blender fisheye

Step by step guide to set up fisheye camera output in blender - 

https://artisticrender.com/how-to-create-a-fisheye-lens-for-the-camera-in-blender/

using Cycles instead of Eevee.

 

1. Render properties - set the render engine to cycles.

2. Select the active camera -> data properties -> lens -> set to panoramic.

3. Panorama type -> Fisheye Equisolid.

4. Adjust the Field of view FOV to 180 etc.

firefox - proper indic fonts for google docs

https://www.google.com/search?q=firefox+default+font+for+indic+change

https://kohageek.blogspot.com/2018/04/configure-firefox-browser-to-view-unicode-fonts.html

In my case, Firefox -> Hamburger menu -> Settings -> (Search for Font) -> Advanced

set separately for Devanagari, Telugu, etc.

 

 

Admin console of Google Workspace - mini QA

Copy-pasting from my responses to some queries about admin.google.com - 

1. Two of us have admin access for user actions. But both of our interfaces are different. Kindly check and provide only the role that is required to perform the above roles.  

If you go to admin.google.com (Desktop interface) -> (LHS panel) Account -> Admin roles and click on Super Admin, you will find that one is a Super Admin while the other is a User-management Admin. That is the difference. Is the user management admin role working fine for <username>? If yes, you can add <theotherusername> to user management admin role, and then remove <theotherusername> from super admin role.

2.  Around a month or so ago, I noticed that my drive shows the quota in the following manner.  Is it normal? If I have changed some settings, kindly help in retrieve it back to normal. 

My guess is that you are looking at a shared drive, or at the paid google workspace login.

 OK - I found that Drive now shows this "1.13 TB used out of 100 TB" for all Super Admins, which is the total usage by all users.

3. Can the users reset their password themselves, just like normal Gmail?  If yes, kindly change the settings accordingly. If not, we'll continue to reset it manually. 

Account Recovery - Admin Console (google.com)

Currently this setting is set so that users can reset their own passwords. But they should know how to do it:) 

Google Play Console - even test tracks need to be updated with latest API level

We got an email 4 days back, "We've detected that 1 of your apps is targeting an old version of Android. To provide users with a safe and secure experience, Google Play requires all apps to meet target API level requirements before Aug 31, 2023." 

Apparently this is due to old apks in other (testing) tracks, not the Production release track.

Went to Policy Status page via the link,

Clicked on the right-arrow for view details, found a link in Details,

"View affected app bundles" which listed the four old closed testing tracks, two of which were even paused. 

 

So, since it doesn't seem possible to delete any tracks, updated all of them to the latest release.

https://stackoverflow.com/questions/51300238/delete-a-closed-track-in-google-play-console

 

Even after the updates were accepted after review, the banner saying "Policy violation" etc was still being shown on the console. Then I sent an appeal - which is via this page, 

https://support.google.com/googleplay/android-developer/answer/10357403?hl=en

for which if you are logged on to the developer account, some of the fields are auto-populated. I came to this page via

play.google.com/console
--> Dashboard
--> ? icon,
--> entered "api target  level" and hit next
then went to Get Developer Support.
to access the contact form for appeal.
Appeal the decision here link.

After the appeal reached the concerned team, the warning on our developer console vanished, and they also confirmed that the only way is to update all the tracks, tracks can't be deleted.

 

using google vault

vault.google.com allows Google Workspace users to have an archive for their emails (and other products like Google Drive also, optionally). Helping CV to set it up, this video offers a good walk-through. For every e-discovery request, we need to create a new "Matter" with search terms etc. The reports generated via the "Matter" can be exported as an mbox file with full contents, or metadata only, etc.

Instead of Thunderbird, which seems to be able to import mbox files through ImportExportNG addon which is only supported up to version 102 and not for the current version 115, standalone mbox viewers are also available, just to view/print/save attachments from mbox files.

Examples

Get MBOX viewer - a true freeware with transparent interface! (coolutils.com)

https://github.com/eneam/mboxviewer

etc

 

Azure portal VM start-stop permissions for guest user

Revisiting my older post on the subject:

I had created that custom role inside the particular VM I wanted to delegate access for. So, the custom role "VM-start-stop" or whatever was not available for another VM I had created in the same resource group. The point to note is that the role should be created with the desired scope - so, 

• if I navigate to the resource group --> IAM --> Add +  --> Add custom role, then that custom role would be available throughout that resource group. 
• if I navigate to Home --> Subscriptions --> (a particular subscription) --> IAM --> Add +  --> Add custom role, then that custom role would be available throughout that subscription.
• if I navigate to the VM and go to IAM --> Add + , I don't see the option to add a custom role.

https://stackoverflow.com/questions/23668154/allow-users-to-start-stop-particular-azure-vms

According to the stackoverflow post above, "Virtual Machine Contributor" role gives more permissions than we want to delegate. So, after trying out this role (which works for start/stop VMs, but also allows some more permissions), decided to recreate the permissions given in my earlier post. So, copy-pasting from the JSON tab of the previously created role, only the actions and dataActions parts in permissions,

"permissions": [

            {

                "actions": [

                    "Microsoft.Network/publicIPAddresses/read",

                    "Microsoft.Network/virtualNetworks/read",

                    "Microsoft.Network/loadBalancers/read",

                    "Microsoft.Network/networkInterfaces/read",

                    "Microsoft.Compute/virtualMachines/*/read",

                    "Microsoft.HybridCompute/machines/*/read",

                    "Microsoft.HybridConnectivity/endpoints/listCredentials/action",

                    "Microsoft.Compute/virtualMachines/start/action",

                    "Microsoft.Compute/virtualMachines/powerOff/action",

                    "Microsoft.Compute/virtualMachines/restart/action",

                    "Microsoft.Compute/virtualMachines/deallocate/action"

                ],

                "notActions": [],

                "dataActions": [

                    "Microsoft.Compute/virtualMachines/login/action",

                    "Microsoft.HybridCompute/machines/login/action"

                ],

                "notDataActions": []

            }

to a newly created custom role in the Resource Group, (we have to click Edit in the JSON tab, and then Save after entering the permissions).

Then, to see it under our VM, had to go to roles tab, choose the custom role filter, and then refresh.

Later, under Role Assignments, selected the VM contributor rule, and removed that role since it was too permissive.

setting timezone to Indian Standard Time on Ubuntu 22.04

Probably similar for other systems running systemd, I guess.
sudo timedatectl set-timezone Region/City,

so

sudo timedatectl set-timezone Asia/Kolkata

Sony Trinitron Micro Latch-up

The error-code in the service manual of the Sony Trinitron TV used in the waiting area, model KVXJ29M50, indicated "Micro Latch-up". Apparently that means that some CMOS component is damaged?

https://eng.libretexts.org/Bookshelves/Electrical_Engineering/Electronics/Introduction_to_Physical_Electronics_(Wilson)/03%3A_FETs/3.12%3A_Electrostatic_Discharge_and_Latch-Up

Currently not digging further.

Edit 16 Aug - saw some waveforms of "board C6" on oscilloscope, they seemed to be OK - at least not flat, even for R, G and B, even though R is not coming on screen. Perhaps damage inside the picture tube? Or on the board C6? Anyway, not digging any further, just leaving it, as it works as a "monochrome" or "duo-chrome" monitor at present. 

some notes on using supervisord to start a flask web app on Raspberry Pi or Ubuntu-based systems

Using flask on Raspberry Pi - https://towardsdatascience.com/python-webserver-with-flask-and-raspberry-pi-398423cc6f5d

One of the methods suggested to start a python flask web app on boot was to use supervisor. Since I installed it with apt, the suggested method to create a configuration file did not work.

Also, it is apt install supervisor, and not apt install supervisord.

https://www.google.com/search?client=firefox-b-lm&q=supervisorctl%3A+command+not+found

sudo apt install supervisor

https://github.com/Supervisor/supervisor/issues/376

https://forums.raspberrypi.com/viewtopic.php?t=126950

looks like /etc has a supervisor directory and a conf.d directory inside it!

sudo service supervisor status

not supervisord

Also, supervisor seems to search for and use supervisord.conf files, so we must ensure wrong or backup files are renamed to something else. 

INFO exited: flask_app (exit status 1; not expected

This was apparently due to not using the user=ouruser in the config file.

Final configuration was just by appending the following to the existing /etc/supervisor/supervisord.conf file.

[program:flask_app]
user = ouruser
command = /usr/bin/python ourapp.py
directory = /home/our/path/to/WebServer
autostart = true
autorestart = true

Then, to add a favicon, we need to add a route:

https://stackoverflow.com/questions/40696745/adding-a-favicon-to-a-flask-server-without-html

Authentication, or running on android - 

https://www.digitalocean.com/community/tutorials/how-to-add-authentication-to-your-app-with-flask-login

Flask on android | How to run flask server on an android | python web development | we make geeks - YouTube

https://www.youtube.com/watch?v=eDlauNGVYf4

UFW and limit ip address would be easier.

https://pimylifeup.com/configuring-ufw/

(Had to install python-is-python3 on one of my systems for using python nameof script.py instead of python3 nameofscript.py)

 

Edit: problems with this approach, and alternate solution in another post.

notes about controlling "smart lamps"

The final solution was something like this,
https://github.com/hn-88/python-tuya-fadein-web

Connecting to the web app running on our internal network using our mobile browsers to have local controls. 

Making the flask python app start on booting the "server" - using supervisor. I'll write a separate post on configuring supervisor on Raspberry Pi (in general, Ubuntu-based systems).

Link dump:

https://www.makeuseof.com/build-diy-alexa-smart-switches-home-automation/

Add Dimming or Brightening Bulbs to a Routine - Amazon Customer Service

https://github.com/beville/flux_led

flux_led shows no devices. And there is no ip address for the wipro smart batten 

(Actually it did have an ip address, our router was not displaying the client list properly - it would show only those clients which have recently requested ip leases.)

https://www.makeuseof.com/tag/control-philips-hue-lights-arduino-and-motion-sensor/

Apparently philips hue uses a bridge, which does have an ip address.

on alexa app, info by long press is "Connected via Smart Life"

https://www.google.com/search?q=smart+life+lamp+bulb+control+with+python+

https://ardha.dev/blog/controlling-tuya-smart-bulb-using-python

https://stackoverflow.com/questions/73062266/stuck-at-1004-sign-invalid-from-tuya-api

Local Control of Tuya Devices // No Cloud Needed // LocalTuya - YouTube

TradeFace/tuyaface: Python Tuya client that allows you to locally communicate with tuya devices (github.com)

from

https://github.com/rospogrigio/localtuya

which is mentioned in above video.

Will try the nmap solution at

https://www.reddit.com/r/homeassistant/comments/nvtwh3/how_to_identify_tuya_device_on_network/?rdt=40848

nmap -sU -p 137,5353 --script nbstat,dns-service-discovery <<your target range here>>

Install localtuya custom integration and when trying to add devices,
it will list found ones with their ip addresses and IDs.

sean6541/tuyaapi (github.com) - but this seems to be 5 years old, may not work.

https://www.quora.com/What-protocols-are-used-by-Tuya-smart-devices

->

https://github.com/codetheweb/tuyapi

->

https://github.com/clach04/python-tuya

->

https://github.com/jasonacox/tinytuya

Trying tinytuya.

https://github.com/rathmerdominik/tuyactl/blob/master/tuyactl

also uses sleep.

But in testing, even if the set_brightness_percentage() calls are not separated by sleep calls, it still works synchronously, taking more than half a second per call.

Exploit for jailbreaking tuya devices (not used)

https://rb9.nl/posts/2022-03-29-light-jailbreaking-exploiting-tuya-iot-devices/

via

https://github.com/jasonacox/tinytuya/issues/249

fade in and fade out lights using "smart lamps"

Testing out various options, found that these two smart lamps do the job for us.

Wipro Next Smart Wi-Fi 20W CCT+RGB LED Batten | 16 Million Colours | White Tunable | Dimmable |Scheduling |Scene Creation |Smart Grouping of Lights |Amazon Alexa & Google Asst. Compatible |Pack of 1 

https://www.amazon.in/gp/product/B08D19X3LS/

Amazon Basics - 9W Smart LED Bulb with Alexa, Google Assistant & Wi-Fi Enabled (Pack of 1, Cool White, B22 Holder)

https://www.amazon.in/gp/product/B0BBMRCFGW/

Apparently these are Tuya compatible, since they are controlled using the Smart Life app. 

We could control these using either the Smart Life app or the Alexa app, but in case future updates cause them to need a subscription or something, I thought of writing python controls for them with the help of tinytuya.

 Some notes. 

The Amazon basics lamp - for pairing, need to put on and off exactly 5 times: doing 10 times does not seem to work. First need to set up in smart life app, only then can we set up in Alexa app.

1. Smart Life App -

• way to add - add a 'home', add a 'room', add devices
  
  Click on the + in the top right to create "scene" (fade in fade out etc)
  Tap to run
• minimum delay between actions is 1 second
• does not seem to support drag and drop to rearrange actions in a scene.
• does not seem to support "local" control - the control type seems to be set to 'cloud' only.
  

2. Alexa App - 

• When adding device in Alexa app, choose Smart Life as the device type.
• After adding to "office" group, More... for custom routine.
• minimum wait time is 5 seconds. For shorter wait, a workaround as suggested at reddit is to make Alexa say something with 0 volume.
  
  https://www.reddit.com/r/amazonecho/comments/qtz9ob/create_1_second_wait_in_routine/
  
  Action: +
  >(Device Settings) Set Volume 0
  >Next>Action: +
  >(Alexa Says)>(Customize)>Count 1
  >Next>Action: +
  >(Device Settings) Set Volume 5
  >Next>Save<  

3. Writing a python script: I'll write a separate post on my notes for writing this set of scripts.

 

Excel as a sport

Interesting concept - Excel as an E-sports event!

https://www.theverge.com/2023/8/3/23817220/excel-esports-espn-the-ocho-championship-start-time-watch

passport renewal from a rural location

A follow-up post from my earlier post on applying for a passport.

• Since I had a "Post office passport seva kendra" or "POPSK" in a nearby town of Hindupur, I could avoid going all the way to Hyderabad.
• In the online application process, I had to change my Passport regional office from Hyderabad to Vijayawada, since Hindupur comes under Vijayawada. But for doing this, the procedure is to create a new account - the old account's office cannot be changed. I checked this with the toll-free helpline too.
• After filling up the online form, after the last step of making the payment, there is no option to choose a date and time of convenience - the appointment date and time is intimated. So, if we need an appointment on a certain date, we have to keep checking the available dates at their website. Booking an appointment in the evening after 8 pm, like I did this time, seemed to result in an early morning appointment.
• The only required documents shown at the passportindia.gov.in website were the old passport and a proof of identity, for which Aadhaar was available. I took my Voter Id card (EPIC - Election Photo Identity Card) too, and also the old copies of all the documents I had taken for the initial passport application. That turned out to be useful. The passport officer asked me for proof, since I was born out of India and was a citizen by descent and not by birth. Note to self - take the old passports of parents on next renewal too!
• The officer also asked me to take fresh xerox copies of the first and last page of the passport on a single sheet, as well as page 4 & 5 (blank pages). I wonder why. Perhaps as proof that I had not gone out of the country after receiving the first passport booklet in 2014?
• The officer also refused to recognize my class X SSLC certificate (which is a booklet as they used to issue in Kerala in 1990), saying that it is Transfer Certificate and not X class certificate(!) and finally took my MSc certificate as proof that I'm a matriculate for "non-ECR category"!
• But overall, very smooth. Only 2 people in a tiny office - the passport officer and a lady to take the passport photos and biometrics - thumb/other fingerprints. As soon as we entered at 9:35, he started working with us, since ours was among the first appointments, even though the appointment time was 10 am. 
• And the drive to Hindupur was very pleasant, via a National Highway to Lepakshi. Left at 7:45, reached Hindupur POPSK at 9:15 after stopping for coffee and a restroom break at AP tourism hotel next to the Lepakshi bull. On the return, started at 10:10, reached back at 11:45 after taking the non-toll road, since we had to wait at a level crossing for a goods train for a few minutes. 

pdf find and replace

 There was a request to find/replace some boilerplate text in a few thousand pdfs. One option was to use  https://pdfreplacer.com/

and see if it does the job with a watermark.

(You may need to find/replace one line at a time).

Just for interest, looked at scripting options, found it was quite complicated. Some notes below.

Learning more about pdf

https://automatetheboringstuff.com/2e/chapter15/

Visual studio

file 

open with 

binary editor

But does not find the text, because it is zlib compressed.

Is there a C++ library to extract text from a PDF file like PDFBox for Java? - Stack Overflow

Code to extract plain text from a PDF file - CodeProject

podofo/podofo: A C++17 PDF manipulation library (github.com)

https://pdfreplacer.com/

($30 for no watermarks)

libreoffice - Command line tool to search and replace text on a PDF - Ask Ubuntu

Repository search results (github.com)

google search github for pdf tools

Hosted docker-based,

https://github.com/Frooodle/Stirling-PDF

(but no find-replace text in api

https://app.swaggerhub.com/apis-docs/Frooodle/Stirling-PDF/0.10.3#/Convert/processPdfToRTForTXT

)

https://www.google.com/search?client=firefox-b-lm&q=github+pdf+bulk+find+replace+text

https://gist.github.com/artlbv/ec1153feb8644c613b4c7110b25e49ed

https://coolaj86.com/articles/extract-text-from-pdfs-even-protected-ones/

above gist does not work in our case, since the text is not encoded as ANSI/ASCII text but using the font's encoding, in hex.

font subsetting

https://stackoverflow.com/questions/9871585/how-to-find-and-replace-text-in-a-existing-pdf-file-with-pdftk-or-other-command

 

How to replace text in PDF file with Embedded Subset Fonts? | VeryPDF Knowledge Base

pdftotext seems to be from xpdf and not pdftk

how to extract text from pdf with embedded subset fonts - Super User

list of pdf tools and what they can do

PDF Tools and How to View and Edit the Code of a PDF File in Linux | Baeldung on Linux

exploring mutool

and qpdf too

4. Text - Developing with PDF [Book] (oreilly.com)

to see how we can convert the Tj Tf etc to text

pdf generation - How to convert the PDF content code to the type like "(<0034>) Tj"? - Stack Overflow

The <0065> does not just mean 0x65 in ASCII - 

pdf generation - PDF Tj command with angle brackets? - Stack Overflow

python possibilities

Search and replace for text within a pdf, in Python - Stack Overflow

Blog post with 2 ways to replace Replace Text in a PDF File with Python - ThinkInfi

Both are not useful.

Method 1 changes the formatting

Method 2 makes the pdf into an image.

https://www.codingunit.com/c-tutorial-binary-file-io

since xpdf pdftotext seems to work, checking out the code

https://www.xpdfreader.com/download.html

PDFDoc.cc

Page.cc

https://www.codingunit.com/c-tutorial-binary-file-io

https://stackoverflow.com/questions/2094253/editing-pdf-with-xpdf-or-with-something-else

Although not a library in traditional sense, Pdfedit has scriptable editing capabilities. But it requires QT. PodoFo probably fits best at your requirements. There's also PdfHummus.

http://pdfedit.cz/en/pdfedit_windows.html

(unmaintained)

https://github.com/podofo/podofo/issues

(maintained, but unstable)

https://github.com/galkahana/pdf-text-extraction/tree/master/TextExtraction

(apache license.)

IDR solutions guide

page 16 says

The Font encoding object specifies how to translate this value into a character (and these are defined in Appendix D of the PDF Reference specification)

But these seem to be similar to Unicode / ASCII, with d being  0x64 - not 0x47 as seen in our document at an offset of 18020 or so.

pg 27

RUPS free tool for diagnostics of PDF

Diagnostic tool for updating PDF Syntax and debugging PDF code | iText PDF

Using RUPS

Thence to find the encoding Identity-H

Identity-H encoding - Adobe Support Community - 10400841

https://community.adobe.com/t5/acrobat-discussions/identity-h-encoding/td-p/10400841

So, looking for a ToUnicode CMap

found the stream,

to understand bfchar and bfrange in cmap

How are Embedded CMAP tables defined in a PDF File? (idrsolutions.com)

https://blog.idrsolutions.com/how-are-embedded-cmap-tables-in-pdf-file/

we see a pair

<0047> <0064>

So, 47 in the pdf is mapped to unicode 64?! Yes!

We can use

http://brendandahl.github.io/pdf.js.utils/browser/

instead of RUPS to browser (but RUPS can also edit/change the PDF.)

Unfortunately, there is no space 0x0020 mapped? So we can't just replace the characters with 0x0020 or something similar.

The "correct" way to do it might be to use the "stamp" function to overwrite the offending text with what we want.

The "hacky" way might be to replace the characters with something else and make them not visible, if that is feasible.