Running on Linux is pretty easy. I've set up enough testing stations that, except for the dependency list, I have just about every other step memorized.
The only trick to it is making sure you don't put a zero where a capital O should be in the one line.

Though admittedly if they did it wouldn't be the first time. I'm still working on recovering from the sales and savings hit from last year's dick-swinging contest price dive.

Being robust against that attack is less of a concern than against other types of attacks. You'd need a massive cartel of people heavily invested in BTC in order to collect enough coin to drop the price that far, and it's likely that, in an economy driven largely by greed, you wouldn't be able to find that many big players interested in getting such a lousy return on prior investments of capital or mining efforts.

Dropping to 1 block per day would mean a 99.3% drop in global hashrate. Diff tends to follow price, but when prices drop diff drops at most half as far, probably on account of all the industrial low-cost-power miners who can ride out drops better than home users and poor planners. If a 95% drop in price were to happen, the short-term effect should only be a roughly 50% reduction in hashrate, for 72-block days, at least for a time. Even the big industrial miners would have to operate out of pocket while hoping for coins to rebound. If the price didn't come back up soon, then we could have trouble, but it wouldn't land all at once.

The guys who bought in that massive dump at low prices will also be hoping for coins to rebound, and possibly wrangling their own cartels of manipulators to pull off some token large buys in order to swing the price upward.

In the meantime, Bitmain might use the opening to push one of their crappy forks, so it's possible an alternative would gain a bit of traction, but most of the alternative options have the same susceptibilities.

Ooh, firmwares. Maybe in early December I'll have time to test out some of that, rig up S9 space heaters for the new shop. I've seen BM1387 go below 70W/TH at the chip and landing a couple machines below 80W/TH wall would be nice.

Honestly, I'd recommend talking to 419mining. Tell him you bought it from MinersSupply and he'll probably take care of you anyway, since MS kinda got booted as an official reseller and doesn't have stock anymore.

I'm a bit confused with your reporting though - does it say no chips found, or 2 chips found? Because [oo] would indicate it's detecting 2 chips. Either way, yeah it definitely sounds like a hardware issue.

What's the situation there? Has the ISP itself banned mining, or is it reporting to government because mining is against a civil law?

Because if you come here for advice on how to circumvent laws, you'll probably be disappointed.

Right now we're packing up the shop. The last units built in our current location are currently on the tester and will be dispatched to a reseller next week. The first truckload of tools and materials rolls out in the morning.

Good news is, stuff for the next batch is already on order. Most of it's already here. Production will start in the new shop not long after Thanksgiving. I recently redesigned the PCB so the new batch units will be quicker to assemble and quicker to test and verify, so that's good for you guys.

The other thing the R4 did was push air across only about four rows of chips instead of about fifteen rows of chips. With long paths like in the for-some-reason popular long tube design, the air cooling the rearmost chips is already quite hot so it doesn't work as well. Keeping the path short means you still have relatively cool air even at the last stage, which means you can reduce the linear flow rate (and therefore the blower speed and noise) and still get as effective cooling.

Spondoolies proved how bad a long path could be with their rack gear ages ago, and yet Bitmain still upped the ante for S7, S9 and the like. When they landed on S15 they had finally realized it was dumb and flipped things around.

Changing the S9 case won't shorten the flow path. You could remove all the heatsinks and turn them sideways for a more R4-like flow, or you could pull individual boards and put fans directly on them instead of blowing from one end.

I would switch over to my in-house server but it looks like the ISP's static routing is hosed so my stuff's been down for a couple days. Hopefully I can get everything updated and going again soon.

Anyone who's tried to get ahold of me by email since Saturday morning, drop me a PM.

The new fancy multi-core ones anyway. The 5V regulator on the R606 is the same circuit that powers the D/1 port pair on the GS Hub, and should be good for the older models but the new Pi 4 is hungry.

Well there you go. The Pi4 can be a pretty hungry machine, and the internal 5V, which is located at the hot end of the board and powers that output jack, is good for 3A but shared with internal loads like the USB interface chip, onboard micro, sensors and lights. If you overloaded that regulator and damaged it, the miner won't mine and the lights won't work, but the fan would still kick on.

Are you testing it with the Pi powered separately? Try that and see what happens.

My guess is there's an issue with the 5V circuit. Not sure why there would be, but if you're in the US and bought it from Eyeboot it traveled across the Pacific twice so something could have gotten jostled in shipping. The batch he received was tested for a full 3 days and power-cycled numerous times. Point of fact, I have yet to see one with a failed 5V circuit at any point in testing. But your symptoms sound like what would happen in that case.

What you should do is talk to Eyeboot about warranty work. If you can wait a bit I can repair it for you, but as noted above I'm breaking down my shop and moving to a new location this month so I wouldn't be able to look at it until after Thanksgiving, and either way you need to contact Eyeboot. The reseller is always the first point of contact for support.

So when it powers off, the fan is still powered and spinning? Or is it spinning under its own inertia? Does the orange light go off as well?

Usually when an R606 cuts out like this it's because of the power supply. But if the fan is still powered that would indicate something else.

We've got some miners to test, and some to ship out to resellers in the next week or so, and then production will be stalled while we move into the new facility I've been building since summer.

It's gonna be actually insulated, which is great because I didn't want a 100KW data center running just to keep the pipes from freezing.

But it does mean we'll have probably a three-week lull in production of basically everything. But fresh batches are already in the works for as soon as we get machines back up and running.

Probably an 8A brick? Then it's possible trying to run high speed and high voltage could be causing problems. 8A should be sufficient to see upwards of 600MHz around V4 even with heat problems. I had one on the tester last week with only one heatsink that pulled 7A at 550MHz. Usually it'd be under 6A.

Assuming everything's working right, I'd expect around 700MHz on V6 to be possible with that brick. Maybe you lost the chip lottery and your unit isn't terribly overclockable.

Okay that's great. But what's your power supply?

That doesn't make a lot of sense to me. What's your power supply?

Heat has nothing to do with the speed re-targeting, unless it's getting so hot that your power supply can't keep up. All pods are tested for 550MHz at V4, and if one's having a bit of trouble it'll be re-tested at V5. The balance and quality of the chips is a crap shoot; some can only hit the speed at V5, and some can do it at V1.

587.5MHz on V5 is about right. Higher speeds require higher voltages.

Then there's the ASICBoost thing, where it can't break 600MHz without AB enabled because of USB traffic limitations.

The lowest voltage at which it'll run stable for a given frequency is the voltage you want. Means less power and less heat.

The colder you can keep it, the better. I'd shoot for keeping it under 70C.

This isn't really the place to talk about 2Pacs, but there were two "versions" - the regular domestic green one, and the bitshopper-specific black one. In the original Compac days bitshopper licensed the design and did his own manufacture in Germany, with black PCBs and I believe a Japanese-made black heatsink with beefy cap screws. Honestly it looked pretty sweet. So when we started making the 2Pacs but BM1384 ASICs weren't readily available for a proper factory, I manufactured all of them in-house and we set up special batches for bitshopper using their original color scheme. Functionally and electrically there is no difference, just a different coat of paint.