Monday, June 17, 2024

USB-C cables have vastly different capabilities

USB-C is a wonderful standard. Apart from the ability to plug it in either way up, it supports Power Delivery and much higher speeds.  I've run in to some mystifying scenarios where a cable would charge but not carry data, would carry data but not charge, and lower than expected data throughput.

I'm a registered Apple developer and was keen to try out the macOS 15 Sequoia beta but I'm not an idiot so decided to install it on an external disk on my Mac mini.

The steps were:

  • Used Disk Utility to format a Samsung T5 drive as APFS
  • Downloaded the Sonoma macOS installer via the App Store
  • Installed Sonoma macOS 14 on the external drive
  • Booted from the external drive and enabled Beta Updates in the Software Update settings and chose macOS 14
  • When all that was done I booted from the external drive.
Sequoia seems quite stable and I've been able to build my applications with the Xcode beta.

But.... it did seem rather slow! I used the Blackmagic disk speed test to measure the drive and got Write: 38.9 MB/s & Read: 39.3 MB/s. 

The USB3.1 port is capable of 10Gb/s or 1,250MB/s so something is up.

Changing USB-C cables to one marked as 10Gbp/s I re-measured and got Write: 263.6MB/s & Read: 344.9MB/s. Almost 10x improvement but well short of what's possible.

Today, I bought a 1TB Western Digital drive. Blackmagic gave me Write: 804.1 MB/s & Read: 906.8 MB/s.

The Mac mini has two Thunderbolt/USB4 ports that are capable of up to 40Gb/s so I have ordered a Thunderbolt enclosure.

Oh, and just a note that the speed of the internal SSD on a mac is pretty amazing. Blackmagic measures the internal SSD as Write: 3049.6 MB/s & Read: 2856.4 MB/s. I guess that's why I noticed the slow speed.

Summary: If things seems slow, try another USB-C cable. (I wish they had a standard for markings on the cables with their capabilities).

Friday, June 14, 2024

How to spread misinformation

I've been seeing a lot of misinformation lately and noticing the techniques used. While discussing this with friend Tony, I wondered if ChatGPT could list ten effective ways to do this.

"Spreading misinformation effectively, while unethical and harmful, often employs psychological and social manipulation. Here are ten techniques commonly used:

1. Emotional Appeals: Crafting messages that trigger strong emotions, such as fear, anger, or joy, can bypass critical thinking and make the information more memorable and persuasive.

2. Repetition: Repeating a false statement frequently can make it more familiar, leading people to believe it's true (the "illusory truth effect").

3. Selective Presentation: Sharing only specific facts or stories that support the misinformation while omitting contradictory evidence creates a biased view.

4. False Authority: Citing fake experts or misrepresenting the credentials of individuals can give false information an air of credibility.

5. Echo Chambers: Sharing misinformation within closed or like-minded groups reinforces the false beliefs, as members are less likely to encounter opposing views.

6. Manipulated Content: Using doctored images, videos, or audio clips to create convincing but fake evidence supporting the misinformation.

7. Bots and Trolls: Deploying automated accounts and paid individuals to amplify misinformation across social media platforms, making it seem more widely accepted.

8. Headline Manipulation: Crafting sensational, misleading, or clickbait headlines that can spread quickly even if the actual content is less inflammatory.

9. Conspiracy Theories: Developing complex narratives that explain events through secret plots, which can be compelling and difficult to disprove.

10. Astroturfing: Creating the illusion of widespread grassroots support or opposition through fake reviews, comments, or social media activity.

These methods exploit cognitive biases, emotional responses, and the structure of information dissemination in the digital age to spread and entrench misinformation."

I wonder if some sort of score-card could be made that analyses a published story in terms of these factors?

Sunday, June 09, 2024

Raspberry PI Pico with USB-C not working? Try a USB-A cable.

It's a pity the Raspberry Pi PICO comes with the horrible micro USB socket. I purchased what, I guess, is a PI Pico clone board with a USB-C socket on it from AliExpress but was disappointed when none of them seemed to appear on the USB bus.

On a whim, I plugged one in to my computer with a USB-A to C cable and up it came! My guess is that somehow the USB-C wiring isn't right and a proper USB-C port doesn't power on. Anyhow, they work fine and I've now flashed MicroPython v1.23 on to them and all works well.

Thursday, June 06, 2024

A nice review of WSPR Watch in German magazine FUNK Amateur

My thanks to Stephen Rapley, VK2RH, for passing on this review of ham radio software that gave my WSPR Watch app a high rating. 

It's in the latest June issue of German amateur radio magazine FUNK amateur.

Translating the text with Google Translate I get:

"Who can hear me? Where can I be heard? Who can I receive? These questions can be answered with the WSPRwatch app. Your own reception reports and those of other stations can be displayed in a list, on a map, Figure 1, or as a graphical evaluation. However, the information does not only come from WSPRnet, as the name suggests. Timebase DB, Reverse Beacon Network and PSK Reporter can also be selected as data sources.

Users have a wide range of setting and filtering options at their disposal.

The data can be exported in CSV format for other applications. 

Linking the app to your own account adds further information to the spots displayed, such as the name and exact location of the call sign owner."

Stephen added "You appear at the top of a list of apps ranked by “Nutzerbewertung”! (User ratings)."

Thanks again Stephen.

Wednesday, June 05, 2024

Technology news on ABC radio

Had a nice chat with the brilliant Dom Knight about technology news including Ticketek, Windows Recall, Face recognition failures, Google AI summaries, "filmmaker" mode on TVs and what to expect from Apple's developer conference next week. 

You can hear it here.

I've been working on the acoustics in my home studio. It was a bit of an echo chamber so this week I hung a blanket from the ceiling behind me and it seems to work nicely.

Sunday, June 02, 2024

Scientific glassblower still going at UC Berkeley

I don't normally post links but this story caught my attention and deserves more coverage.

It's not that long ago that I bought some test tubes from a place on Harris Street, Ultimo, Sydney. They had a scientific glass blower working at the time. 

Sending audio to two devices on macOS

This morning I saw that the (very active) Bendigo Amateur Radio and Electronics Club were transmitting RTTY on 7.035. When I'm not in the shack I often leave an AirSpy HF+ running connected to SDR++ in server mode. This means I can sit by the fire and tune around on HF.

To decode a digital mode audio must be fed to FlDigi. I use BlackHole for routing audio from one app to another but I wanted to hear the audio as well so needed a way to send audio from the SDR++ app to both my speakers and a loopback for FlDigi. 

The trick on macOS is in the Audio Midi Setup app. Create a new "Multi-Output Device" and set the outputs to both BlackHole and your normal audio output (I use headphones with speakers plugged in to them. 

The audio output in the SDR software is set to send to the Multi-Output Device.

FlDigi is set to get input audio from BlackHole.

The final result is audio I can hear but also decode in FlDigi. Here's a bit of the BAREC broadcast.

Thanks BAREC for running the digital broadcast. I think it's great to have some regular signals on the band that aren't just FT8.