workstation stability

hp z440     hp z840
before <     > after

 

I hate cables

pronóstico eclipse

 Latest forecast from Aragon:

sasp launchers

 

shell

#!/bin/bash
cd /home/steve/sasp-314
. venv/bin/activate
export QT_OPENGL=desktop
#QT_QPA_PLATFORM=xcb (if you want the old x server)
export QT_QPA_PLATFORM=wayland
python setiastrosuitepro.py

launcher

[Desktop Entry]
Comment[en_US]=sasp-314
Comment=sasp-314
Exec=/home/steve/runsasp-314 ; exec bash
GenericName[en_US]=sasp
GenericName=sasp
Icon=/home/steve/sasp-314/src/setiastro/images/astrosuite.png
MimeType=
Name[en_US]=sasp-314
Name=sasp-314
Path=
StartupNotify=true
Terminal=true
TerminalOptions=\s--noclose
Type=Application
X-KDE-SubstituteUID=false
X-KDE-Username=

graxpert gpu torch

Kubuntu 26.04 running siril. 
remove whatever is at:

~/.local/share/siril

open a terminal there. Then it's simply:

pip3.12 -m venv venv
. venv/bin/activate

Siril was complaining about not being able to load some CUDA 13 stuff, so from here:
 https://pytorch.org/get-started/locally/

choose your poison

Then it's simply:

pip3 install torch torchvision

Launch siril now, and we get hw accellerateion: 

02:06:46: Testing CUDAExecutionProvider...

02:06:46: OK: CUDAExecutionProvider ran successfully

02:06:46: Testing CPUExecutionProvider...

02:06:46: OK: CPUExecutionProvider ran successfully

02:06:46: === Summary ===

02:06:46: OK: Working providers (in priority order):

02:06:46:   - CUDAExecutionProvider

02:06:46:   - CPUExecutionProvider

02:06:46: → Best available provider: CUDAExecutionProvider

02:06:46: Cached execution providers to /home/steve/.config/siril/siril_onnx.conf

02:06:48: Using inference providers: ['CUDAExecutionProvider', 'CPUExecutionProvider']

02:07:03: Finished denoising

nvidia not upgrading

ubuntu 26.04 Not upgrading:

 libnvidia-cfg1-595 libnvidia-extra-595 nvidia-dkms-595-open libnvidia-common-595 libnvidia-fbc1-595 nvidia-driver-595-open libnvidia-compute-595 libnvidia-fbc1-595:i386 nvidia-kernel-common-595 libnvidia-compute-595:i386 libnvidia-gl-595 nvidia-kernel-source-595-open libnvidia-decode-595 libnvidia-gl-595:i386 nvidia-utils-595 libnvidia-decode-595:i386 linux-modules-nvidia-595-open-7.0.0-28-generic xserver-xorg-video-nvidia-595 libnvidia-encode-595 linux-modules-nvidia-595-open-generic libnvidia-encode-595:i386 nvidia-compute-utils-595

easy fix:

sudo apt dist-upgrade

nvidia purge

Because e.g. you're going to install a different gpu.

sudo apt purge 'nvidia-*' 'libnvidia-*'

 sudo apt autoremove --purge

Reboot afterward so the Nouveau driver is loaded again.

Shutdown, remove the power cable and hold the on button for a few seconds before removing and replacing hardware.

Then it's:

sudo ubuntu-drivers install

That's it.

tomates

planta madre

1 semana

2 semanas

3 semanas