I want to ask the level1techs community about their opinions on the relative security of different blockchain consensus strategies. Specifically I wonder about blockchains exploring the niche for decentralized data storage networks. Some consensus strategies in that space include “proof of space,” “proof of space-time” and “proof of replication.” I would not want this discussion to be limited to any examples I explicitly point out.
My personal interest in this is predicated on finding the most objectively secure crypto to mine regarding blockchains that use storage space as a mining resource. There are a few coins exploring this niche somewhat recently. As most mining networks are centered around specific computing resources aside from different encryption, I find that this storage capacity based niche is a comparatively new one that is underdeveloped both as networked communities of individuals centered on blockchains as well as an efficiently monetized decentralized sector for data storage hardware. As such I feel this underdevelopment represents opportunity for early adopters. I would like to find objectively the most secure coin as I believe it will have the most longevity and price potential in the long run.
I thank you all for your insights
I really like the idea. I have poked around at Storj and they seem to really be onto something.
I just think holding an immutable record that data once existed for all of eternity, for data that is itself entirely ephemeral, seems like a bit of a waste.
Generally I find the innovator in the space is Filecoin at this time, its at least the most popular, where Storj has languished somewhat in interest both in development and price performance as a surrogate for general interest. Chia, however is a newly founded coin by Bram Cohen, who at least has a reputation to leverage for Chia in the near term if not a longer term better design/development than Filecoin.
I wrote this post mainly to gather insight for the new project “Chia.net” as a competitor to Filecoin. Bram Cohen is the lead dev for Chia and seems to have based a lot of his criticism on filecoin, which tells me he sees it as his main competitor and is designing his project to out compete it. While Bram is the creator of bittorrent, I don’t care to appeal to his authority or his bias for his own project, what I am interested in is sussing out the best crypto in this data storage hardware mining niche and pointing my idle hardware towards it. In the very near term I intend on capitalizing on potential pumping of the Chia price for at least the purpose of getting more hardware to pay for itself. Chia has just recently been released and I will be speculating on its early adoption phase, then focusing on the coin that shows the most merit for monetizing decentralized storage space.
Its hard to comment on the general real world functional security of each project. However major protocols rely on similar consensus strategies:
Both Filecoin and Chia rely on some form of Proof of Storage to prove that data storage providers have the data they claim to.
Filecoin has a major difference such that it also incorporates a form of Proof of Replication such that the protocol verifies that the storage provider is storing up to n distinct replicas of the data. As far as I can tell, this occurs at the same storage provider.
I’m coming around to the idea of letting my computer hum along playing FarmVille Bingo with itself, and maybe making some money…
its not a bad idea as long as you take some profits early. Chia itself will only be mine-able for the next few weeks, and I think this style of release will cause a lot of FOMO from non-mining traders waiting on the wings to jump in once trading goes live. If i calculate based on the market cap of file coin and the market supply of chia, I am guessing the unit price per chia will be 400-500 dollars. I expect at least a speculative spike to this price before market forces balance out, I have other reasoning involved. That being said and if I am correct that will mean successfully winning one block reward (2 chia at this time) would be enough to justify a lot of scenarios. Frankly when it comes to crypto speculation I could easily see traders over shooting that price target.
I will again note, I have no loyalty to chia, I am just trying to make a nice entry point and then switch over to objectively the leading coin in the niche of storage space based mining. I would love if any in this forum can speak to the technical merits or lack there of for chia versus other coins in the niche.
be prepared to burn up at least 400gb of SSD space if you want any chance of winning a chia block reward.
Seems odd, the “beginner” page suggests something like 1 person a day will unlock a coin, but there is also like a 1 in 4000 chance to complete/win.
I presume the fuzz the numbers as just a demonstration/guide, but it just seems odd to read, as the chances should scale; if chances are 1 in 4000, then there should be more and more unlocked each day, or the odds should go down with more competition.
Probably just simplified
Yes, Id say they just used easy numbers for demonstration purposes. From what I see the relative mining difficulty has already exceeded the FAQ numbers. It still doesn’t seem like a bad chance though as long as you can get whole plots per day at least. Yes the odds are going down, about 200 PB have come onto the network since the op of this thread. The GUI has live stats that will tell you your chances of winning at any juncture, to be honest they did a fair job on designing the original wallet software compared to many other coins with their own blockchains that Ive seen. I’m fairly convinced that the user friendliness and the name branding of “BRAM COHEN!” plastered in press releases bodes well for trader FOMO incoming. what may really seal the deal is that they have positioned it as a a smart contract platform with colored coins implementation. If chia manages to attract projects I think file coin stands no chance against it
It mentions quickly creating “plots” of 101GiB, using SSD to quickly grow out the plot to size, then flush it to slower HDD, that’s the impression I got.
there are two steps to mining chia (farming). to make the “plots” and then to farm them. Making plots is significantly faster with SSD, but you dont need the speed for farming so your plots transferred later to a HDD is ideal. You need about 100gb final storage space for a plot and a bit over 200gb more as temporary space during the plotting, so about 300gb in total to make a plot, where 200 is reclaimed if the plot is made successfully
I started farming Chia 2.5 weeks ago and would consider myself slightly late to it. Currently at 22TiB of plots and have won two blocks. Estimated time to win is 6-7 days but there is no way I will be able to keep up with netspace growth.
yes, it seems making plots will be pretty crucial to winning rewards, difficulty is growing fairly fast
I tried to write a dockerfile for Chia:
# run updates
RUN apt-get update && apt-get upgrade -y && apt-get install git screen nano lsb-release sudo -y
# checkout and install chia
RUN git clone https://github.com/Chia-Network/chia-blockchain.git -b latest --recurse-submodules
RUN chmod +x ./install.sh && sh install.sh
# create mountpoints for ssd temporary storage and persistend hdd storage
RUN mkdir /ssd && mkdir /hdd
# copy over ".chia" folder if it exists in the build folder
COPY .chia /root/.chia
# manually activate virtual env (replaces the ". ./activate" command)
RUN python3 -m venv $VIRTUAL_ENV
# run Chia init
RUN chia init
You can build it by running:
sudo docker build -t chia-docker .
You can login to the container by running:
sudo docker run -it chia-docker
Activate the Python environment in the container:
Trying to run the daemon or any command dependent on the daemon:
Unfortunately fails with:
OSError: [Errno 99] error while attempting to bind on address ('::1', 55400, 0, 0): cannot assign requested address
There is not much information why this error occurs, does anyone here have an idea?
the best i can do is mention that their dev community talks on keybase @chia_network.public
that’s the general channel, there are several daughter channels
It’s trying to bind on the ip6 loopback address
for anyone thinking about being “too late” to mine chia, i believe that it will be possible to still take advantage of at least the early days of mining if farming pools come about to regularize the chances for block rewards for sharing among said pool. There are already talks of some of the chia devs themselves working to that end. The key would be doing it in a verifiably decentralized way where all of us degenerates don’t need to trust each other. It will have been advantageous for smaller farmers to have plots made before pools become prevalent.
To quote my thesis on the potential unit price of Chia where 400-500 USD per unit represents Chia with a market cap equal to Filecoin’s, I would also add that the vast majority (21mm chia) currently existing is part of the chia “prefarm.” Depending on those holders’ behavior the price could be far higher than 400-500 USD due to lack of liquidity in future exchange markets. To explain further, several of the top cryptos have such token distributions weighted heavily to devs and dev adjacent holders in the outset of their release, and often among those other coins the original distributions go several quarters or even years before significant sell pressure from those holders hits the markets. If Chia is in fact not a confidence scam, it is largely in the best interests of those prefarm holders to not dump on the market for as long as they can bear. This is why I’m so interested in sussing out the technical security and competitive advantages of Chia as a measure of prefarm devs’ intentions to maintain a long term project. The reason i mention this is to justify the smaller but more steady rewards of pooled mining.
That all being said, caveat emptor, not financial advice, just my opinion of the market based on past experience in crypto. Personally I will be recouping my initial speculative investment early and letting the rest ride or converting it to the storage coin that leads the niche.
my approach is actually not to use any ssds, cause i dont want degraded hardware by the end of my foray into chia. Rather than one beefy plotter, I have made many slow plotters. I suppose you could do this in many ways, I did it in the form of having several rasp pis with their own partitions on large HDDs (no more than 2 pis with their own partitions on a single HDDs). I figure if i dont succeed in making a profit at least I will have pis to play with later.
that being said I have not been terribly successful in making plots but that is for unrelated reasons like my current wifi being shoddy and interrupting what im doing. if i had not had this wifi problem i figure my rate is about 4 plots per36-48 hours.
I think I will have paid for my costs if I end up getting just 1 block reward, relevant to my chia unit price thesis
i would have certainly bought ssds from my local e-recycler if i would have gone that route.
I see thanks for your perspective. One thing I could not figure out how long are plots supposed to last until I need to write a new one. Do you have any links that explain this?
I have been told;
“When the time to create a k32 becomes too short the minimum k size will increase, but k32 is considered safe for 10+ years”
others have said no less than 5 years
I have a used Intel P3700 that only had 184TBW out of 7.3PBW when I got it but also a lot of other SSDs will last a lot longer than the rated endurance. Samsung 970 Pros are MLC drives and near identical to their enterprise couterparts, Samsung just wants to upsell you.
NVMe drives are worth it if you want plots fast because each parallel plotting queue can pump out 3-4 plots per day. A single 2TB drive can do 6, maybe 7 plots simultaneously provided you have the CPU and RAM for it.
Plots don’t expire except when the plot K size can be created faster than a block is constructed. That is to say that you could create a plot that specifically solves/wins the block within the block construction time frame. Generally the Chia team retires the K size ahead of that time so that type of attack could never occur, which is why K31 plots were retired despite not meeting that threshold yet.