teri

I heard that indeed the Italian government did not do a good job teaching their kids about fascism. In other places such as Germany and Switzerland (and many others, I hope) most students get to know how fascism works and how harmful it is. At least when I was in school, this was part of the teaching program. But in Italy? Apparently not.

This sounds really bad and I'm afraid it's even worse than it sounds. For example, end of 2022 Italy introduced a law which bans certain types of harmless assemblies (rave parties) of more than 50 people (participants can be put to jail for up to 6 years). This is paving the road for cutting possibilities of people to protest against their government. (Imagine, there is a a protest and somebody plays music. The government will see a rave party.) Yet the Italian government seems to have no problem with 3000 neofascists visiting the hometown of Mussolini. And now this.

Cool, never saw any of those!

No, haven't seen. What is RHEL doing?

Spontaneous idea of how to use copyright law for keeping Meta out of the Fediverse (more for fun):

Introduction: Parts of the Fediverse, including Mastodon, are software licensed under the APGL license. This license is a great choice because it forces the ones running the software to grant users access to the source code. GPL for example would allow to run proprietary services based on GPL code. The AGPL does not. Companies like Meta and Google will likely not use AGPL code because it might force them to also publish their proprietary systems behind the scenes. However, this does not help much for keeping the Fediverse save. They simply implement their own software which will not be open source.

Therefore we may need another approach. Defederating is the simplest and in my opinion currently the best. It's easy and keeps people in control.

However, there could be some 'automatic' approach using copyright law. It's a hack which allows to use existing law to regulate the way instances can federate.:

• instances would Federate only if the other side can provide a certain piece of information called X
• X is protected by copyright law, therefore by default, instances are not allowed to provide X
• However, X is released under a license which for permits to copy and distribute X under certain conditions
• The conditions allow to tune who can legally federate
• Conditions could be
  • The server software must be AGPL licensed
  • The instance must not be owned by a company with a certain amount of annual revenue

Open question is, who owns the copyright of X?

The EDA/CAD vendors generally seem to aggressively kill competing projects. You cannot trust any project as long as it is not open source under a copyleft license (plus does not require contributors to sign contributor license agreements).

The decent PCB tools I know of is KiCAD and eventually LibrePCB, which I never tried.

I once got inspired by this: https://scoollab.web.cern.ch/diy-particle-detector It uses photodiodes instead of BJTs. Advantage is that you can get some which are easy to decap. I did not get it to work :/ but also did not spend so much time. I think what is really important is to properly shield the circuit from electromagnetic radiation, use a battery (low-noise) and also shield the detector from light.

KiCad. Stay away from closed-source tools. They'll all try to press out the max amount of money sooner or later. Or get bought and discontinued for eliminating competition.

Difficult to find datasheets...

Second picture: To me looks like Q1 (3400) is an NMOS which connects the LEDs to power. The 'gate' is marked yellow. It is the 'control' input of the MOSFET. The 4 resistors RD RC RB RA (purple) probably limit the current through the LEDs. If you're lucky, then an easy hack would be to bypass this transistor: Remove Q1 and connect the red and green marks via a mechanical switch. You'd need to scratch away the white coating until you get copper. Then solder wires there. As a consistency check you could measure the voltage over Q1 (red and green marks). Measure once when the module is plugged in but switched off, and once when the LEDs light up. If you see a voltage while the LEDs are dark, then this would partially confirm my guess. As a test: Before removing Q1 you could also try to bypass Q1 with a resistor (~ one, two kOhm) while the LEDs are off. If you see LEDs now lighting dimm, you know that Q1 is the one switching the LEDs.

Other possibility: It might possible that the timeout is computed with an R-C circuit. for 30min you need a rather large resistor because there can't be very big transistors. R1 (blue) is the largest one. Experiment could be to remove it and see if timeout is still 30min. Or put another 10k resistor in series and see if the timeout gets shorter.

Wondering: What voltage do you connect to the module? Please be extra careful if you have somewhere mains supply voltage.

For the first one: Would be good to know what U1 is exactly. Can you read any number/code on it?

If you are lucky it is simply a 'switch' which switches on once you press the button and switches of after 30min. In this case you could add a wire which constantly switches the lights on. Then add a normal mechanical switch to the power supply cables.

If you are unlucky, U1 also regulates the current through the LEDs. Then cannot be easily replaced. Is there anything on the other side?

In general, that's definitely a problem too. Just look at Windows 11. Now M$ wants you to buy a new laptop such that their OS can run at all. I use a >10y old laptop with Linux. No performance issues for everyday use.

Going from Google-Android to LineageOS boosted the battery life of my old Smartphone by a factor 3 roughly. There where some Google-things consuming battery in the background, probably spying on me. Also using simple Apps from F-Droid helps to get a smooth experience.