[science-iwg] DOIs for scientific software

[Martin Hammitzsch <martin.hammitzsch@xxxxxxxxx>]

Hi Everyone,

You may remember our discussion on minting DOIs to software releases /
packages hosted by the Science and LocationTech WGs:
- http://dev.eclipse.org/mhonarc/lists/science-iwg/msg00815.html
- http://dev.eclipse.org/mhonarc/lists/science-iwg/msg00819.html

Summarized I understand that many of us would be interested in a
concept how the Eclipse Foundation and Community could implement a
working solution.

So far we come up with a document draft collecting general ideas,
thoughts and questions:
https://docs.google.com/document/d/10GNuXtjJNOJZ9AVzfeZP_cybYILjsh6KQ6f-htQH190/edit?usp=sharing

Another document summarized how and why the Eclipse Foundation could
make use of DataCite DOIs and which steps could be the next ones to
address the details:
https://docs.google.com/document/d/1WoQxo8C81S62CgVWjTVqLwpuB8WuQKYMJnd6PRHkBeE/edit?usp=sharing

I suggest that one of the next steps should be a meeting in Ottawa
with the National Science Library to present the idea, to discuss
options, to touch contractual matters, and to get support for drafting
a detailed concept.

Please let us know what you think.

Best,
Martin


On 18 April 2015 at 12:01, Philip Wenig <philip.wenig@xxxxxxxxxxxxx> wrote:
> Hi folks,
>
> we will work out a draft how this could be implemented the next weeks. It
> will figure out the scope, the benefits for the Science and LocationTech WGs
> as well as the costs and obligations it would have.
>
> We also try to cover all mentioned thoughts and criticism as well as the
> upcoming ones.
>
>
> Best,
> Philip
>
>
> Am 17.04.2015 um 23:11 schrieb Martin Hammitzsch:
>
> Hi Andrew,
>
> Regarding the metadata I agree. It's something we might use as a first
> starting point together with a related landing page that more or less shows
> the metadata in a convenient way plus additional information such as metrics
> and some interactive elements like links or a web based code repository
> browser.
>
> Unfortunately, the DOI thing isn't that easy. Either Eclipse becomes a DOI
> minting agency itself or an existing partnership with a partner who already
> is a registered agency supports this idea. Thus the DOI infrastructure
> provided by the partner already exists, is maintained, and finally can be
> used for minting DOIs, for maintaining the metadata, and for properly
> redirecting the DOI to the landing page.
>
> Furthermore, an agency is a member of the International DOI Foundation
> (IDF), as far as I know, and has to commit to certain regulations, e.g. if
> an agency shuts down services than the other member/agencies have to sort
> out how DOI labeled digital objects are taken over by the other members. So
> if the Eclipse Foundation wants to mint DOIs itself it finally means the
> foundation is part of a kind of social commitment and has to consider
> further implications. So having the opportunity to get support by a partner
> who already takes care of all of this in the role of a DOI minting agency
> would possibly be highly beneficial in a first phase.
>
> And last but not least, the IDF membership and the minting of DOIs costs
> money, as far as I know.
>
> So the one thing is the technical implementation and the set-up of some
> additional processes that possibly can be put on top as a kind of simple
> layer realising the idea without touching what already exists and is working
> for the WGs.
>
> The other thing is the DOI related stuff, that should be well thought out
> since this is another world we depend on if we want to assign DOIs to
> software releases. Unfortunately, this is the currency in sciences. However,
> we may also consider the use of the Handle system at first, which is the
> foundation of the DOI system. As far as I know, its no problem to make use
> of the Handle system at no cost. This would allow a first proof of concept.
>
> Just some further thoughts ... I am not an expert in this library and
> publisher driven DOI world.
>
> Best,
> Martin
>
> On 17 Apr 2015 21:58, "Andrew Ross" <andrew.ross@xxxxxxxxxxx> wrote:
>>
>> Dear Philip, Martin, and All
>>
>> I like the idea FWIW. And it sounds quite reasonable to implement.
>>
>> Part of the meta data could be a git commit and direct link to the code
>> repository.
>>
>> I suspect the team/community could determine the meta data fields without
>> too much trouble.
>>
>> How would the DOI be generated? I'm wondering if this is something we can
>> automate and simply give to all releases and maybe even all builds.
>>
>> Kind regards,
>>
>> Andrew
>>
>> On 16/04/15 08:04, Philip Wenig wrote:
>>>
>>> Hi folks,
>>>
>>> Martin gave an excellent talk about the publication of software at the
>>> EGU meeting yesterday. Scientists get reward for papers but not for
>>> software. Papers are cited, software not. That's a drawback for a scientific
>>> career. As we know, software has become crucial in scientific research
>>> nowadays. Hence, I'd like to share some of Martin's ideas with you:
>>>
>>> * Certain Science/LocationTech WG releases/packages get a DOI
>>> * The DOI points to a landing page with appropriate metadata to cite the
>>> release/package
>>> * The release/package can be cited and will be ranked on Google Scholar
>>> and other systems
>>>
>>> = This could be an incentive for more scientists to publish their
>>> software under the umbrella of the Science and LocationTech WG.
>>>
>>> What's your opinion?
>>>
>>>
>>> Best,
>>> Philip
>>>
>>> PDF of Martin's presentation:
>>>
>>> https://docs.google.com/presentation/d/1tulhwtEK9l1yGaAX1YWsMR7OrQAwWDee4N2M1Yy7bIo/edit?pli=1#slide=id.gada00e2fc_0_866