How to find more project members?

[Thomas Sødring]

On 02/07/2017 11:22 PM, Petter Reinholdtsen wrote:
> [Thomas Sødring]
>> Unfortunately, none of it worked!
> That is not quite true.  Some of it worked.  After the dots presentation
> at NUUG in 2015[1], I was ready to get involved, but shortly after I got
> access to the code it was yanked from the public and my motivation to
> help develop dots floored.

Fair enough - agree with that! But I was referring to the efforts with
the archive community and trying to get the involved.

>
>  [1] <URL: https://www.nuug.no/aktiviteter/20140211-noark/ >
>
>> I do not think we will find any active help within the archive
>> community.
> Neither do I, but I hope we can find some within the free software
> community among the people assisting the archive community. :)
>
> But I really hope we can get some help from the National Archive with
> certification, both to interpret and clearify the specification and to
> verify that our implementation is acceptable.

The approval process for Noark 5 is a bit of a joke! It simply does not
serve its purpose. It's a long explanation as to why, but the NA will
either have to let us get temporary approval for Noark 5v4 or Noark 5v3.
If the NA hasn't finished the interface I can't see them stopping us
getting temporary approval and allowing us to submit for updated approval.

Personally, I'd like to see help coming in giving feedback on the code
structure, how to make it run quicker, reduce garbage collection by the
jvm, how to scale it etc. I wonder if the current version can even
encounter a 409 Conflict as it's single threaded. We also need to make a
conversion process on top of libreoffice for conversion to archive
format. These kind of tasks could be handled by others.

>> But no-one can lock this core down!
> I appreciate this, given my previous experience.
>
>> I think easiest way to build a community for this variant of a Noark
>> core is to find Java heads. There are lectures in SOA and springsource
>> at UiO, so there might some relevant people there.  Simula and HiOA
>> are in talks about establishing a "centre for digitalisation" for the
>> public sector. I'd be happy to see nikita there and that might be a
>> possible avenue to explore. Uninett and USIT might have people that
>> are interested, but they have enough with their own projects.
> Would you be interested in presenting the project at a meeting at USIT?
> I might be able to organize it.
I would be happy to. I presented the project a few years ago at UiO in a
meeting between the administration and some people from USIT. But what
we have now is much more concrete!
>
>> However, in some ways there is not really that much left to do and
>> maybe it will be easier to attract a community once the "product"
>> exists and is approved by the National Archive.
> I believe we need to get it to a point where it can do something useful
> before most people will be interested in joining.  Once it is useful,
> the need for improvements to make it more useful will arise. :)
>
> I am not sure we need to get as far as having approval from the National
> Archive.

>> I personally think the easiest way is to hit a JUG. Find a meeting and
>> present and see if something comes out of it.
> Could be.  I also suspect a weekly blog message syndicated on
> planet.nuug.no would help raise the visibility of the project.  Would
> you be willing to write about the progress in a weekly blog message?  I
> can add the RSS to planet.nuug.no.

We could try there, but I have had a blog for the project.

https://blogg.hioa.no/arkivetsrolle

As you can see I'm terrible at updating it. The reason is I'd rather
spend the hour or two I'd spend writing blog posts on coding. I
understand the need to profile and show development is happening though!
I need to think about this a bit more.

>
>> But your interest in the project Petter has really helped drive it
>> forward. For a long time now it has just been me. Just having someone
>> to discuss, someone to question what's happening gives the project
>> more meaning. You have highlighted so many issues that I probably
>> would have easily missed.
> I'm happy to assist, and look forward to when I can import and search
> the content.  And I would love to analyze the metadata across several
> institutions to try to locate documents that are only listed in one of
> them. :)
That would be cool and in fact something I think we could write a paper
about!  We are lucky to have something like OEP and it's time that we
started showing how it can be reused. Maybe something we can collaborate
on later in the year?
>
>> There are others, but they are not relevant at the moment. There's
>> definitely little understanding of open source and the editors of the
>> relevant archival publications will be weary of giving us what they
>> will deem as free advertisement space. Some vendors have referred to
>> the project as competition, and I think that label has stuck! But I
>> could for example publish information about the nikita-core and block
>> chain and that will probably be OK as I'm not trying to push the
>> nikita-core.
> Sad to hear this.  Btw, which other NOARK 5 core implementation exist?
> Who are our "competitors"?  And which client implementations exist?  Who
> are our potential users?
>

You can take a look at the following links:

Final approval:
http://arkivverket.no/arkivverket/Offentleg-forvalting/Noark/Noark-5/Godkjenning-av-Noark5-systemer/Endelig-godkjente-loesninger

Temporary approval:
http://arkivverket.no/arkivverket/Offentleg-forvalting/Noark/Noark-5/Godkjenning-av-Noark5-systemer/Midlertidig-godkjente-loesninger

Make a note when some vendors got temporary approval and when they got
final approval! The final approval really is just about bringing real
data to the national archive and showing that real data can be
extracted. And once you have final approval, there is no requirement to
get reapproved for later versions.

My opinion and understanding from listening to vendor descriptions is
that the big 3, SI, Evry and Acos got their temporary approval based on
Noark 4 systems upgraded to Noark 5. Take that with a grain of salt
though, but I was at a presentation by Gecko (Evry) and the lead
developer said this himself. A Noark 4 system is kinda a monolithic
record-keeping system, compared to Noark 5. dots and braArkiv I believe
are written from scratch as Noark 5 systems and then there are some
various systems built on top of other systems. What the big 3 have now,
I do not know. They say they have made big investments in the products.
>From feedback I am hearing from the archive community this seems to be
true and the systems are really good at creating extractions.

One company has recently said they will develop a Noark 5 open source
implementation. I am not sure they would want me naming them here.  I
think some of the existing vendors may also release as open source when
we get approval. I don't think anyone wants to "give away" their product
but if open source establishes itself, they will want to dominate that
part of the market.

I don't see nikita as a threat to any of the vendors. Just like Linux
lead to to innovation in smart telephones, I think nikita might lead to
innovation in the record keeping sector. Some municipalites will only
choose Microsoft and the  purchase process can be defined so that
experience tells so much that it's impossible for new comers and
difficult for open source as there may be requirements that lock things
to Microsoft technology.

Client implementations?

I don't think there exists any client implementations. As far as I know
people buy Noark as a stack, front end and back as one. edudots, that I
built on top of dots, split the front end from the back end and exists
as an open source front end project to Noark 5v2.

Who are our potential users?
Non-commercial entities mainly. Norsk Arkivråd have expressed an
interest using this as a course tool on Noark 5. It frees them any
single vendor. Researchers working on digitalisation of public sector.
No need to pay expensive vendors on an hourly basis, when they can play
with this themselves and move whatever they do to commercial Noark 5
cores. Fagsystem vendors might want this either under their products or
to test against. NTNU have a course on record keeping and archives as
well. They may want to integrate this into their teaching. I will use
this across all three years  of the bachelor program and will use the
interface in a course on metadata and interoperability. And another set
of users, being like you guys that take and use the core for something I
never anticipated!


 - Tom