Storj vs Filecoin?

Can someone explain shortly why We are better than Filecoin?
if We are better, then we should take some opportunnity and show it, because filecoin is getting some huge hype right now, and i wonder if rightfully? (if for example storj have better solution)
Is theres significant storj advantage?

1 Like

The Filecoin network utilizes IPFS as its back-end storage. As far as I know, there’s no default encryption included in storing files on the network. Furthermore, the files are stored as a complete unit on a single IPFS backend node.

Storj encrypts the files and then breaks apart the file into pieces before uploading. The pieces of the file are then distributed spatially across many Storage Nodes in IPv4 space.

So, in Filecoin… someone has your complete file… which is not encrypted by default.

In Storj… no one has your complete file… and the files are all encrypted by default.

However, I may not have the most recent information on the most recent Filecoin network architecture.

7 Likes

Storj has Client-side encryption (E2EE by default w/ macaroon (edge-based) IAM). FC not encrypted by default.

Storj, Data is erasure coded and spread across 80+ uncorrelated endpoints with different owners/geo’s. Filecoin, pay per one replication.

Pricing is static at half single region AWS (with global availability) and can pay in fiat (Important for enterprise today). Filecoin is dynamic per contract and payment only in crypto.

On Filecoin, you have to pay per replication, and negotiate with each node. This means much more complexity for devs. This also means a higher expansion factor, see:

Pricing: Object storage on Storj (with distribution across 80+ endpoints) is currently cheaper than a single endpoint on Filecoin ($0.31/GB $0.10 on Storj at time of writing)

11 Likes

The two are very different. Storj is more of a turnkey solution that is relatively mature and easy for a new miner to participate with low costs. FileCoin is still fresh and changing, but requires far more intervention as a miner to figure out how to effectively mine. Despite the recent launch of mainnet, prices and ROI are not understood. (Filecoin launched with a spike in trading at more than $200 now settling to ~$40 each.)
The network highly depends on people developing applications to interface with the lotus backend. Miners are expected to put up a significant investment in hardware and collateral, starting in the thousands of dollars and in many cases reaching millions. The big miners are heavily invested and reap most of the block mining rewards.

6 Likes

Wow, @stuberman, somebody make a yt video about that! All those people out there makes videos now about FC and thinks its nodes are for a small guy LOL

1 Like

There are plenty already on YouTube - here this one shows three different rigs

holy cow, those rigs look super complicated at least from my point of view. HUGE investment for the hardware and of course there’s then the collateral side of it.

I have to say that without a doubt, Storj has the advantage over Filecoin on both fronts.

5 Likes

yeah i was looking at filecoin a couple of days ago because i kinda had the misconception that i chance of running it… but turned out i would have better odd’s as a snokeflake in hell…
aside from that they have some concepts i really don’t like…

you can mine filecoin, and to be allowed to mine filecoin you need to store random data or maybe customer data, so when you lock down a block of data you are then granted the ability to participate in a “mining pool” and then whatever the mining pool successfully mines the miner will have a chance of winning depending on how many blocks of data he has locked.

while this is going on one also has to burn filecoin’s to prove ones node isn’t fake, basically it works something like you spend filecoin to prove you are real or something… might be sia tho… but i think it’s on both of them.

so the more free space you have then with filecoin you can fill it with random generated data, which then is verified periodically through some mathemagical wizardry and the more you store the more you are allowed to mine, out of the 1billion possible coins, of which there is only like 15million right now…

so filecoin starting at 200$ means it is or should be thought of as being a company valued at 200billion $ heheh and now at 40$ billion … sure one could go the other way as say only the coins that are mined has value… but there is basically an endless supply of coins that can be mined, if one only has the space and the computing power for both… because you cannot have one without the other and if you just store data for the network, somebody else in you mining pool with get the mining rights, thus you really want to have both things running to equal each other…

i cannot help but seeing this as a bit of a crypto version of a pyramid scheme, big datacenters will just move in a grind through the mining making any profits to be made from small setups basically impossible… also the exchanges are xchanging iou’s for filecoin… because in the future filecoin’s will be worth much less, but most people don’t seem to get that…

but its simply a matter of math… there are 15mil coins now accessible and eventually there will be 1billion… the whole thing that makes it cool is that you can use storage to give you access to mining valuable coins… but the coins are pointless, it’s really the storage that is valuable, and one just locks it up with random data one cannot delete without getting punished.

losing collateral…

filecoin seemed interesting until i started to dig into it… if anyone things filecoin is a good idea i would be very interested in hearing the reasons… because i just don’t see it… but maybe i’m just missing something… kinda stopped digging into it when i realized my 8 core, 48GB ram, 80TB and ½Gbit fiber internet wasn’t even remotely close in any aspects, infact my hardware is so old it doesn’t seem worth trying to do filecoin…

my cpu types are to old… but granted they are 10 years… my ram is to small… tho i could get to an acceptable level by putting 200+GB ram in it and then i need to add a graphics card which i could…
so i would need to replace my PSU, buy 12x16GB ram, OC my cpu’s or replace them, and ofc get atleast 1 graphics card if not more than one to be able to actually mine anything close to what my storage would unlock for me.

minimum graphics cards for mining is like 1080, but i doubt anything less than a 3080 or so would be an efficient choice today… and maybe a couple is really what one wants… because in the end it becomes how many watts you spend to do the computation and 3080 is :smiley: good at computation.

not sure i would ever buy a 1080 tho… i bet tensor cores are great at mining… so would have to be 20 series+ at the very least… ofc maybe the amd cards could be for consideration here… even tho i’m not really much of an amd fan…

but difficult to not see that they do seem to have taken the cpu crown for the next couple of years atleast, their graphics cards are still crap for gaming… but from what i hear fairly okay for mining…

i would say filecoin… mostly because i would have wanted to participate, atleast to test it out… but i cannot… :confused: for me it’s not even worth trying to make my server able to run filecoin

4 Likes

Do Filecoin infrastructure make it completely decentralized ?

yeah i think so… because it doesn’t use satellites but blockchain … but there are certain disadvantages to that model also… like say if you want to change your ip address, then you need to pay a fee for changing the blockchain, like if you made a crypto coin transaction.

so basically all information like that will need to go through the blockchain, even if the data isn’t stored directly on it.

Have you tried to store something on “full decentralized” solutions available today?
Do you have an access from smartphone or from any other device in the same time? From work office, for example?
How much time you will need to restore your data from there on the clean hardware?

In my opinion - the tech is a nice toy, but if it is not practically usable, why is it needed?
Seems only to mine coins and cut some profit until it is gone…

3 Likes

No, there is no full magic decentralized storage solution. That wasn’t a critic about Storj, i think your model is superior in terms of usability and overall performance. It was just a simple question to try to see the big picture. I addition, to me, a so called decentralized infrastructure like Filecoin, loses instantly this label if only big actors can participte in the network because datacenter grade hardware is mandatory.

1 Like

Yes, I do not mean that model is bad and I do not want to compare. I mean that it’s not usable today. The “full decentralization” is a buzzword like blockchain.
Everything should have a practical usage, otherwise it’s just spent resources and time.
I was excited about “full decentralization”, but as turned out - it’s not practical usable, but idea is nice.
So, today the “full decentralization” is practically used only in blockchains, but their purposes and reasons for use are questionable, except few.

However, even there, as soon as come a big players or interested countries, the decentralization is failing too fast - the majority of consensus start to belongs to a few players…

In case of filecoin seems it’s embedded into model, at least now. Only big players can participate.

2 Likes

Well, i just tried Filecoin as a customer. Uploaded a 700MB file (bigger ones failed) I had an average upload speed of about 0.4MB per second. My FTTH internet is capable of 40MB/sec upload.

2 Likes

thats terrible, even for a test network…

but to be fair to filecoin as i understand it, they just started it recently and decided to test everything while developing it… which is why their model is like it is… with the whole mining deal.

they hope that people will keep wanting to mine worthless coins so they can keep their operation running… so if people just quit they will most likely fail because their entire development model is based on an endless supply of people to test on, which the crypto mining gives them…

so they could very quickly run into some serious issues, especially if thats the speeds they can muster, then on top of that the whole 15k $ hardware and collateral required for entry…
which might be working for free if the coins won’t become popular

and all the allocated space will be used for random junk data created just for that, using internet bandwidth + storage capacity… it’s just such a bad model i cannot believe it

i must be misunderstanding something

Shame on me, i did not test on tardigrade.io to compare. What is to be expected right now speedwise ?

i do think its a main network, no ?

1 Like

no clue i just storj xD
there have been some going back and forth about that on the forum from time to time, about bandwidth speeds… i’m not aware of the actual speeds people get on avg… but from what i understand people with like high bandwidth have tested and gotten mixed results… like getting 200mbit or 100mbit on a 500mbit connection…

ofc the geolocation factor also comes very much into play… like if you are sitting on a local high speed network in Siberia doesn’t mean one should expect full local internet bandwidth to the rest of the world and thus stuff like storj wouldn’t see full bandwidth either…

i’m unaware of any comprehensive testing which is published about storj speeds, but i would be very interested to look at the data if it’s out there somewhere… i’m sure there are lots that have done tests… but i mean proper tests… multiple systems, multiple locations, not being limited by geographical internet limitations… you know… seeing what the network will actually do when there is open lines and normal latency

1 Like

I think you should test it on the same location and hardware to be comparable.

not to my understanding…but my knowledge on filecoin isn’t great so cannot really say 100% one way or another… it was how i understood it… mainnet is like what they are working towards or their name of the network… like storj’s is called tardigrade…

i think when they launched the spacewar recently, they had just created their testnet or whatever… if you want to know exactly you will have to dig into it… i doubt there is anything simple about it…

i could barely figure out how it was suppose to work from their descriptions…
mainnet sounds better than testnet, or there is multiple versions of their network, i think that was it actually… there are like multiple branches doing different developments or something… and then the mainnet is the unmodded network…

kinda like crypt coins branch, they seem to have done the same because some companies want to do something special with the tech or something… i duno…

it was a very confusing read, and when i found out i was like 10000$ out from even being able to host their storage solution on my server, there wasn’t much point for me to keep learning about their stuff