[Scipy-organizers] Publication and review in SciPy

[sheila miguez]

Hi all,

On Wed, Oct 30, 2013 at 11:17 PM, Jacob Barhak <jacob.barhak at gmail.com> wrote:
> Hi Sheila,
>
> You mention a few recent sources for publishing externally to SciPy. And you
> have the same idea as I have of newer publication types and you mention a
> few examples.
>
> If such a publication has an added value to github code+paper publication,
> then it makes sense to channel such an external instrument. An external
> instrument can offer the following advantages:

For the list, I only have experience to talk about the technology and
DOI depositing.

> - A larger/different base of subscribers for advertising
> - Recognition from funding organizations and traditional community
> - Improved publication technology
> - Access to a certain community of experts for reviewing purposes
> - Registration with certain archiving entities that gives the publication an
> ID such as DOI or PMID

We can also package up materials to store on figshare as a simple way
to get DOIs.

To become a depositor, I think it may be $275/yr, but we would want to
double check with crossref.

http://crossref.org/02publishers/20pub_fees.html

> - Advanced technology to store the paper/code

I think that github is sufficiently advanced to store the materials if
we are not storing more than 2 GB. If we do want to store more than
2GB we should talk to someone at github.

Will we get papers with large datasets? I think we could ask for
sponsors to cover the costs of depositing, or provide infrastructure
for us.

> If neither of these or similar advantages are present in the external
> publisher then sticking with github publication model seems a good idea.

I like planning for the simplest solution, which to me seems to be
github for storing everything including the src for the web page that
would publish the materials online. This is something we can talk
about to see if people agree.

> Perhaps an external source is not necessary for SciPy - perhaps other
> advertising methods should work better.

> Never the less, if you consider an external publishing partner, I believe a
> test is a good idea. And the sooner the better. I agree with Sheila that
> last minute in bad.
>
> And note that Partnering for the sake of appearances may be detrimental
> rather than instrumental.

I am not familiar with prestige and appearance when it comes to this
domain, so I will let others talk about it. I tend to have more
ego-investment in techie stuff.

> And Sheila, thanks for supporting the github review process idea practiced
> in 2012.

I pointed someone from github at this thread, and he wanted to know if
there was anything github could do to support this type of activity.
Do you have anything to suggest?



-- 
sheila at codersquid.com