The Saga Of 32-Bit Linux: Why Going 64-Bit Raises Concerns Over Multilib

July 9, 2019 by Tom Nardi 137 Comments

The story of Linux so far, as short as it may be in the grand scheme of things, is one of constant forward momentum. There’s always another feature to implement, an optimization to make, and of course, another device to support. With developer’s eyes always on the horizon ahead of them, it should come as no surprise to find that support for older hardware or protocols occasionally falls to the wayside. When maintaining antiquated code monopolizes developer time, or even directly conflicts with new code, a difficult decision needs to be made.

Of course, some decisions are easier to make than others. Back in 2012 when Linus Torvalds officially ended kernel support for legacy 386 processors, he famously closed the commit message with “Good riddance.” Maintaining support for such old hardware had been complicating things behind the scenes for years while offering very little practical benefit, so removing all that legacy code was like taking a weight off the developer’s shoulders.

The rationale was the same a few years ago when distributions like Arch Linux decided to drop support for 32-bit hardware entirely. Maintainers had noticed the drop-off in downloads for the 32-bit versions of their distributions and decided it didn’t make sense to keep producing them. In an era where even budget smartphones are shipping with 64-bit processors, many Linux distributions have at this point decided 32-bit CPUs weren’t worth their time.

Given this trend, you’d think Ubuntu announcing last month that they’d no longer be providing 32-bit versions of packages in their repository would hardly be newsworthy. But as it turns out, the threat of ending 32-bit packages caused the sort of uproar that we don’t traditionally see in the Linux community. But why?

Continue reading “The Saga Of 32-Bit Linux: Why Going 64-Bit Raises Concerns Over Multilib” →

New Part Day: The $15 ESP32 With Cellular

July 9, 2019 by Brian Benchoff 51 Comments

Cruise around AliExpress for long enough and you’ll find some interesting new hardware. The latest is the TTGO T-Call, an ESP32 breakout board that also has a cellular modem. Yes, it’s only a 2G modem, but that still works in a lot of places, and the whole thing is $15.

On board the TTGO T-Cal is the ESP-WROVER-B, the same module you all know and love that features a dual-core ESP running at 240 MHz with 4 MB of Flash and 8 MB of SRAM. Add to this WiFi and Bluetooth, and you have a capable microcontroller platform. Of note is that this board includes a USB-C port, ostensibly wired so that it behaves like a normal USB micro port. That’s neat, 2019 is the year USB C connectors became cheaper than USB micro connectors.

In addition to the ESP32 module, there’s also cellular in the form of a SIM800 module. This module has been around for a while and used in many, many cellular-connected projects and products like the ZeroPhone. This module is only a 2G module, and that’ll be going away shortly (if not already) in built-up areas, but this can serve as a building block for modules that have more Gees than a 2G module. That said, if you’re looking for a WiFi and cellular bridge for fifteen bucks, you could do a lot worse for a lot more money.

Rock, Paper, Neural Net

July 9, 2019 by Brian Benchoff 6 Comments

You might think the game of Rock Paper Scissors is just the random chance, but that’s not true. There is a strategy for Rock Paper Scissors, multiple ones in fact, and the best human players can consistently beat any Joe Schmoe off the street. But what about computers? [Paul] answered that question with a tiny little keychain dongle that can beat you at Rock Paper Scissors.

This is a neural network, and you need to train a neural network, so where did [Paul] get all that data? roshambo.me offers thousands of paper rock scissor games, and trained the network on more than 85,000 human games, along with about 10,000 simulated games. Rock Paper Scissors isn’t a complicated game at all, and the entire neural network is stored on an ATtiny1614 microcontroller. The calculations are done as floats, even. That’s how non-computationally intensive this project is.

Building a neural network is one thing, but putting it in a handy keychain enclosure is something else. This handsome device fits on a PCB just larger than a 2032 coin cell battery and is enclosed in a 3D printed case. The buttons are 3D printed as well, with some clever application of fiber optic as light pipes for the LEDs. The end result is something that is slightly better than random chance at Rock Paper Scissors and shows off some matrix programming skills. Check out the video below.

Continue reading “Rock, Paper, Neural Net” →