Discussion thread for 42.2 Release Candidate 1 (RC1)

Leap 42.2 Release Candidate 1 (RC1) was released today <https://news.opensuse.org/2016/10/18/release-candidate-available-for-opensuse-leap-42-2/>.

This thread is for discussion.
[HR][/HR]13.2 and Leap 42.1 running on 64-bit AMD 4-Core CPUs with AMD Graphics
Leap 42.2 running in Oracle VirtualBox VMs

I have done the 13.2 → 42.2 beta 3 → 42.2 RC1 upgrade on my test VB install.
I haven’t encountered any problem on the second part. I find RC1 absolutely stable for the set of applications/packages I use.

The only stopper to me to do right now the 13.2 → 42.2 RC1 upgrade on my primary hardware is the packman repo. All other repos I use have their 42.2 version already in place, except packman.

I have just completed my second install.

Installs mostly went smoothly.

Here are the “issues” that I ran into:

Cosmetic:

There were places during the install where it was still saying “Beta”. That should not be there on a release candidate (probably an oversight).

During the install, it downloaded “kaddressbook5”. I suspect that was a mistake.

In more detail, this was after the software install, when it was setting up the newly installed system. The last package actually installed before the finishing steps was “kaddressbook5”. But the later message about downloading “kaddressbook5” is probably bogus. I’m guessing that it was re-using a message buffer that wasn’t properly updated. It was probably downloading something to test the network. Traditionally, this has been the release-notes. It seems unlikely that it would download “kaddressbook5” for that. I’ve seen this also in earlier Beta releases. I wasn’t paying enough attention with Alpha releases.

So those are the cosmetic issues. I haven’t decided on whether to file a bug report.

Functional:

One functional issue. My first install was to a box with Nvidia graphics, using the nouveau driver. Plasma 5 does not much like nouveau. So it is not much of a surprise that it wasn’t perfect.

I set LIBGL_ALWAYS_SOFTWARE=1 in the environment before starting plasma 5. I actually did this with a line in the shell startup file.

Plasma 5 came up nicely, without any problems that I could see.

So I removed the environment definition (commented out) from shell startup file, and logged out and then logged in again. And plasmashell apparently crashed leaving a crippled session and a coredump process. I aborted the session (with CTRL-ALT-BACKSPACE twice), then added back that environment definition. And it has been fine that way.

I have not yet come across any other functional issues.

I should add that both installs were to encrypted LVM. That worked well. For the second of those, I did not use a separate “/boot”, so I had to give grub the encryption key. That also worked well. So all looks good for now, apart from the issues mentioned above.

upgrade my machine today from 42.1
now, Network Manager not showing available wifi
also sometimes it crash, i only open chromium when it happen


Fujitsu LH530 with KDE Plasma

reconfigure Network Settings fixed Network Manager Problem
let’s see with the crash


Fujitsu LH530 with KDE Plasma

Question.

Did the checkbox "Auomatic Configuration return in (one of the first screens of) the installer?

I don’t think so.

I’ve done 3 installs, and I don’t recall seeing that. As far as I know, it is gone for good.

Terrible >:)

I know I can revert to my own configuration wishes for at least the NIC configuration by pulling the ethernet cable, but not forgetting to pull and replug the cable during each and every installation in time frustrated me severely until the point that I stopped installing 42.1 out of fear that would start kicking the machine (no joke).

I’ve tried pulling the cable. And that works. But, like you, I usually forget.

It would be nice if we could get that network config screen back without having to pull the cable.

I am also not sure what the situation is with a Wifi NIC. You can not pull the cable there :(.

And I am not sure if the “not automatic configuration” included more then the network configuration. It could mean that there are things configured where I only will find out when it is too late (or maybe never), but that still consider harm my policy as an admin or even the security of my systems.

I don’t know what happens with an open network, since I don’t have one of those. But, otherwise, it has to go to the network screen for you to configure the network key.

As for other configuration - I think there used to be a printer section. But the network part was the only one where I remember having to do anything interactive.

I sometimes consider pulling the ethernet cable for a different reason. Without a cable, it gets the time setting correct. If I have the ethernet cable connected, it gets it wrong. On a computer with Windows, it first decides that the clock should use local time. Then it use “ntp”, and sets the internal clock. I check the box to use UTC, but it still sets the internal clock 5 hours off (the difference between Eastern time and UTC). It configures the computer correctly, but that depends on NTP during the first boot fixing the internal clock problem. If I disconnect the ethernet cable, then it cannot do NTP, and it does not screw up the internal clock time.

I will try the RC later on, the beta 3 was okay but does this do anything better?

I did a clean Leap 42.2 install on both my Viper ufei desktop and HP 17T ufei laptop; both went exceedingly well & yes, it did say Beta on the 1st install page. This was a win10, Leap 42.2 & Tumbleweed multiboot install on both.

I was expecting a video problem on my desktop with my 4-monitor display and the Noveau driver, but found that it drives my 4 screens good enough to suffice until the Nvidia driver is provided.

There are, of course, some nitnoids, i.e., I did away with the desktop folder and placed the Home, Trash and Links to applications (that I made) on the main display, however they appear on all my screens and the Links to applications say just that, not the name I assigned.

I have used various multimedia repos and have vlc working. but kaffeine and xine crashes, but that will go away when Packman gives us the 42.2 repro.

My plan is to play with the install untill GM is released and then do a GM clean install.

I was not a Leap 42.1 fan and have been using 13.2 as my production OS; I will move it all to Leap 42.2 at GM.

Have fun!

In addendum, I will have to get Packman’s BroadcomWL src rpm and rebuild it to get my HP 17T laptop wireless working.

info
On my desktop, i used the PackmanTumbleweed repo and kaffeine and xine work but VLC does not start; on my laptop, I used the sle12 repo and VLC worked and kaffeine and xine crashed.

One can pick and choose for the moment.

I just upgraded from 42.1 to beta 1 and then upgraded from there as the releases came out. The only real problem I had with the betas was wifi. I had to reset it on every boot. Since rc1 it’s been working as it should.
I will continue to just upgrade.

Well, last night it was working very well, I must have updated/added something it didn’t like because now it boots to a set of blank screens. I tried nrickert’s setting LIBGL_ALWAYS_SOFTWARE=1 in the environment (post #3), but it didn’t work for me. Also re-installed nouveau files.

When I have time, I will install the nvidia driver the hard way, but, it will be after the weekend.

I did find that if I boot to run level 3 and login as root and “startx” I can get an X session with all the tools – neat!

Had this “boot to black screen” also here with an old Nvidia card (no NVIDIA driver installed) after updating from Beta to RC1 yesterday. Today booted again, waited a little longer and had my GUI back again!

But I have in dmesg:

   23.480090] SFW2-INext-DROP-DEFLT IN=eth0 OUT= MAC= SRC=fe80:0000:0000:0000:020f:1fff:fe90:ae58 DST=ff02:0000:0000:0000:0000:0000:0000:00fb LEN=319 TC=0 HOPLIMIT=255 FLOWLBL=482250 PROTO=UDP SPT=5353 DPT=5353 LEN=279 
   23.516453] SFW2-INext-DROP-DEFLT IN=eth0 OUT= MAC= SRC=fe80:0000:0000:0000:020f:1fff:fe90:ae58 DST=ff02:0000:0000:0000:0000:0000:0000:00fb LEN=184 TC=0 HOPLIMIT=255 FLOWLBL=482250 PROTO=UDP SPT=5353 DPT=5353 LEN=144 
   23.730303] SFW2-INext-DROP-DEFLT IN=eth0 OUT= MAC= SRC=fe80:0000:0000:0000:020f:1fff:fe90:ae58 DST=ff02:0000:0000:0000:0000:0000:0000:00fb LEN=319 TC=0 HOPLIMIT=255 FLOWLBL=482250 PROTO=UDP SPT=5353 DPT=5353 LEN=279 
   23.981392] SFW2-INext-DROP-DEFLT IN=eth0 OUT= MAC= SRC=fe80:0000:0000:0000:020f:1fff:fe90:ae58 DST=ff02:0000:0000:0000:0000:0000:0000:00fb LEN=319 TC=0 HOPLIMIT=255 FLOWLBL=482250 PROTO=UDP SPT=5353 DPT=5353 LEN=279 
   24.181614] SFW2-INext-DROP-DEFLT IN=eth0 OUT= MAC= SRC=fe80:0000:0000:0000:020f:1fff:fe90:ae58 DST=ff02:0000:0000:0000:0000:0000:0000:00fb LEN=295 TC=0 HOPLIMIT=255 FLOWLBL=482250 PROTO=UDP SPT=5353 DPT=5353 LEN=255 
   55.340496] SFW2-INext-DROP-DEFLT IN=eth0 OUT= MAC=33:33:00:00:00:fb:b8:27:eb:9b:23:d1:XX:dd SRC=fe80:0000:0000:0000:23ce:03f6:0c1e:YYYY DST=ff02:0000:0000:0000:0000:0000:0000:00fb LEN=115 TC=0 HOPLIMIT=255 FLOWLBL=0 PROTO=UDP SPT=5353 DPT=5353 LEN=75 
   87.369365] SFW2-INext-DROP-DEFLT IN=eth0 OUT= MAC=33:33:00:00:00:fb:b8:27:eb:9b:23:d1:XX:dd SRC=fe80:0000:0000:0000:23ce:03f6:0c1e:YYYY DST=ff02:0000:0000:0000:0000:0000:0000:00fb LEN=115 TC=0 HOPLIMIT=255 FLOWLBL=0 PROTO=UDP SPT=5353 DPT=5353 LEN=75 
   87.377675] SFW2-INext-DROP-DEFLT IN=eth0 OUT= MAC=33:33:00:00:00:fb:b8:27:eb:19:2f:1a:XX:dd SRC=fe80:0000:0000:0000:d42f:f02b:f0b9:YYYY DST=ff02:0000:0000:0000:0000:0000:0000:00fb LEN=115 TC=0 HOPLIMIT=255 FLOWLBL=0 PROTO=UDP SPT=5353 DPT=5353 LEN=75

Kind of IPv6 spam? Although in Network Settings IPv6 is DISABLED from beginning and the Network card is configured for IPv4 DHCP (no IPv6). Can this be stopped?

Has anyone taken a look at the following LTS Kernel 4.4 feature?

  • Support for NFSv4.2 file CLONE using the btrfs ioct.

[HR][/HR]I’ll set the VMs I use for testing here to “Network Bridge” mode, and see if it works here with the NFS Servers I have available.

I did a clean install (no data from previous versions; / and /home both formatted to ext4) of openSUSE 42.2 RC1 without any problems.

The system works fine EXCEPT that i can’t use my scanner (Canon CanoScan LiDE 700F) any longer (see https://bugzilla.opensuse.org/show_bug.cgi?id=1005673 for more details).

Any suggestions how i could investigate the source of the problem (libusb ?) in more detail?

Regards

susejunky