subuser-security / subuser

Run programs on linux with selectively restricted permissions.
http://subuser.org
GNU Lesser General Public License v3.0
889 stars 65 forks source link

UnicodeEncodeError: 'ascii' codec can't encode character '\u2192' in position 88043: ordinal not in range(128) #325

Open mailinglists35 opened 7 years ago

mailinglists35 commented 7 years ago

I'm getting this trace at the end of an apparent successful subuser add command Attaching log since it's pretty big for pasting. Installed on Ubuntu 16.04 with kernel hwe 4.10 amd64 from launchpad subuser_0.6.1-3_all.deb

Is it something to worry?

mailinglists35 commented 7 years ago

subuser.txt

timthelion commented 7 years ago

This appears to be fixed in git. Unfortunately, I haven't had time lately to make a realease. I'll move it up my todo list.

On 07/14/2017 09:45 AM, mailinglists35 wrote:

subuser.txt https://github.com/subuser-security/subuser/files/1147682/subuser.txt

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/subuser-security/subuser/issues/325#issuecomment-315294691, or mute the thread https://github.com/notifications/unsubscribe-auth/ABU7-MsV02NcB7fS-cLkPl12cHobz8Bfks5sNxylgaJpZM4OX716.

timthelion commented 7 years ago

I've made a release to pypy. I'm currently waiting for feedback before I notify distro maintainers.

If you want to help me test it out you should be able to do so with:

sudo pip3 install subuser
mailinglists35 commented 7 years ago

sure. would that conflict with the version already installed from launchpad deb package?

timthelion commented 7 years ago

Yes, no, there is a two way conversion process for your .subuser folder, you'll need to run subuser repair after installing the pypy(pip3) version. Also if you have both installed at the same time, then the deb package will probably be first in the $PATH. You'll want the executable at:

which subuser

/usr/local/bin/subuser

On 07/28/2017 07:32 PM, mailinglists35 wrote:

sure. would that conflict with the version already installed from launchpad deb package?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/subuser-security/subuser/issues/325#issuecomment-318715582, or mute the thread https://github.com/notifications/unsubscribe-auth/ABU7-OLd5CPCv3b2PVjV3JvQYpUNH6cBks5sShsPgaJpZM4OX716.

mailinglists35 commented 7 years ago

hm, after apt install python3-pip:

#pip3 install subuser
Requirement already satisfied (use --upgrade to upgrade): subuser in /usr/lib/python3/dist-packages
You are using pip version 8.1.1, however version 9.0.1 is available.
You should consider upgrading via the 'pip install --upgrade pip' command.

should I proceed as it suggests? (--upgrade)

timthelion commented 7 years ago

Aha, I guess they do conflict then. Well, the deb version doesn't work for you, so I guess you might as well go and uninstall it anyways, right?

On 07/28/2017 07:51 PM, mailinglists35 wrote:

hm, after |apt install python3-pip|:

|#pip3 install subuser Requirement already satisfied (use --upgrade to upgrade): subuser in /usr/lib/python3/dist-packages You are using pip version 8.1.1, however version 9.0.1 is available. You should consider upgrading via the 'pip install --upgrade pip' command. |

should I proceed as it suggests? (--upgrade)

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/subuser-security/subuser/issues/325#issuecomment-318720357, or mute the thread https://github.com/notifications/unsubscribe-auth/ABU7-GnU0ciwaZfampGY_IeyO_LkOK1jks5sSh-tgaJpZM4OX716.

mailinglists35 commented 7 years ago

uhm, the deb version actualy does work (although i see very high virt/res ram consumption of xpra server and client*). i was just worried about that message.

*

  PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
 2102 user     20   0 2138.2m 247.0m  49.1m S   0.0  1.6   9:53.51 firefox-esr
 2285 user     20   0 1747.3m 222.1m  54.5m S   0.0  1.4  14:34.90 Web Content
 1938 user     20   0 1158.0m  88.3m  21.5m S   0.0  0.6   1:01.95 xpra
 2030 user     20   0 1060.5m  54.1m  20.1m S   0.0  0.3   1:02.36 xpra
 1973 user     20   0  549.8m  28.2m   5.0m S   0.0  0.2   0:10.30 Xorg
 1064 root      20   0 1359.1m  26.1m  18.4m S   0.0  0.2   0:01.73 libvirtd
  925 root      20   0  633.4m  18.6m   5.2m S   0.7  0.1   1:07.25 docker
 1920 user     20   0   74.2m  16.2m   2.8m S   0.0  0.1   0:00.33 subuser
 2022 user     20   0  172.0m  14.7m   5.7m S   0.0  0.1   0:33.87 docker
 1930 user     20   0  236.0m  14.5m   5.6m S   0.0  0.1   0:34.10 docker
 1883 user     20   0   29.7m  12.1m   3.6m S   0.0  0.1   2:50.30 Xvnc4
18433 root      20   0   25.6m  11.8m   5.6m S   0.0  0.1   0:15.69 Xvnc4
 2034 user     20   0  180.0m  10.9m   4.1m S   0.0  0.1   0:35.04 docker
  305 root      20   0   34.5m   8.0m   7.6m S   0.0  0.1   0:01.61 systemd-journal
18274 root      20   0  186.3m   7.3m   6.4m S   0.0  0.0   0:00.02 virtlogd
43007 root      20   0   95.2m   7.2m   6.3m S   0.0  0.0   0:00.36 sshd
timthelion commented 7 years ago

The performance in general is my biggest gripe with subuser. I started the project with that happy go lucky "cpu, ram, and disk space is cheep" attitude. But it turns out that cpu time is MY time when I'm interacting with the computer. I really am not %100 sure what exactly is the cause of the various performance issues with xpra, but it seems to be some weirdness, because it works a lot better on my crappy old desktop than on my newer faster laptop... I keep hoping that the open sourceiness will lead a mad genius with too much time on her or his hands to fix all the problems for me ;).

On 07/28/2017 07:59 PM, mailinglists35 wrote:

uhm, the deb version actualy does work (although i see very high virt/res ram consumption of xpra server and client*). i was just worried about that message.

*

|PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 2102 user 20 0 2138.2m 247.0m 49.1m S 0.0 1.6 9:53.51 firefox-esr 2285 user 20 0 1747.3m 222.1m 54.5m S 0.0 1.4 14:34.90 Web Content 1938 user 20 0 1158.0m 88.3m 21.5m S 0.0 0.6 1:01.95 xpra 2030 user 20 0 1060.5m 54.1m 20.1m S 0.0 0.3 1:02.36 xpra 1973 user 20 0 549.8m 28.2m 5.0m S 0.0 0.2 0:10.30 Xorg 1064 root 20 0 1359.1m 26.1m 18.4m S 0.0 0.2 0:01.73 libvirtd 925 root 20 0 633.4m 18.6m 5.2m S 0.7 0.1 1:07.25 docker 1920 user 20 0 74.2m 16.2m 2.8m S 0.0 0.1 0:00.33 subuser 2022 user 20 0 172.0m 14.7m 5.7m S 0.0 0.1 0:33.87 docker 1930 user 20 0 236.0m 14.5m 5.6m S 0.0 0.1 0:34.10 docker 1883 user 20 0 29.7m 12.1m 3.6m S 0.0 0.1 2:50.30 Xvnc4 18433 root 20 0 25.6m 11.8m 5.6m S 0.0 0.1 0:15.69 Xvnc4 2034 user 20 0 180.0m 10.9m 4.1m S 0.0 0.1 0:35.04 docker 305 root 20 0 34.5m 8.0m 7.6m S 0.0 0.1 0:01.61 systemd-journal 18274 root 20 0 186.3m 7.3m 6.4m S 0.0 0.0 0:00.02 virtlogd 43007 root 20 0 95.2m 7.2m 6.3m S 0.0 0.0 0:00.36 sshd |

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/subuser-security/subuser/issues/325#issuecomment-318722273, or mute the thread https://github.com/notifications/unsubscribe-auth/ABU7-EhaxtHqGt3uWcia1vBXUCgI-L00ks5sSiGRgaJpZM4OX716.