clicks, I usually do it the Linux way: click on the tray icon, hold the button down, move it to the menu item I want, and release the button. I find this really convenient. But even in this case Remmina treats it like a double-click, even though it was actually only one click and release!! -- darkpenguin
Bug#787950: How about accepting this patch?
I think this situation is a disaster. Not having periodical checks on your arrays, or rather, having them fail quietly without even notifying the admin, can lead to grave consequences. (People tend to underestimate how critical it is to have periodical consistency checks, especially in RAID5's). Finding this bug report is not exactly easy. I've spent a whole day trying to find out why do I not have consistency checks, and I could not advance any further than "why does checkarray not do anything". Not to mention it was a very lucky coincidence that I even noticed not having consistency checks. So, how about accepting this patch and having checkarray use bash by default? I see no downsides to this; bash is an "essential" package, so most people do have bash installed; even if they don't, they will still not have critical maintenance fail silently, but get an error message. Even if this is fixed in more recent kernels, 4.1-LTS is one of the most popular ones, if not THE most popular, and it still has this problem. But still... "One byte at a time"? Was this the reason why Linus wrote one of his most famous quotes about this being incredibly stupid rather than simply "inefficient"? %) Is there at least a bug filed against dash about, um, "optimizing the inefficient behaviour potentially leading to major problems on most widely used LTS kernels"? -- darkpenguin
Bug#816779: Remmina doesn't always honor scale ratio lock
Package: remmina Version: 1.1.1-2 Even if the "Maintain scale ratio" button is enabled, and this is saved in the connection, Remmina does not always honor it. This happens in some cases on window size change, and sometimes even right after connecting. How to reproduce: - Connect via VNC to a computer whose screen is larger than your screen. Expected behaviour: - A connection window appears that takes all available space on the screen. It has the remote screen in the center, scaled down with correct aspect ratio, and black borders around it. Actual behaviour: - A connection windows appears that takes all available space on the screen. It has the remote screen stretched to take all available space, with wrong aspect ratio and no black borders. - If I click the "Lock ratio" (or whatever it's called) button two times in a row (disabling and re-enabling it), Remmina applies the ratio lock, and black borders appear. Having to make three additional clicks right after starting each connection simply because Remmina forgets to apply its saved options is annoying. I'm using Debian Jessie x64 with MATE and all available updates - just installed it few days ago, so it's almost "a clean installation", and the Remmina package from the repository with VNC and RDP plugins. I haven't tried to reproduce the bug with RDP - I don't have a suitable RDP server available. -- darkpenguin
Bug#813010: caja: Shift-Del behaviour is wrong when renaming files
And, no, it's not only up to the developers to figure what causes the bug. Both Mate and Debian are community-driven, so if you find a bug you would like to get fixed as a user, you should help developers with some testing which helps locate the issue. The more information you provide, the faster can others fix your problem. Remember, many people here don't get paid for fixing these things, so they might as well say that they don't care about a bug if it takes too much effort to debug and fix it. Yes, I know; and I absolutely agree - when reporting a bug, I make sure to test everything I can think of that I would do if I was the project developer. Actually, I like fixing UI problems myself, and I would do it if I could, but I'm not yet good enough to deal with things like this... What I meant is, "it's there in Debian, it's not there in Mint; the only people who can find out exactly where it all went wrong is the Debian people". Also, the bug report page for Debian said something like, "if you're unsure, don't send the report upstream; maybe it's only there in Debian - we'll direct it upstream if we have to". > Could you please test caja upstream and report back? Linux Mint is > not upstream, the MATE project is. > > And if it turns out that the bug is still present upstream, then > you're barking up the wrong tree as Debian just packages most > upstream software, we don't develop it. > > Also, if turns out to be an upstream bug, then you should yell at the > Linux Mint people to fix it. I'm extremely sorry, but it seems that I've provided some wrong information. I've just tested it again. The bug IS there in Linux Mint Debian Edition 2; however, it is NOT there in Linux Mint 17.3 . (I mixed up different Mint distros.) So the problem is upstream, anyway, and it's not with Caja - it's with something else that's different between Debian and Ubuntu. -- darkpenguin
Bug#813010: caja: Shift-Del behaviour is wrong when renaming files
The next problem is to find the bugtracker for MATE. Unsurprisingly, googling "bugs mate" yield unwanted and graphic results. %) -- darkpenguin
Bug#813010: caja: Shift-Del behaviour is wrong when renaming files
Package: caja Version: 1.8.2-3+deb8u1 Severity: normal Dear Maintainer, When renaming files in Caja, pressing Shift+Del does not "cut the name"; instead, it works as if you wanted to "permanently delete" the file. Ctrl+X works as expected, however. To reproduce: - Press F2 on any file to rename it - Press Shift+Del to cut the highlighted part of the name Expected behaviour: - It cuts the highlighted part of the file name. Observed behaviour: - Renaming stops, and a confirmation dialog appears: "Are you sure you want to permanently delete this file?" This buggy behaviour is present in Caja 1.8.2 in the current Jessie, but not in Caja 1.8.2 in LMDE2. -- darkpenguin
Bug#813010: caja: Shift-Del behaviour is wrong when renaming files
When renaming files in Caja, pressing Shift+Del does not "cut the name"; instead, it works as if you wanted to "permanently delete" the file. Ctrl+X works as expected, however. Meh, you people are really nitpicky. I would have expected that Shift+Del will actually delete the file, not cut the text here. I find Shift+Del for Cut and Shift+Ins for Paste much more convenient to use, and I use those shortcuts all the time. When that does not work as expected, it is very uncomfortable... But yes, I'm as nitpicky as one can get about "user interface should be as polished and comfortable as possible", because it directly influences the user's experience. I think it is very important. But usually people are too lazy to file bugs about such minor things, which is why we have a lot of amazing software with UI so buggy that no wonder people prefer proprietary software sometimes... =/ > I think it's arguable whether this is a bug or not. But in any case, When I am in the process of editing a file name, the "text edition shortcut" should obviously be a priority - it should cut selected text. Why would I even want to rename a file before Shift-Deleting it?!.. Even if I did, I'd just press Escape to cancel or Enter to confirm, and then delete the file. this is up to the MATE upstream developers what the expected behavior of Shift+Del is in this case. In fact, you should not test any other Linux distribution, but download caja in source from the upstream website, compile it yourself and test how the vanilla version behaves. If the upstream version behaves as the Debian package, it's an upstream bug. If not, they probably changed behavior upstream and caja just needs to be updated in Debian. I highly doubt that someone intentionally changed this behavior in Debian. I believe Linux Mint is as vanilla as it gets with MATE. And their version works correctly, so they think this behaviour is not expected. It may be that Mint "fixed" this already, but I think it's been more than a year already - why is it still not in Debian?! So I think it's a problem with something other than Caja itself - some Debian-specific hotkey-snatching subsystems or something. Maybe this problem was never even there in Mint. Either way, it's up to the developers to find out why is there a problem with Debian, but not with Mint. -- darkpenguin
Bug#766957:
I can maybe take a stab at backporting to 0.206, as it's on my radar to do anyways. Note this though: I don't think we should outright disable SSLv3 in a stable release. There is a code commit in the pull requests queue waiting for inclusion that allows the specification of SSLv3 being disabled - this in turn would allow 0.206 users to disable SSLv3 if they wish and not change a default that would maybe cause undesired confusion and results. Downstream in Ubuntu, the Security Team will not include the SSLv3-disabled-always changes, but may consider the pending Configure disabling of protocols commits. -- Thomas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#767456: disable SSLv3 by default
Okay, so after poking #debian-security on OFTC, Thijs said the following: (Or at least I believe it's Thijs): [2014/11/01 11:25:15] thijs_ teward: I think the ideal package does not have SSLv3 included in its default settings. With Apache in Debian is quite the case because /etc/apache2/conf-available/ssl.conf will disable SSLv3 so any vhost using SSL without explicitly overriding the SSLProtocols will not have it [2014/11/01 11:26:28] thijs_ that nginx disables it in a configuration example is good, but I would think it's much better if someone creates a vhost without explicit protocol specification, it would not do SSLv3 To that end, I went fussing around with the code of the SSL module. Attached is a patch which should do the trick, and disable SSLv3 support if ssl_protocols is NOT defined. Before this patch is included, though, we should really consider whether we actually *want* to disable SSLv3 by default and potentially break nginx configurations which need SSLv3 and don't have ssl_protocols defined. At the very least, a NEWS entry needs to be added for this. If this change is accepted, I'll make a blog post about it, but only if it's included. -- Thomas disable_sslv3_default_protocol.patch Description: Binary data
Bug#767456: disable SSLv3 by default
fixed 1.6.2-3 thanks Confirmed: This was done already. The commit this was done in was this one: http://anonscm.debian.org/cgit/collab-maint/nginx.git/commit/?id=9a4e0f0a698bee2b03b7f417ad9286e5eb22141e 1.6.2-3, which had this fix already, was uploaded and accepted to Unstable on 2014-10-16, according to the package tracker (https://packages.qa.debian.org/n/nginx.html) This is confirmed in the 1.6.2-4 changelog in Unstable (http://metadata.ftp-master.debian.org/changelogs/main/n/nginx/unstable_changelog). Dissection of the package 1.6.2-4 also shows that the default SSL stanza has an ssl_protocols line of `ssl_protocols TLSv1 TLSv1.1 TLSv1.2`. Coupled with the OpenSSL updates made by the Debian security team to support TLS_FALLBACK_SCSV to prevent the protocol downgrade attack from TLSv1 to SSLv3, POODLE is effectively mitigated with the now-default config stanzas for SSL. (This assumes also that a user is using the default SSL config sections. A large portion of (albeit newer) users do use the default config stanzas, or at least use it as a base, and it can be argued that competent administrators will already disable the vulnerable protocols in their own site configs separately.) -- Thomas Ward On Fri, Oct 31, 2014 at 7:28 AM, Thomas Ward tew...@dark-net.net wrote: I thought this was already done? I checked the packaging myself and this change was already in there, or at least in git. (the default ssl stanza in the config has SSLv3 dropped from the ciphers list in the git tree for the Debian package already, I checked the commit logs myself) -- Thomas On Oct 31, 2014, at 03:37, Thijs Kinkhorst th...@debian.org wrote: Package: nginx Version: 1.6.2-2 Severity: important Hi, Please disable the legacy SSLv3 protocol by default for installations of nginx. It doesn't need to be disabled completely per se, but should not be available on a default installation. This helps to defend against the recent POODLE attack (CVE-2014-3566). Thanks, Thijs -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#762494: Please update the Lua module to upstream git master
Apparently upstream has posted a new tagged version of the Lua module, v0.9.13rc1. Can we update the package to include this tagged revision, to address this bug? (Also note: bug retitled) -- Thomas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#764527: Package Tracker URL Problem Resolution and Request to Look Into Updating the Package
Source: libpff Severity: wishlist Hello. Firstly, it should be made known that packages.qa.debian.org is showing that there is a problem with the URLs for this package. According to what I could find, the libpff project moved away from sourceforge. It looks like it was located on Google Code for a while, but has since moved to GitHub at https://github.com/libyal/libpff (this assumption is based on the owner of the libyal group on GitHub being the same person who filed the ticket at SourceForge requesting project removal - http://sourceforge.net/p/forge/site-support/3913/ and is also based on Google searches for the libpff project.) Secondly, the version currently provided by Debian appears to be several years old. Downstream in Ubuntu and elsewhere, it looks like users are trying to use these tools but are forced to compile the software to get the newer software. While this is typically not an issue for power users, for the less technically inclined with the coding and compile-from-scratch side of OSS it is a disadvantage. To that end, it may be prudent to investigate whether the software can be updated in Debian. -- Thomas -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#762494: Please update the Lua module to upstream git master
Source: nginx Severity: wishlist Hello. I would like to request that the Lua module be updated to the version in its latest git master branch. I am requesting this early, ahead of future upstream versions of nginx uploaded to Debian. This request is based on the Mainline branch, which I package in the downstream Ubuntu PPAs, and directly base its debian/ folder off of Debian Unstable. The latest mainline branch, 1.7.5 changes the API and causes the Lua module to fail to build in its current form. While this API change does not change here, there is added into the module an `if` block which determines which code to use based on the version of NGINX it's being built with. As this is related to a future hypothetical fail to build error, I would like the module to be updated to include the fix ahead of future API changes in the nginx versions. -- Thomas
Bug#587657: locally installed plugins aren't loaded
On Thu, Jul 1, 2010 at 3:57 AM, Niels Thykier wrote: With the upgrade from 3.5.2-3 to 3.5.2-4 all locally installed plugins aren't loaded anymore, while they appear in the Installed Software tab. Neither uninstall/install of the plugins nor an clean ~.eclipse/ and ~/workspace folder solves the problem. ~/workspace/.metadata/.log is empty. Please let me know if I can provide any further information. Can you please provide the contents of [1] from your original ~/.eclipse (if you still have it) and the one generated when start eclipse with a clean ~/.eclipse. [1] .eclipse/org.eclipse.platform_*/configuration/org.eclipse.equinox.simpleconfigur ator/bundles.info Same problem encountered, and I just solved my problem temporarly. As pointed out by Niels, I made a quick diff of these two .info file and it turned out to be the problem of org.eclipse.equinox.concurrent. Previously, this package's location was at: `org.eclipse.equinox.concurrent,1.0.1.R35x_v20100209,../../../home/lisnaz/.eclipse/org.eclipse.platform_3.5.0_155965261/plugins/org.eclipse.equinox.concurrent_1.0.1.R35x_v20100209.jar,4,false' a jar under local plugin directory, I don't know why it went there but it must be there for reason. After modifed the path to the newly generated .info's location: `org.eclipse.equinox.concurrent,1.0.1.R35x_v20100209,plugins/org.eclipse.equinox.concurrent_1.0.1.R35x_v20100209.jar,4,false' then every plugin works fine. This solves my problem, but it may also means something to Hannes's problem. -- Best regards! Vincent Gu lis...@gmail.com NE RD
Bug#544515: MD Listing in the United States
Many of these lists are on sale too. Just email me here: jillian.gr...@marketlistgurus.at By emailing disapp...@marketlistgurus.at you will have your email taken off -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org