Currently Chia/MMX compressed plots will only work properly on Flexpool. Max has been using us since September and several changes were needed to properly support them. The major one being self-adjustable difficulty. You may notice issues on other pools. On the bright side, we have a 0.7% fee which is lower than most pools.
I have a supplemental Ubuntu Gigahorse guide that I’ve posted in the Gigahorse release post.
And yet, Spacefarmers.io and others also have those capabilities and always have. Not trying to be rude, it’s great you have worked with MMX to get this going, but you aren’t the only game(pool) in town if someone wants to pool with these plots.
On a small scale you’ll be fine on a large scale there are advantages to Flexpool. I’ve explained this to many including in MMX discord and Max has confirmed why. He’s also messaged a few pool ops to tell them some of what they need to implement.
Long-term FlexFarmer is a massive advantage in terms of power, resource use, and equipment wear. Once it supports compressed plots it’s going to represent an even bigger advantage as you’ll be able to farm more plots using less power than the Chia farmer as it has better code. Max said that too.
And yes thanks to extremely hard work we’re the only pool that’s kept up with HPool and Space as new/old farmers all switch to them. Given that 99% of new farmers join one of the 2 largest pools I would say we’re doing an excellent job.
If someone is already running a full node on RPi, not much of power on the RPi side is saved (compared to the rest of the farm).
I was running my node on i5-8250U NUC, and that box was drawing ~10-20 W at the wall (USB only). I did run FF on it, but cannot say I saw much of saving, or at least on the level to justify the switch. Although, I think having something like FF even just as a backup is really a good thing.
Sure, resources are much lower, as no node, no NVMe, no worries about blockchain, are needed.
Equipment wear (I think you mean HDs), is just FUD, or maybe I should call it baseless marketing pressure. Those drives run basically idle, so sure there are big improvements, but still on the idle end, so who really cares. Last time you provided power stats, the results were well within the error margin (as expected), basically meaningless and you were trying to draw some conclusions from that test (any comparisons, tests that don’t have error ranges are just garbage).
The cost for that (at least for me) are basically BS logs that no existing chia log analyzer can bite (so no farmr, chiaDog, …). This one was one of the main reasons to not use FF or me. A couple of other reasons where there, and I left FlexPool.
And I agree with you that the fact that majority of people join SpacePool is highly disappointing, as they are basically killing the pools ecosystem. I don’t think anything can be done about it.
Well for compressed plots it really becomes relevant if you can do better per a plot which is the plan. Max and Alex have both said its doable as FF has better code. So less power and resource use per plot means more plots farmed for the same hardware or a higher compression.
It reduces I/O and I/O is limited under warranty so it definitely plays a factor but its just one small part of the car. So you reduce your risk of I/O related failure. This is something you can test yourself, we haven’t tested post Direct I/O but previous versions of FF saw around a 40% reduction.
And definitely its good to have it as a backup since even if you don’t use it if your node goes down while your with FF you can just DL FF and start it up while you fix the node.
I know I get hammered for marketing but as you agree, if I didn’t everyone would just go to 1-2 pools. We only stay in this thanks to constant effort on my part and great software on Alex’s part. The pools that used to be near us have all fallen way behind as their customers all switched to the top 2 pools. My marketing looks weird only because no one else bothers. I honestly don’t do that much, a few times a week is not a lot? Someone criticized me for constantly posting on reddit when I hadn’t posted in over 2 weeks. We charge a lower fee yet we struggle because everyone goes to xxxpool because they are the biggest. Imagine if everyone just moved to China or Russia based on the principle that bigger is better. We keep things interesting and provide an alternative. We also provide a lot of competition and based on fee, service, and products we’re the #1 pool.
The main advantage of flexfarmer, for me, is that it is so lightweight that you can run a copy on your main farm, run a copy on your pi with a couple of usb disks, run it on a server doing storj as it’s main job…
If you have a suspect disk or controller you can put a dedicated flexfarmer on it where the chia software might just stop farming altogether
At the moment I have three flexfarmers on WIFI warming the house in different locations
Every time we had a similar exchange, it ended up either with you stating that you have no clue about the technical side or telling me to see psychiatrist as you had no more bullets. So, thank you, but I will pass on this nonsense you just brought up, as again all that is way over your head.
Also, you read me wrong. I am not against your marketing, rather I really appreciate your efforts, and I give you a lot of credit for helping people.
So, maybe try to own the original problem that was posed, as that may be the key why people are leaving Flex.
I’ve never seen a HARD DRIVE that limited I/O in the warranty.
SSD drives on the other hand, limiting I/O and it’s resulting wear MATTERS - but that’s mostly a writes issue.
