Raspberry Pi Zero 3W and Pi 4 Model A spotted in device trees

Hardware & feature wishes for Zero 3W / Pi 4 Model A

• Many want better Wi‑Fi more than more CPU/RAM: 5 GHz 802.11ac, wider channels, and WPA3.
• Requests for stronger/better antennas; current Zero 2W/older Pis seen as weak. Some hack in external antennas; others argue RF switching adds too much BoM cost at this price point.
• Desire for updated Bluetooth (e.g., BT 5.x with LE Audio).
• Strong demand for USB‑C (especially for gadget mode), with frustration at lingering micro‑USB OTG.
• Memory asks cluster around 1 GB RAM for Zero 3W, with some arguing a cheaper 128–256 MB variant would be fine if it kept the price low.
• Some want more integrated features (UPS, faster PCIe, SDR, better networking), while others strongly prefer keeping the boards barebones and cheap.

Pricing, availability, and market shift

• Repeated nostalgia for the original $5–$10 Zero; many say the Zero line only makes sense if it’s in the “so cheap you just hack on it” range.
• Persistent frustration about years of poor availability and scalper pricing; some say the Zero is effectively dead if supply doesn’t improve.
• Examples of formerly cheap SBCs (e.g., Orange Pi Zero) now costing several times more; blamed on RAM/NAND prices, COVID-era shortages, and sometimes “revenue maximization.”
• EU VAT/customs changes are seen as having severely hurt hobbyist access to cheap Chinese hardware.
• Some pessimism that computer hardware will ever be truly cheap again; others think cycles and new tech will eventually push costs back down.

Alternatives and ecosystem

• ESP32 (and Pico/Pico W) frequently cited as the new “cheap hack” platform: far lower power, enough for GPIO/automation, but not a Pi replacement for many workloads.
• Some argue the full-size Pi’s niche is shrinking (ESP32 for GPIO/IoT, mini PCs for “real computer” tasks, TVs for signage).
• Competing Pi‑Zero–form‑factor boards (Radxa, Banana Pi, Orange Pi) offer better specs on paper, but many say RPi’s software support and ecosystem remain a decisive advantage.
• Counterpoint: Rockchip and others now have good mainline support; Armbian is highlighted as key for making non‑RPi boards usable long‑term.
• A few suggest using AI tooling to reduce porting/SDK pain, while noting a gap between “works for me” hacks and upstream‑quality support.

Wireless reliability concerns

• Several reports of instability and crashes with the CYW/BCM43455 Wi‑Fi used in Pi 4/CM4 and other boards, especially under load.
• Debate over whether this is a chip design flaw, firmware bug, or thermal issue; suggestions include firmware re‑implementation via decompilation, though Wi‑Fi is seen as very complex.
• One commenter notes similar historical RF hardware defects in other vendors’ parts that were never seriously recalled.

Raspberry Pi organization, design choices, and openness

• Some feel enthusiasm has cooled due to perceived anti‑consumer behavior, higher prices, and focus on industrial/ OEM customers over hobbyists.
• Criticism of long‑standing reliance on Broadcom: weak docs, binary blobs, odd boot chain (GPU-first), and hostility to open source; others respond that very few users actually care about blobs.
• Speculation that refreshes (like Zero 3W / Pi 4A) are driven by DRAM discontinuations and possibly new EU regulations.
• Irritation that ARM SBCs still rely on hard‑coded device trees in the kernel, instead of self‑describing hardware like PCs.

Use cases and perceptions of power

• Some remain highly impressed by how much can be done with tiny, cheap boards (e.g., Pico W sensor platforms, multiple Pi 400s as small servers).
• Others argue the Pi 5 is far from PC‑class performance in absolute terms, though its perf‑per‑watt and physical size are acknowledged.
• Diverging views on whether increasing Zero performance is even needed versus just modernizing the original low‑end design.