betacentauri

You can try this, but I think you need to use twisted Agent to workaround it

import ssl
try:
  _create_unverified_https_context = ssl._create_unverified_context
except AttributeError:
  pass
else:
  ssl._create_default_https_context = _create_unverified_https_context

https://forums.openp...dpost&p=1206862

I’m not sure whether it is already fixed. But we have to be sure that it is. Otherwise users will “kill” us.

You can also try this
https://github.com/E...ugin/ui.py#L364
I mean the self.onShown.append(...)

Mixing (one file spaces, other tabs) I don't really prefer. Maybe only if we want to move slowly to spaces. (Mixing in one file is bad and shouldn't be done).

 

I really prefer tabs. I don't understand why spaces are the standard. Doesn't make sense for me. For me as long as tabs are allowed we should use them.

You can always use tab to complete the commands/filenames.
So you can say /media/hdd/202 <tab>
then you see most likely
/media/hdd/2020
With another tab all files matching 2020 should be displayed. Then enter eg 0422 <tab> and you should see the files for the 20200422. I hope you understand what I mean.

Please try this ddrescue. This time self build. It's a static binary, so no libs are needed. It's quite large as it's static and not stripped.

At the moment there is no public image available afaik.

Or see here:

https://bbpress.org/...modify-mo-file/

It's the problem that the virtual keyboard disappears for a tens of a second when you move the cursor to another character. You can see it only on the TV and not via webif.

Afaik blindscan issue is now fixed in RC.

From your PM I see:

Found mountpoint for rootfs file: /mnt/usb/sda1

The empty cmdline is also not ok and not normal.

Both imply that some kind of multiboot mechanism was installed by the user. I heard that that was available for solo4k, but I never used or really saw it.

 

If the user has detached the USB stick / HDD(/mnt/usb/sda1), he should attach it again. If root is still present on that stick/hdd, box should work properly again.

 

If that's not possible, please post output of "mount". Is the /dev/mmcblk0p4 used for rootfs?

Please don't do it yet. My idea would be to flash a proper kernel (from OpenPLi). The kernel will use /dev/mmcblk0p4 as rootfs. It works if really a rootfs is available on /dev/mmcblk0p4. If not, we might have a bigger problem than before. So better wait and think twice before bricking the box.

But maybe with proper kernel, USB flashing works again....

Du könntest mal probieren ob ein Neustart einmal am Tag das Problem behebt. Wenn ja, geht der Box vielleicht der Speicher(RAM) aus. Ist aber nur geraten.

On a quad4k you can watch the recording.

I don’t think the “ERROR reading PES” errors are related. I already saw these messages without a crash.
The crash seems to be in libavcodec, but it can be that the calling part already delivers bad data. So this backtrace doesn’t really help.
If you have the knowledge disable the signal handler in e2. Then a core dump is created during next crash. And this core can be analyzed with gdb.
Or add tons of debug messages to find out where it crashes.