Are there block explorers which respect your privacy?

[Pablo-wood]

If we run our own full node on bitcoin core the issues of selling our meta data to block explorers will no longer be a thing to worry about. Even Blockchair can't be vouched for 100%. Every non-decentralised system will always want to make profit from their services.

Mempool.space if self hosted aligns better with the idea of privacy and sovereignty. It's open-sourced with no third party dependencies

[dkbit98]

Can you link to some sources about that? No issue

What more sources do you need?
Just freaking open blockchair website and type any random bitcoin address, you will literally have confirmation for my words.
I am not going to recommend that browser to anyone, and one more reason is because they are also supporting a bunch of shitcoins.
Anyone can use whatever explorer they want, but for me there is nothing better than mempool.space.

[d5000]

Just freaking open blockchair website and type any random bitcoin address, you will literally have confirmation for my words.

Ah yeah, you're correct, sorry. I had only checked transactions and blocks at Blockchair, never addresses

Now okay, one could argue that if they respect their own Privacy Policy, they display third-party "risk scores" as a service to the user, without transmitting visitor information (IP adress, browser fingerprint etc.) to these third parties. But that's where the "vague points" in the Privacy notice I mentioned here -- that they only use personal data to improve their services -- could be a loophole to indeed transmit this information in real-time to the risk score analyzing companies, because their risk scores "improve" the service of Blockchair.

Even if they don't do that, at least it is a big temptation for the users to click on these risk score ads, which open the chain analysis companies' websites -- and I think if you do that with your own addresses, then that's the moment they'll have the visitor information they wanted.

I can totally understand your discomfort and that would be also a reason to not recommend Blockchair.

BTW, I'll add recommended block explorers to the OP from now on (only really convincing one until now is Mempool, Blockchair may be better than average but needs a warning).

And @all: I know of course that the best way is to host a full node and run a self-hosted block explorer. But that's not the topic of this thread, I'm searching for services with a "better than average" privacy.

[X-ray]

If you truly want privacy you should run your own full node and use its RPC and create the UI while you're at it or use existing open sourced frontend for these. Even I learned that RPC provider sometime selling our data to MEV bots to frontrun transaction in blockchain like ethereum, so I'm sure they are collecting data to some degree.

I personally have no problem using the public blockchain explorers out there but just know that running these sites require money and it's most likely coming from selling the data. Personally I don't really think a true privacy exist though.

[ranochigo]

I don't think that is meaningless. Of course you have to compare the previous block hash included in the first block of your chunk with several sources (e.g. at all block explorers you know if you're paranoid ). But if you know e.g. that you received a Bitcoin transaction one day before, but want to check the correct amount without having to fire up a SPV wallet (which potentially may expose you to Electrum server privacy attacks) then such a service may be quite neat as you can check all blocks of the last 24 hours approximately. You do all the analysis on your own device, without exposing the addresses you own, and without needing a full node.

You need to know the exact block hash to download, and having the block itself doesn't provide any form of security, which is why SPV wallets require at least the block headers. Having the block headers still doesn't give you any good information about the required transactions. This means that you must know which are the blocks to download, which means you also have to trust someone to tell you the exact blocks and download those. In essence, I'm describing an SPV wallet; its probably easier to use an SPV wallet and start querying random addresses to obfuscate your actual address.

Also if you want to do some onchain analysis of certain blocks on your own (e.g. to look for hints which addresses could belong to a single wallet) and you know approximately the timeframe where you're searching, that would also be an use case for such a service.

In the end, that kind of service could basically enable everything a normal block explorer can (even what walletexplorer.com does, i.e. search for "connected" addresses which may belong to the same wallet), only that you need slightly more resources on your device, but far away from the resources you need for a full node.

If you want to index all the transactions, you need a full node. If you only want to know a subset of them, you need to get your information from someone that is running a full node. Else, you would be randomly searching the blockchain, without any validation which opens you up to a host of eclipse and sybil attacks. There were several wallets which implemented BIP157 and BIP158 which supposedly provides more privacy than an SPV wallet.

[roemer]

I have not seen any debate about privacy in regards to using block explorers previously. I wonder how well in depth this data is actually being used in regards to the addresses on blacklists that crypto investigators are tracking/watching

[d5000]

You need to know the exact block hash to download, and having the block itself doesn't provide any form of security, which is why SPV wallets require at least the block headers.

In this post I wasn't looking for a replacement for a wallet, but for a block explorer - only with a little bit more privacy. If you use a block explorer, you also have to trust the provider of the service to deliver you the correct data. And of course the type of service I was thinking about should provide you all necessary data to check that, including blocks / block headers all the way down to the genesis block.

This means that you must know which are the blocks to download, which means you also have to trust someone to tell you the exact blocks and download those.

Ideally such a service should provide a web interface where you could enter the timeframe (e.g. you remember that you should have received a transaction between day X and day Y, so you would download all the blocks in-between). So the service provider doesn't have to know the exact blocks. Once the blocks are downloaded, you could then cut off the connection to the Internet and process the data offline via a JavaScript interface, e.g. searching for addresses and transactions. Or directly process it with command line tools or so.

The reason why this would be more privacy friendly than a SPV wallet is of course that on the SPV wallet the block data are on the server and you query addresses/transactions (which are then known by the server), while in the service I was thinking of, the server would only known the blocks you're requesting. Indeed technically it would be similar to BIP 157/158, only presented in another way and with more data analysis features.


I've looked a bit about privacy notices of some block explorers (five highest ranked on Google plus walletexplorer which I checked recently):

- httos://walletexplorer.com - shares all data with Chainalysis, basically no privacy
- https://blockstream.info - no separate privacy notice for the block explorer, according to Privacy Policy may share data with third parties, probably also bad privacy
- https://blockchain.com - Privacy Policy Section 7 says that they share data with a lot of different service providers including "providers of KYC or AML services". Probably bad privacy.
- https://btcscan.org - Privacy Policy (for all Redot services, no separate notice for block explorer) - mentions "compliance" and "research" companies as some of the parties they transmit data. Looks also bad for privacy.
- https://bitaps.com - Privacy policy doesn't mention that they transfer data to third parties, but also doesn't deny it. Quite unclear privacy.
- https://blockexplorer.one - Privacy Policy mentions GDPR (European data protection guideline), says that they don't transmit personal information to third parties, but they mention "Controllers" which could process data. Also a bit unclear.

May continue this list if there's interest.

[ranochigo]

In this post I wasn't looking for a replacement for a wallet, but for a block explorer - only with a little bit more privacy. If you use a block explorer, you also have to trust the provider of the service to deliver you the correct data. And of course the type of service I was thinking about should provide you all necessary data to check that, including blocks / block headers all the way down to the genesis block.

The actions that you describe mimics an SPV-type of implementation closely. Since you mention that you are looking for a service that provides you with all the blocks / block headers, then you can just enable txindex on your full node, which indexes the transactions and you can run a block explorer on top of that..

Ideally such a service should provide a web interface where you could enter the timeframe (e.g. you remember that you should have received a transaction between day X and day Y, so you would download all the blocks in-between). So the service provider doesn't have to know the exact blocks. Once the blocks are downloaded, you could then cut off the connection to the Internet and process the data offline via a JavaScript interface, e.g. searching for addresses and transactions. Or directly process it with command line tools or so.

That'll be exactly like how SPV works. You would be looking for a more resource-heavy SPV, which could work if there is a demand for it.

The reason why this would be more privacy friendly than a SPV wallet is of course that on the SPV wallet the block data are on the server and you query addresses/transactions (which are then known by the server), while in the service I was thinking of, the server would only known the blocks you're requesting. Indeed technically it would be similar to BIP 157/158, only presented in another way and with more data analysis features.

Yeah, I think generally the key issue here is resource tradeoffs. Indexing more blocks requires more resource, and it would probably start to look more like a full node after awhile. Anyhow, would it be better if we use a blockexplorer with Tor, but only query one address with one circuit? Since you are essentially linking each address to one circuit only, I assume it would leak practically no usable information?