Clément Elbaz [ARCHIVE] on Nostr: 📅 Original date posted:2014-01-08 📝 Original message:Hello all, It seems there ...
đź“… Original date posted:2014-01-08
đź“ť Original message:Hello all,
It seems there was a problem with my first email (thank you Mark for the
heads up), so I'll copy paste it there :
-----------------------
Hello all,
I'm designing a program that needs some metrics computed from the Bitcoin
block chain (some address balances, or the occurrence or not of a specific
transaction). The kind of infos you get from http://blockchain.info/,
provided you trust this website (my program do not).
My program should run on lightweight/embedded hardware. The execution
environment provides access to the Bitcoin network but not enough resources
to set up a trusted node along with my program. Also, my program trusts the
global Bitcoin network but no individual node.
I would need a way to ask an untrusted Bitcoin node to compute some 'metric
request' on my behalf and having the result of that metric request
validated by the network.
Is there any available or work-in-progress projects that would come close
to this need ? Or should I do it myself ? :-)
Thank you all,
Clément Elbaz
-----------------------
On Wed, Jan 8, 2014 at 8:44 PM, Clément Elbaz <clem.ds at gmail.com> wrote:
> Some more thoughts :
>
> If no such project exist yet, I thought it could work with an alternate,
> small and fixed-length 'metric request block chain' of some sort.
>
> It would temporarily stores structures defined as [metric request |
> current block number when request was made | hash of the response] instead
> of financial transactions.
>
> These structures are verifiable so it could work the same way as a regular
> financial blochchain.
>
> It should not be part of the main Bitcoin protocol but could be a plugin
> interacting with the data managed by the fullnode bitcoin software.
>
> Also, metrics requests can be expensive to compute and validate, so it
> would make sense to pay a fee everytime you ask one.
>
> Does any of this makes any sense to you ?
>
> Thanks,
>
> Clément
>
--
Clément ELBAZ
06. 09. 55. 78. 41
clem.ds at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxfoundation.org/pipermail/bitcoin-dev/attachments/20140108/6dc30995/attachment.html>;
đź“ť Original message:Hello all,
It seems there was a problem with my first email (thank you Mark for the
heads up), so I'll copy paste it there :
-----------------------
Hello all,
I'm designing a program that needs some metrics computed from the Bitcoin
block chain (some address balances, or the occurrence or not of a specific
transaction). The kind of infos you get from http://blockchain.info/,
provided you trust this website (my program do not).
My program should run on lightweight/embedded hardware. The execution
environment provides access to the Bitcoin network but not enough resources
to set up a trusted node along with my program. Also, my program trusts the
global Bitcoin network but no individual node.
I would need a way to ask an untrusted Bitcoin node to compute some 'metric
request' on my behalf and having the result of that metric request
validated by the network.
Is there any available or work-in-progress projects that would come close
to this need ? Or should I do it myself ? :-)
Thank you all,
Clément Elbaz
-----------------------
On Wed, Jan 8, 2014 at 8:44 PM, Clément Elbaz <clem.ds at gmail.com> wrote:
> Some more thoughts :
>
> If no such project exist yet, I thought it could work with an alternate,
> small and fixed-length 'metric request block chain' of some sort.
>
> It would temporarily stores structures defined as [metric request |
> current block number when request was made | hash of the response] instead
> of financial transactions.
>
> These structures are verifiable so it could work the same way as a regular
> financial blochchain.
>
> It should not be part of the main Bitcoin protocol but could be a plugin
> interacting with the data managed by the fullnode bitcoin software.
>
> Also, metrics requests can be expensive to compute and validate, so it
> would make sense to pay a fee everytime you ask one.
>
> Does any of this makes any sense to you ?
>
> Thanks,
>
> Clément
>
--
Clément ELBAZ
06. 09. 55. 78. 41
clem.ds at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxfoundation.org/pipermail/bitcoin-dev/attachments/20140108/6dc30995/attachment.html>;