Testing with blockchain

[Thomas Sødring]

On 02/07/2017 02:32 PM, Petter Reinholdtsen wrote:
> [Thomas Sødring]
>> I have an open mind about this. I'm not convinced it's necessary, but
>> it's relatively quick and cheap to show a proof-of-concept. A lot of
>> the discussion was about avoiding the hype-factor and trying to
>> identify if and where block-chain is relevant.
> Is it clear what kind of problem we are trying to solve here?
No, I  don't think so. There is a push in the academic record keeping
world to take a look at  block-chain as integrity mechanism. I think
some of this push comes from countries where there is not much trust in
government so you might want for example want a public record showing
the transfer of properties.

So I think people are asking in Norway whether or not we also should be
looking at this. Some people have clear thoughts on how it should be
used and some are just curious.
> The trusted timestamp mechanism can be used to prove for those with
> access to the archive that some data (say the archived PDF) existed at a
> given point in time and have not been modified by anyone without the
> secret key of the timestamp service.  In other words, it can prove the
> archive is unmodified to anyone with the public key of the timestamp
> service.
>
> What you talk about here seem to be solving a different problem:
Yes, probably! My impression from the meeting was that we were a little
unsure how to proceed and I offered to take a look at creating block
chains based on events in the Noark core. I will demonstrate this and
see if it leads to further discussion /investigation.
>
>> What I like about this is that as soon as a document is registered in
>> a core, a record will be available in a distributed fashion. Sneaky
>> people can't pretend the document isn't there. So maybe I can register
>> a document before it hits OEP, even though I may not be able to
>> publish title.
>>
>> The downside is that let's say a Child Protection Agency document with
>> some metadata is published on a block chain it will probably be
>> impossible to erase an accidental publication. I awlays find it's
>> easier to discuss things when there's a practical implementation.
> This seem to discuss proving that some document actually made it into
> the archive.  Is that a problem worth solving?  If so, why?
To provide a greater degree of transparency of government. As it stands
today you have to trust the Noark system to trust government. Perhaps
with block-chain, a new level of trust can be established.
 
>
>> I read this before, it's interesting and definitely something I will
>> consider. If block-chain can't give anything more than trusted
>> timestamping, then following it is just hype!
> I am quite sure it isn't 'just hype', as it has many very useful
> applications.  But I do not quite understand which problem it would
> solve for a public archive.
>

Block-chain definitely isn't hype, but record-keeping and block-chain
may be hype. One of the things we discussed is to avoid chasing a
technology that isn't relevant. The problem that was brought up that
block chain may solve could for example be diploma verification. I
believe MIT do this. So apparently MIT create a hash of your diploma and
add it to a block-chain and then anyone can verify it without having to
contact MIT. A similar example is that a lot of electronic documents in
Norway are signed digitally, not with a private PKI key. Trondheim
kommune for example had to develop a special physical stamp that they
use when such documents are required to be proved authentic. The
document is printed and stamped and signed. Perhaps block-chain can help
here. So it's more about looking for other ways to prove authenticity
for electronic material.


 - Tom