There are a log of Gmail users getting an unexpected surprise from Google if they use their Gmail password in other applications (Outlook, Thunderbird, etc.) These apps no longer work.
This is due to a security change Google is rolling out. Google now requires 2-Step Verification is activated and all ‘less secure’ apps must use an App password instead of a Gmail password.
Despite the surprise, this is a good security practice. I am all for it.
If you have your SkyWeather2 systems’ Email notification service configured with your Gmail password, these emails will be rejected by the Gmail servers. To fix this, you need to configure an App password for your SkyWeather2 system.
Fear not! I wrote up a ‘How-To’ document that walks through the process. It is pretty painless and only takes a few minutes.
You can find the document here.
As noted in previous posts, the latest version of SwitchDoc Labs SkyWeather2 software release (May 7, 2022 – Version 027.6) broke my installation script. The problem involved updates to the MySQL database schemas.
New releases of the SkyWeather2 software are updates. This means there is an assumption that you have a working installation of a previous version. My installation script starts from a clean Pi OS and installs all software and dependencies from scratch.
This has not been a problem historically, but the latest SkyWeather2 update makes changes to the existing database schemas, and assumes the databases exist. To correct this problem, I need to be able to create the databases before running the update scripts.
SwitchDoc Labs support has been terrific in helping me fix this issue. I was provided access to all of the information (e.g detailed database schemas) I needed to get my install script working again.
You can find the latest version of the install script here. It is recommended you install SkyWeather2 software on a Pi running the 32-bit Buster version of Pi OS.
It is interesting to note, my install script runs fine on the 32-bit Bullseye version of Pi OS with legacy camera support enabled. This is not officially supported but those of you that like to live on the edge can hack around and report any issues.
Thanks again to the great folks at SwichDoc Labs for supporting me in getting this issue resolved.
On May 7, 2022, SwitchDocLabs released Version 27.6 of their SkyWeather2 software suite. A few days later, I was notified by a user that my SkyWeather2 install script failed.
pi@skyweather2:~/SDL_Pi_Skyweather2 $ sudo mysql -u root -p < WeatherSenseWireless.sql
ERROR 1046 (3D000) at line 5: No database selected
This error was not caused by my script. It is caused by an error in the SkyWeather2 SQL scripts.
A couple of years ago, I bought a SkyWeather2 system but did not purchase the software SDCard. I was not willing to pay an additional $35 on top of the cost of the system. I soon discovered downloading the SkyWeather2 software on Github and running the main Python script does not work.
Several folks have brought to my attention my SkyWeather2 automated installation script fails to complete. The problem has been corrected and a new version of the script is available here.
The latest script has been tested with SkyWeather2 Version 027.1 and Raspberry PiOS dated October 30, 2021.
In a previous post, I raved about the SwitchDoc Labs SkyWeather2 KickStarter project. When you purchase the kit, there is an option to buy an SD Card with all the required software installed. I passed on it, and cloned the SkyWeather2 software from Github.
To get it all working, I had to sort through all of the dependencies. This took time.
As far as I know, there is no documentation anywhere that describes how to get the SkyWeather2 system up and running from scratch. So – Ol’ Sopwith decided to do something about it.
I created a bash script that installs all of the required application and Python module dependencies. In addition, I wrote a detailed 18 page “How-To” guide.
Hopefully, this makes it easier to get your SkyWeather2 up and running if you choose the DIY route.
Anyone who has ever messed with a Raspberry Pi knows the drill. You download the latest Pi OS release image and burn it to an SD Card. Then you slide the card into the Pi and power it up. The Pi boots, you login with the default credentials, and you run ‘sudo rasp-config’.
You configure your WiFi access point, locale, keyboard, and timezone. You enable SSH, I2C, Camera, and whatever else you need for IO. Next, you run ‘sudo apt update’, ‘sudo apt upgrade’, and reboot. Finally, you log back in and install all your favorite software that is not installed on the base image (p7zip, pip, i2ctools, midnight commander, etc.)
For casual Pi users, this is a one-time or rare task. For experienced Makers who have gone through this drill dozens, if not hundreds of times – it is a real pain. For Makers who write lots of code and/or software installation scripts, this process is beyond irritating.
Greetings smoke-eaters. It has been a while since ol’ Sopwith published a blog post. That is because I have been working hard on another clock project.
For those of you who do not want to read the rest of this post, here is the datasheet.
A while ago, I came across a very cool Thingiverse project by a French Maker comrade named jeje95. He created a replica of the time-circuit in the Delorean filmed in the classic movie Back to the Future. Certainly one of the best movies ever produced, with a near cult-like following to this day.
jeje95 also produced a pretty funny video of his creation, that I highly recommend you watch.
I produced one of these clocks, and as all Makers are obligated to do, I decided to make it better. I ditched the Arduino for a Raspberry Pi Zero, added a much better real-time hardware clock (RTC), and a whole lot more.
Way back in June 2019, I posted an article about the revival of the DOTKLOK project.
I worked hard and made some changes to the design. Unfortunately, I could not find the same LED displays that were used in the original project. The Chinese company SURE no longer makes these displays. I literally hounded them to find me some new-old-stock (NOS) and came up empty.
This means I have to find a different source of LED displays. I did find some alternatives, but they are a different size and use a completely different API specification. To get this to work, I have to redesign the case and rewrite the source code that drives the clock.
I have not given up on the idea, but there is a lot of work here. Stay tuned.
This is the final entry in the 4-Part blog posts re. my Prusa i3 MK3S printer. The previous posts can be found here: Part-1, Part-2, Part-3.
Once I completed assembly of my printer, I followed the instructions to do a self-test and calibration. The self-test failed while performing the “Loose belt pulley test” on the X-axis. The manual said to be sure the motor pulleys are not loose, and to be sure the screws are tightened against the flat part of the pulley shaft.
First, I checked all the pulleys and they were not loose. This was puzzling to me, because I could not get rid of the error. I finally decided to re-tighten the X-Y belts. This fixed the problem. I was quite surprised how tight the belts had to be in order to pass the self-test. If you run in to this same situation, check your belt tension.
The next challenge I had was getting the print head depth calibrated correctly. I followed the instructions using the ‘sheet of paper’ method, but the hard part for me was getting the “first layer” calibration correct. This involves getting the perfect Z-axis height at the beginning of the print. The instructions really did not provide a lot of help – they simply describe, in general, what the thickness of the print layer should be.
In Part-1 of the Prusa i3 MKS3S 3D printer kit build, I gave a brief overview of my build experience. In Part-2, I provided more details of putting the printer together. Here in Part-3, I finish up describing the challenges I had during the assembly process.
NOTE: Remember, the below steps are just highlights of my assembly experience. They do not match the assembly steps in the manual.
Step-7 – Installing the X and Y belts. As the build continued, the number of parts to physically work around as you assemble components increases. I found the section on belt installation challenging. The toothed belts are provided flat – not a continuous loop. You have to bend a belt end tightly around a bolt on the belt tensioners, and run them around pulleys.
Impression-7 – This part of the assembly took way longer than it should. It was very difficult to determine how to set the correct tension. This caused me great pain later in the assembly. I would say the belts mechanisms are one of the most important parts of a 3D printer.
Step-8 – Assembling the PSU. You may recall in Step-5, I assembled the PSU mounting bolts on the wrong side of the rail. Now that the printer was 75% assembled, it came to haunt me. When I went to mount the PSU, the screws were on the wrong side.
Impression-8 – The manual was very predictive. “Incorrect placement of the PSU holders will lead to issues later,” is the ground truth. I had to disassemble a lot of parts to correct this issue.