Sharing the data

[Giancarlo Macchi]

Dear Benajamin and Stefano,
Thank you very much.
Well, i think that one solution would be the following procedure.

a) Create a single compressed file.
b) Include all the necessary metadata
c) Name it with a complex, long enough pseudo-random name
d) Upload it in as many servers you can. Those suggested by Stefano
    both *seem* *quite promising*
e) Provide the name along with the checksum in your paper or
    electronic publication, in order to allow to the reader to know
    that this is exactly the file generated by the author of the article
f) Pray the best for webcrawlers

What do you think?

  Giancarlo



Stefano Costa ha scritto:
> Il giorno mer, 21/06/2006 alle 09.50 +0200, Giancarlo Macchi ha scritto:
>   
>> Dear Friends,
>> I just want to ask you some thing related with the topic of 
>> data sharing.
>>
>> In this moment I'm finishing an Landscape Archaeology article
>> where I made some assumptions on the basis of point distribution
>> maps. As Roberto correctly stated in one of his previous posts
>> the data "should be available" to the reader in order to validate
>> or refute my assumptions.
>>
>> So the question is very simple: is there a secure place to
>> publish such files. I mean, even if my lab today run a web server
>> and other services, I don't know if 20 years from now there will
>> be an archaeology department, I will be still an archaeologist,
>> the University of Siena will be interested on keeping my old
>> files in some server etc etc etc. 
>>
>> Do you know about the existence of institutions and repositories
>> for such purposes?
>>     
>
> Giancarlo,
> I think that http://www.opencontext.org could be a suggested place for
> some of such purposes. The Alexandria Archive Initiative works towards
> open archaeological archives to be published on the web (which could
> however have a slightly different meaning from "share the data").
>
> Another interesting project can be found here http://edna.itor.org/en/
> and it comes from the Netherlands, and this is more linked to "data
> preservation".
>
> Maybe there are others, we could try to collect links to all of them and
> evaluate/discuss which are the best-fitting.
>
> But I'll try to give you a more "scientific" answer as well. I guess
> it's quite different from Benjamin's previous post, but I hope we can
> have some good discussion about this one.
>
> If you are publishing your data as freely available on the WWW (free,
> not just gratis..), everyone with an Internet connection can search
> through them, browse, download, re-use, re-distribute. You see, it's
> much like free software.
>
> If we assume that datasets have their own history, then let this history
> be one of spreading, being used (and perhaps abused) and spread again.
> This is the very idea of scholarly literature meant as a medium for
> exchange of information, ideas - and data, of course.
>
> When your dataset will be erased from Siena's server 20 years from now
> (or maybe even earlier - where do 20-years-ago pre/proto-digital
> datasets have gone?), it will depend on the spreading of those data
> whether they are lost at all or they have at some extent survived in
> someone else's archives. It is obvious that if in these 20 years nobody
> makes a copy, then those data actually have little or no interest at all
> for the scientific community (and I guess this is not your case).
>
> Over long time-spans, data and information are not well-preserved by
> choosing a potentially everlasting support. Archaeology teaches us that
> everlasting objects do not exist. On the opposite, just like music
> recordings and ancient books, it's the continuous re-encoding that
> guarantees preservation. We have already had some discussion about
> digital file formats, you see as well that different topics are quite
> linked each other (though I suppose it's better to keep them separated
> not to generate confusion).
>
> Briefly, I'm supporting the idea that
>
>  SHARING allows (among the others) PRESERVATION
>
> opposite to the idea that
>
>  PRESERVATION is necessary for SHARING
>
> What do you think about this?
>
> Best regards,
> Stefano
>
>