The Chia blockchain difficulty increases by almost 40% in 2 days (2700 → 3700), while the network size remains roughly constant (30 Ebit). 
1 Like
CNI have been adding the ASIC Timelords, and some have ramped up a bit quick, which sped up the signage points, and thus the difficulty increases. Its just a temporary blip, at least that’s how I understand it.
4 Likes
hm, ok i always thought the difficulty is calculated to get a constant block generation over a certain period of time.
so if the net size does not increase, the difficulty should remain the same, otherwise too few blocks would be generated.
i must have misunderstood something 
You’ve not misunderstood (and I don’t understand it anyway), just something went wrong - it was discussed quite a bit on the Chia off topic Discord.
That would depend on how many rewards ( xch ) were given out ( farmed ) in a set period.
To many diffuculty up, to few diff down.
You got this right, you just didn’t add in everything that impacts it, Proof of Space AND Time, PoST…
while (Timelords) “Time” isn’t rewarded with blocks, it acts as a metronome keeping a steady pace on the network (VDF - Verifiable Delay Function), if time speeds up, more signage points come, if time slows down, less signage points come, and so difficulty adjusts to this as more or less blocks are made. Keep the same pace of time, but add or remove space, the same thing happens.
This is part of the secret sauce allowing us to store our proof of work and turn it into space (plots) which consume far less energy over time, so we don’t have to compute everything on the fly like Bitcoin, Litecoin, etc., in fact once SSDs take over storage in the future, we’ll use 10-100x less energy, so we can increase security with more space while consuming even less energy!
Pretty wild stuff when you sit back and compare the future of Chia to other coins.
P.S. Here are a couple links I’ve been posting on reddit for those who want to dive into things more.
4 Likes
While I agree with your overall positive sentiment towards Chia (it’s the reason I’m also interested in it since its beginnings), it’s good to keep in mind that many blockchain projects failed despite being technically better and trying very hard to innovate. Many things can happen to Chia that we can or cannot foresee.
Some of the examples of technically sound ideas that still failed - not exhaustive just the top of my mind: burstcoin (proof of space), grin (mimblewimble), bitcoin interest (progpow), and I’m sure many more.
Well some of the examples deserve a response, like Burstcoin (now renamed Signa/Signum to try and escape the past) which had a group of developers who actively stole computer resources from people and tried to spin it as “testing” after being caught, so to even consider that in the comparison is being a little too far out and dishonest here… they weren’t a technically superior group to compare anyone to, they were thieves stealing from users.
Thats why Burst failed, not because it was superior and didn’t catch on. (I could go into the others because they too have reasons, but I think you read too much into my point, it was a very simple comparisons to the underlying hardware and future costs and thus average user participation, which is one of the hallmarks of Chia, as well as critical to true decentralization, not just in name only.)
When you look at the underlying hardware chosen for Chia, it will be able to reduce power consumption by 10-100x maintaining the same security, while asics may add more hash/watt over time it still uses the same watts and is out of reach for the average person, so I think its pretty remarkable a coin can hang in there with the best while dropping power use (without switching to PoS and all its inherent vulnerabilities)… all while the home user can still participate, and that is pretty darn wild.
2 Likes
You seem to have missed my point entirely.
Chia’s decentralization is quickly becoming a joke, with the advent of pools that don’t rely on the official pooling protocol (that this forum is used to promote despite my many warnings and CNI doing nothing about it). The time component is what makes PoST possible, but requires asic timelords which are costly to develop and not incentivized by the protocol itself. It’s easy to see how an attack on the blockchain would take place, especially if CNI is financially weak, which it recently admitted. That’s the predictable part. For the unpredictable you have to look into the social science aspect of technology, where Chia faces most of the same problems as BTC in interfacing with pre-existing institutions, albeit at a faster pace given their focus on financial services.
I didn’t miss it, I simply pointed out you misunderstood my point and attempted to insert some new argument that you wanted me to defend now, don’t do that, its never cool. Then when I did respond to your comment (my fault for taking the bait) and showed areas that clearly need improvement for your “comparisons” you took that as a chance to double down and insert more things you expect me to defend unrelated to my original point… and so the never ending circle begins.
“Here are a whole bunch of potential bad things, defend them all”
This is just not going to be a good discussion for a thread titled “What is going wrong with blockchain difficulty?” Feel free to make your own thread and pick those fights with willing participants who like to doom and gloom because they thought of a scenario where things could fail, no shortage of people willing to participate in something like that for any topic in life.
4 Likes
If it were required, what were we doing before they created and implemented them?
Commodity hardware was used before, iirc the most important factor is frequency so high-end CPUs. CNI made their own asic timelords (and distributed them privately to a selected group of people) in order to prevent a bad actor from being first on the network with a custom chip, which would have sunk the network instantly. There is always a threat of a bad actor developing a competing asic, which would open the network to attacks and make Chia blockchain’s security very weak (eg. double spends gallore)
Feel free to believe what you want to believe, then 
I’m merely sharing my opinion so that others can make theirs. With everything you wrote here, at least, there can be no doubt as to the extent of your intellectual honesty.
1 Like
Source?
Thankfully there are also good actors, its not all doom and gloom.
As with many crypto attacks, the vectors are partially negated by cost.
Why throw money at something to attempt to destroy it when you can build and prosper.
Sorry for ot op.
In your threat scenario, this bad actor, why didnt he attack the network before the asics were cooked when it would be much cheaper vs after asic.
He would be stupid not to. Did it happen?
It’s on Discord, I saw it myself a few days or so ago.
The exact statement, or one of a similar nature on the same topic?
Beecause how the reader interprets then rephrases matters sometimes.
( and you know, its aurelius.)
Im not on the discord.
Im still not convinced it was required, even if some peeps said it was.
I mean it was working before, no?
1 Like
Think yourself lucky you’re not on Discord, think of a forum with half a dozen threads for Chia, and then everybody posts in those threads, and it ends up a complete mess. You can get a couple hundred posts appear overnight in one thread.
Needless to say, I won’t be finding the relevant posts for you.
2 Likes
(I can’t believe I am taking the bait again)
Your “opinion” included a comparison to a criminal element who actively stole computer resources from people until they got caught and ostracized from a community causing it to fail, to a registered business that is working hard to comply with laws and conduct itself with known actors who would face legal consequences should they act the same way your “comparison” of anonymous thieves did…
“Bad things happened here, so it could happen there” is not an opinion, its hogwash.
2 Likes
Can you comment on the asic difficulty thing?
What happens if someone other than chia trusted entities has the fastest timelord?
Why did the difficulty change?
1 Like
I answered some of these questions and linked to several resources in an earlier comment so you can deep dive as much as you would like on the Chia documents website, one of the links leads to an explanation of the potential types of attacks and the requirements… with only a single much faster timelord the cost is enormous for the little advantage gained, with an “infinite number” of much faster timelords its much more expensive because infinite is a pretty big number to pay for. 
So getting ahead of the curve and distributing the new timelords worldwide helps add to the security of the network while also mitigating other concerns like dependence on Chia to maintain timelords (the current hardware costs and wattage of CPUs being used today is not small), so well distributed low power asics that are much faster than what the next decade of general CPUs will likely do sets things up for the future.
(hopefully you can click the link above to the previous message with resources and information.)
P.S. Remember you only need one honest timelord for the whole network, so having hundreds of low power superfast timelords worldwide operated independently really helps with decentralization.
2 Likes
doc says asics don’t exist (sulking)