[oi-dev] Mate is here (as well as new test ISOs)

Adam Števko adam.stevko at gmail.com
Tue Aug 16 16:09:49 UTC 2016


Hi,

> On Aug 16, 2016, at 6:07 PM, Udo Grabowski (IMK) <udo.grabowski at kit.edu> wrote:
> 
> On 16/08/2016 17:32, Alexander Pyhalov wrote:
>> On 08/16/16 06:22 PM, Udo Grabowski (IMK) wrote:
>>> On 16/08/2016 17:10, Alexander Pyhalov wrote:
>> 
>>>> We'd like to hear your opinion, which images are necessary for next
>>>> snapshot
>>>> (supposedly, 2016.10). We think that minimal one, Mate and Gnome one
>>>> are enough.
>>>> We are going to avoid delivering Gnome ISO image after next snapshot
>>>> (in 2017).
>>>> 
>>>> Also we are interested to hear if community needs VM images, and if it
>>>> needs,
>>>> then what exactly (qcow2, vmdk or something else).
>>> 
>>> Good work ! But under qemu-kvm, the MATE iso fails by starting Caja
>>> in an endless loop (visible in the lower panel), so the desktop remains
>>> mostly black, and Mate does not stop starting...
>> 
>> Can you start with ssh enabled, login as jack remotely and see if there are any
>> messages in ~/.xsession-errors?
> 
> ssh connection refused (sshd is running) - either misconfigured or not starting
> correctly; if jack has an empty password, the sshd configuration
> denies logins with empty passords. But I can get a terminal after
> the Caja processes seemingly all died, the desktop still black,
> no border decoration, the terminal constantly moving to the lower
> right...after patient double typing  get these .xsession-errors:
> -----------
> mate-settings-daemon MateDesktop-Warning: Call to screen_info_new
> is too frequent, skipping...
> caps.c Normally all extra capabilities would be dropped down, but
> that's impossible because PulseAudio was built without capabilities
> support.
> /usr/bin/dispswitch exiting: too few crtcs: 1
> pygobject_register_sinkfunc is  deprecated (GtkWindow,GtkInvisible,Gtkobject)
> mate-panel: couldn't get background pixmap
> Connection failure: Connection refused
> Initializing gksu extension
> The program caja received an X Window System error.
> This probably reflects a bug in the program.
> The error was 'BadLength (poly request too large or internal Xlib
> lenght error (Details: serial 262 error_code 16 request_code 72
> minor_code 0)
> ---------------
> This goes on for a while, then the caja clients are killed:
> --------------
> Maximum number of clients reached
> caja: MateDesktop-Warning: Unable to open display :0 when setting
> background pixmap
> caja: Gdk-critical: IA__gdk_cairo_create asserttion 'GDK_IS_DRAWABLE
> (drawable) failed.
> recvmsg: Connection refused
> --------------------
> a couple of those messages, then
> ---------------------
> process 3535 (one PID beyond the caja PID): arguments to dbus_message_unref()
> were incorrect, assertion "!message->in_cache" failed in file
> /jenkins/jobs/oi_userland/workspace/components/library/dbus/dbus-1.10.8/dbus/dbus-message.c line 1719.
> This is normally a bug in some application using the D_Bus Llibrary.
> D-Bus not built with -rdynamic so unable to print a backtrace.
> -------------------------
> and more of the caja background pixmap and cairo IA__... messages.
> Then a heap of Xlib messages related to the caja breakdown:
> ---------------------------
> Bug in window manager: Unexpected X error: BadLength (poly request too
> large or internal Xlib
> lenght error (Details: serial 815 error_code 16 request_code 72
> minor_code 0)
> --------------
> the same messages repeating endlessly with other serials like
> 832 and 835. The terminal flickers endlessly while bumping into
> the lower right corner like a dumb zombie and is nearly unusable.
> 
> Hope that helps.

I will see if I can reproduce this. Just to clarify: are you running this on illumos qemu-kvm?

Cheers,
Adam

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 801 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://openindiana.org/pipermail/oi-dev/attachments/20160816/fe8258ac/attachment-0005.bin>


More information about the oi-dev mailing list