Following a request to install a new font in a Moodle instance, I did a local install with
sudo mkdir -p /user/share/fonts/googlefonts
cd /usr/share/fonts/googlefonts
sudo unzip -d . ~/Tangerine.zip
sudo fc-cache -fv
Checking to see how critical are prototype pollution vulnerabilities for front-end code, apparently they are:
One of the many sites which list Mongodb vs Postgres comparisons,
https://www.knowledgenile.com/
Transaction-safety and joins etc are what mongodb doesn't offer, while sharding is its strength.
There was one drive which was being recognized, and another drive which was not. The drive which was being recognized had the original 16 GB FAT32 partition, while the one which was not being recognized had been used as a Linux Mint installation disk, and had been later reformatted as FAT32 using the "Disks" utility on Linux Mint.
Suspecting that block size might be the issue, to show the block size,
sudo mkfs.fat /dev/(device name from above) -s 64 -F 16Then, tried GParted. GParted showed the partition to be of type iso9660!
Unmounted, Deleted the partition, recreated partition table with device -> create partition table (msdos), formatted as msdos, then the Firestick recognizes it OK even though the block size is 8K as shown in
stat -f /media/mac/LM
File: "/media/mac/LM"
ID: 83100000000 Namelen: 1530 Type: msdos
Block size: 8192 Fundamental block size: 8192
Blocks: Total: 1951116 Free: 1318805 Available: 1318805
Inodes: Total: 0 Free: 0
A quick 20 minute video walk-through of Blender 3.0
https://www.youtube.com/watch?v=Rqhtw7dg6Wk
Similar to https://hnsws.blogspot.com/2011/10/getting-started-with-blender.html
Create a Release and Upload Artifacts with GitHub Actions | Thomas Stringer (trstringer.com)
Just add the following, and it will create a zip file with the path -
- name: Upload Artifact
uses: actions/upload-artifact@v2
with:
name: artifact-of-build
path: platforms/android/app/build/outputs/
OCVWarp deployment from Travis-CI to Github Releases failed because the token had expired. When creating a new token, I tried a fine-grained token with write permissions to releases on that repo alone. But that did not seem to work. A "classic" personal token with "repo" scope worked fine.
An interesting video and project via Computerphile on youtube -
SelfDriving/DataCapture.py at main · apt503/SelfDriving · GitHub
Of course, this is just a side project, but the approach is interesting.
How to migrate your WhatsApp data from Android to iPhone | WhatsApp Help Center
Though this method was followed, only the photos and media were transferred from whatsapp, not the chat history. Probably that was because whatsapp had been earlier installed on the iphone, though it was then removed with "delete all data"?
https://faq.whatsapp.com/1195710084347483?helpref=faq_content
then, to transfer music, move content manually from your Android device to your iPhone,
While setting up the Firestick, we wanted to create a new dedicated account for it, but it asked for a unique phone number - so we just used my account.
We found VLC player was stuttering with our 4K videos on USB, but MXPlayer works with hardware acceleration, and plays them perfectly. To access MXPlayer without internet, via Quora,
Settings --> “Applications” --> “Manage installed applications” --> "MXPlayer" --> "Launch"
And it looks like there is no "shutdown" option - we just press "home" to come out of MXPlayer, and then turn off power to the device.
Our Firestick runs FireOS 6.4, but we did not have to do any ADB commands to use FAT32 formatted USB memory. The ADB commands in this pdf may be useful for increasing system usable memory, installing apps and so on, but not needed for us.
As recommended in this troubleshooting video, when connecting for the first time, doing so from the "home" screen might be beneficial.
How to play our own content from USB on Amazon Firestick
We can't just connect the firestick to a computer and transfer files to its internal memory - MTP does not work with Firestick 4k -
Investigating unresponsive server behaviour on one of our servers,
grep -i error /var/log/syslog
showed some lines with WALinuxAgent - this is Azure's agent, I guess, as well as snapd.
snapd[63805]: stateengine.go:149: state ensure error: cannot query the store for updates
To check if we can remove snapd
How to remove LXD from my system - LXD - Linux Containers Forum
So I left it alone, and tried a manual update with
sudo snap refresh
which indicated all packages were up to date. So maybe the error was a temporary error.
One of our Linux servers was becoming unresponsive around 2-5 pm daily over 2-3 days. Checked the logs as in my earlier post but did not find anything except the usual SSH brute-forcing attacks. Changed the port from 22 to something else by
sudo nano /etc/ssh/sshd_config
sudo service sshd restart
And opened the new port on Azure portal Networking screen.
This seems to have worked.
The smart lamp controls with a python flask app as in a previous post had some issues.
So, thought of redesigning the code.
Found that RaspController has a section for running user-defined custom commands. Since both KR and I use RaspController on our phones, the simplest solution seemed to be to just use a python script to do the fadeout of the lights, and show the output of the script on the terminal using RaspController.
And that seems to be working fine.
Since the code on github has the device ids and local keys, instead of taking the time to modify the code to put the private info in separate files etc, I'll just copy-paste the current state of the code here.
import tinytuya, sys
from time import sleep
def moveup(lines):
for _ in range(lines):
sys.stdout.write("\x1b[A")
nowait = bool(1);
# Connect to Device
d = tinytuya.BulbDevice(
dev_id='changeme',
address='our.ip.2', # Or set to 'Auto' to auto-discover IP address
local_key="changeme12",
version=3.5)
print("Connected to d")
b = tinytuya.BulbDevice(
dev_id='deviceidchange',
address='our.ip.1', # Or set to 'Auto' to auto-discover IP address, but that takes 10 sec
local_key="changeme34",
version=3.4)
print("Connected to b")
def fadeout():
d.set_brightness_percentage(100, nowait)
sleep(1)
b.set_brightness_percentage(100, nowait)
sleep(1)
d.turn_on()
print("d turn on")
sleep(1)
b.turn_on()
print("b turn on")
sleep(1)
i = 100
while i > 10:
d.set_brightness_percentage(i, nowait)
print("d brightness {0:3d}".format(i))
b.set_brightness_percentage(i, nowait)
print("b brightness {0:3d}".format(i), end='\r')
moveup(1)
sleep(1)
i -= 10
i = 10
while i > 11:
d.set_brightness_percentage(i, nowait)
print("d brightness {0:3d}".format(i))
b.set_brightness_percentage(i, nowait)
print("b brightness {0:3d}".format(i), end='\r')
moveup(1)
i -= 1
d.turn_off()
print("d turned off. ")
b.turn_off()
print("b turned off. ")
fadeout()
There was an issue once again with the feedback ad-hoc Moodle report - checking the query in DBeaver, the database returns an error "Subquery returns more than one row" for some dates. Eg Oct 1.
Via https://stackoverflow.com/questions/2594829/finding-duplicate-values-in-a-sql-table
Found the duplicate timestamps with
select timemodified, count(*) from ourprefix_feedback_completed
GROUP BY timemodified
HAVING COUNT(*) > 1
1690263769 2
1696119639 5
Have changed the timestamps to increment the duplicates by one, so that the query works without issues.
I tried out some rendering of scenes using the Cycles Panorama camera 180 degree Fisheye Equidistant option. A large number of links to howtos for accomplishing specific tasks in blender - link dump below.
1. Render optimizations - 4K render of one frame takes 6:45 - nearly 7 minutes a frame is not enough - need to check out GPU rendering. GPU needs min compute capability 3 Nvidia or AMD GPU -I'd installed loggly free trial on one of our servers, and had forgotten about it. Installation documentation
https://documentation.solarwinds.com/en/success_center/loggly/content/admin/sending-apache-logs.htm
When it came to uninstalling it, Solarwinds support said just remove the API token. But then the server would continue to send the data to solarwinds' server, only we can't see it, right? Not very customer-friendly. Anyway, their own documentation says
Stop Sending Data to Loggly (solarwinds.com)
sudo bash ./configure-linux.sh -a our.domain.org -r
https://stackoverflow.com/questions/4897359/output-to-the-same-line-overwriting-previous-output
just add , end='\r' to the print()
So, something like the code in this private repo,
https://github.com/hn-88/python-tuya-fadein-web/
while i > 10:
d.set_brightness_percentage(i, nowait)
print("d brightness {0:3d}".format(i))
b.set_brightness_percentage(i, nowait)
print("b brightness {0:3d}".format(i), end='\r')
moveup(1)
sleep(1)
i -= 10
Earlier, I had downloaded some youtube 360 VR files, which were in equirectangular format, converted to either fulldome or pre-warped mirrordome using OCVWarp, and projected on our dome. For example
360° Video of Vidhan Soudha, Bengaluru During On Going Lock-Down
But (recently?) when I tried out 360 VR 4K videos like this one
Tim Peake ISS 360° Planetarium Tour
the format was different - not equirectangular. While I thought about exploring the format to find out some workaround, found that Paul Bourke has already done it in Jan 2020.
Apparently it's 6 faces of the conventional cube map, but resized a bit. Ordered as left-front-right
bottom-back-top
And he mentions the ffmpeg command to convert it to equirectangular, which I tested and seemed to work,
ffmpeg \ -y \ -hide_banner \ -i input.mkv \ -vf "v360=c3x2:e:cubic:in_forder='lfrdbu':in_frot='000313',scale=3840:1920,setsar=1:1" \ -pix_fmt yuv420p -c:v libx264 -preset faster -crf 21 \ -c:a copy -ss 16 -t 10 -movflags +faststart \ output.mp4