Family Photo Chest Part 13: Early Prototype Workflow

Good Morning from my Robotics Lab! This is Shadow_8472, and today, I am merging a couple projects into one and hoping they stick. Let’s get started!

Overview

I’ve been piecing bits of my photo trunk project workflow together now for way too long. Right now, the architecture is looking like I’ll be scanning sets of pictures to a directory on a Network Attached Storage, then I can use a cluster of dedicated microcomputers from another unfinished project to separate and deskew the raw data into individual files. These files will then live permanently on the NAS.

Progress is rarely linear though. My end goal for this week hopped around quite a bit, but in the end I felt like I did nothing but figuring out how not to proceed: ground work with no structure. Routers are hard when you’re trying to learn them on a schedule!

Lack of Progress

In a perfect world, I would have been well on the way to configuring a cluster node by now. In a less unreasonable one, I would have had my Pi 4 OpenWRT ready to support the cluster. Late-cycle diagnostics chased me into an even more fundamental problem with the system: Wi-Fi connectivity.

During diagnostics, I’m learning about how different parts of the system work. Physical connection points can be bridged for a single logical interface, and Ethernet cables can support separate ipv4 and ipv6 connections. I can’t configure the Ethernet (on either logical interface) the way I want because that’s how I’m connecting to the Web UI and SSH. I end up stuck using two computers besides the two router Pi’s (OpenWRT and a Raspian hack router that actually works) because I don’t like switching my Ethernet cables around on the switch, but I need to do that anyway when I have to copy a large block of text. In short: the sooner Wi-Fi gets working, the better.

I understand I am essentially working with a snapshot. It’s been tidied up a bit, but bugs still exist. Wi-Fi is apparently one of those things that’s extra delicate; each country has its own region, among other complexities. On the other hand, I don’t know if that’s actually the case, as diagnostics are ongoing.

Takeaway

I’m probably going to work on this one in the background for a while. The OpenWRT help forum’s polish at least in part makes up for routers being dull to learn. If it takes too long, I do have other projects, so I may need to replace the cluster with a more readily available solution.

Final Question

Have you ever had upstream bugs that kept you from completing a task?

Leave a Reply