Okay, I will share one small victory before I depart for nature or something.
I stopped working at $weDropPackets in May of 2016. I left behind several very successful projects, one of which was called "smpreconfig." For the first 11 years I was at $weDropPackets we were primarily a Motorola(later Cambium) Canopy WISP. The last year we pivoted to being a FTTH provider, but the Canopy product is still in use, and still installed new in areas where fiber does not yet exist. Canopy is an absolutely amazing last mile technology, but the tools provided for managing it at scale absolutely sucked. So, I wrote many, many programs to resolve this limitation.
smpreconfig was a Soekris net5501 with a quad port gigE card in it that used OpenBSD routing domains to successfully have seven gigE ports, all with the address 169.254.1.2/16. Canopy radios default to 169.254.1.1/16. Doing seven routing domains allowed you to interact with seven SMs (subscriber modules) at the same time. This device would detect the link up, reset the default password, set the installation color code, and some other network wide defaults. It also interacted with the inventory system I wrote for $weDropPackets, updating the device's history and location as back in availabile inventory. Whether the SM was new in the box, or returned from the field, simply attaching it to a free port on the smpreconfig device with a default plug in the timing port, and waiting about 30 seconds did the needful.
It was a wildly successful system that I first built in 2012 or 2013. When I left in 2016, there was one other employee who understood how it truly worked and was able to maintain it. Said employee left last February.
smpreconfig failed this week, and no one was able to repair it. This wasn't due to it being an esoteric or undocumented system, this was because $weDropPackets no longer hires people with actual skills. The pay is such shit that no one with skills would work for them.
So, a process with 15 years of institutional entrechment is now completely borked, and they have no expertise with which to repair it.
So, I may not have been able to make a single god damn thing work correctly all week, but it is nice to know that things I built ages ago chooched on for years after I was gone.
The Psychotic Network Ferret
npub1en…n4qpv
2026-08-07 18:23:10 UTC
Author Public Key
npub1envsrcj2vqrr8v4qjg2v9s43lzuuwwvr7vpmv9am8pmdyxhdgezqwn4qpvSeen on
wss://relay.mostr.pubPublished at
2026-08-07 18:23:10 UTCEvent JSON
{
"id": "880c05a5fe1f1558261fac238f43ae560e2d40d498ee484d5c357fe4f8b9d4ea",
"pubkey": "ccd901e24a600633b2a09214c2c2b1f8b9c73983f303b617bb3876d21aed4644",
"created_at": 1786126990,
"kind": 1,
"tags": [
[
"proxy",
"https://mastodon.bsd.cafe/users/nuintari/statuses/117055618441523472",
"activitypub"
],
[
"client",
"Mostr",
"31990:6be38f8c63df7dbf84db7ec4a6e6fbbd8d19dca3b980efad18585c46f04b26f9:mostr",
"wss://relay.ditto.pub"
]
],
"content": "Okay, I will share one small victory before I depart for nature or something.\n\nI stopped working at $weDropPackets in May of 2016. I left behind several very successful projects, one of which was called \"smpreconfig.\" For the first 11 years I was at $weDropPackets we were primarily a Motorola(later Cambium) Canopy WISP. The last year we pivoted to being a FTTH provider, but the Canopy product is still in use, and still installed new in areas where fiber does not yet exist. Canopy is an absolutely amazing last mile technology, but the tools provided for managing it at scale absolutely sucked. So, I wrote many, many programs to resolve this limitation.\n\nsmpreconfig was a Soekris net5501 with a quad port gigE card in it that used OpenBSD routing domains to successfully have seven gigE ports, all with the address 169.254.1.2/16. Canopy radios default to 169.254.1.1/16. Doing seven routing domains allowed you to interact with seven SMs (subscriber modules) at the same time. This device would detect the link up, reset the default password, set the installation color code, and some other network wide defaults. It also interacted with the inventory system I wrote for $weDropPackets, updating the device's history and location as back in availabile inventory. Whether the SM was new in the box, or returned from the field, simply attaching it to a free port on the smpreconfig device with a default plug in the timing port, and waiting about 30 seconds did the needful.\n\nIt was a wildly successful system that I first built in 2012 or 2013. When I left in 2016, there was one other employee who understood how it truly worked and was able to maintain it. Said employee left last February.\n\nsmpreconfig failed this week, and no one was able to repair it. This wasn't due to it being an esoteric or undocumented system, this was because $weDropPackets no longer hires people with actual skills. The pay is such shit that no one with skills would work for them.\n\nSo, a process with 15 years of institutional entrechment is now completely borked, and they have no expertise with which to repair it.\n\nSo, I may not have been able to make a single god damn thing work correctly all week, but it is nice to know that things I built ages ago chooched on for years after I was gone.",
"sig": "4317c2c7c134fdb28e5b24a4ae9362118c95c455f5b9c3103cf9daf5e9b19b083ef0f82820ceca86f5a8701621b1dde93e6e7c0d041fd52a695409fb7078c50e"
}