Fwd: [Scipy-organizers] Publication and review in SciPy

30 views
Skip to first unread message

Anthony Scopatz

unread,
Oct 29, 2013, 12:31:40 PM10/29/13
to numf...@googlegroups.com
Inspirational post by Andy on the scipy-organizers list.  

Basically, we want to continue to revamp our the way our 
proceedings are done, but frankly, we need more people to 
decide that this is a way that they can give back to the 
community.

Be Well
Anthony

---------- Forwarded message ----------
From: Andy Ray Terrel <andy....@gmail.com>
Date: Tue, Oct 29, 2013 at 10:46 AM
Subject: Re: [Scipy-organizers] Publication and review in SciPy
To: Jacob Barhak <jacob....@gmail.com>
Cc: "scipy-or...@scipy.org" <scipy-or...@scipy.org>


Hello All,

Before we start discussing the future of SciPy Proceedings, we need to
have a little perspective on where things are and who is involved.
Let me try to summarize what has been going on.

First, we have to ask why does SciPy have a proceedings to begin with.
 I think one answer is that the current state of journals is hostile
towards publishing ideas our communities hold dear.  Things like
scientific software should be built to be usable and here is a code
showing something that is known but done in a way that is accessible.
As co-chair the intention was to get more recognition for the amazing
work coming out of our community, while a video of a talk is nice and
tweets draw interest, archived well done publications stand a longer
test of time.  By helping our community publish more we help the
scientific software field legitimize itself in the ever increasingly
competitive market (academic and otherwise).

Second, we have to establish a means. Companies such as IOP and
Elsevier make money off such publications.  Even societies such as
SIAM and ACM draw the majority of their funds from journal and
conference proceedings.  SciPy Proceedings is entirely volunteer with
no revenue and this needs to be kept in mind when deciding these
wonderful goals, whose gonna do the work and why.  While I think we
had a much more positive review process last year than in the previous
few years, we still don't have the proceedings up in a readable,
archivable format. Without a surge of fresh hands helping out with
this portion of the conference, (actually reviewing, helping with the
publishing, and so on) I am hesitant to push forward on this pursuit.
(With that said, I've added two positions to the SciPy2014 board as
"Technical Committee Chairs" including Sheila who has a great deal of
experience with challenging the publishing norms.)

Third, a bit of history.  Last year we pursued a route of having both
the proceedings that included our traditional 6 page documents with a
low review overhead and including a focus issue with a more
prestigious open journal, Computational Science and Discovery.  CSD
was a newer journal trying to push changes in the field, but
unfortunately have had many setbacks (including staffing problems)
that have basically hamstrung our interactions.  At SciPy2013 I had
invited several journal editors in scientific computing to be part of
a BOF that ended up being canceled due to their unavailability.  I was
pleased that the Center for Open Science was still able to have a
strong presence.  CSD is still interested in having the focus issue
for SciPy2013, but I am not very confident that it will happen (yes,
I'm emailing them quite a bit these days.)

Finally, where are we.  Last year Jarrod and Stéfan built a review
system based on Github pull requests
(https://github.com/scipy/scipy_proceedings/pulls?direction=desc&page=1&sort=created&state=closed).
 This replaced the previous system of putting reviews on the website,
but SciPy2013 Proceedings website has yet to be built
(https://github.com/scipy/scipy_proceedings/issues/70).  Jarrod has
stepped down as chair and we are still searching for the co-chair for
this role.  I think there is a lot of potential in this approach but
would agree it needs a few steps to make a well-done professional
procedure that will attract more attention.

At this point of the juncture, I am convince if we want to change the
state of things we need to do it ourselves in a professional
well-publicized manner, or find a strong partner (investing in Jarrod
and Stefan's github review system, Center for Open Science, inSCIght).
 I would prefer to take partners at face value and work with them to
produce a product, not a disguised test of their services. Finally in
a time of crisis in scientific research, we need to hold ourselves to
standards that are much higher than our many academic peers.

-- Andy



On Tue, Oct 29, 2013 at 9:38 AM, Jacob Barhak <jacob....@gmail.com> wrote:
> Hi Matt,
>
> The link you sent is relevant yet will take a long time to process - there are many ideas out there in that conference.
>
> As for your second remark regarding partnering. Well, you can have a very basic solution with little effort using github and just specifying how to use it properly. The 2013 github publication model combined with the 2012 open review policy may be a good base. From there on you can always build further. Yet first you should have a solid  simple base.
>
> If you wish to test an external partner for publication it is possible to test beforehand by submitting a paper and see how it is handled - I can help here if you have specifics in mind.
>
> Thanks for your fast response.
>
>         Jacob
>
> Sent from my iPhone
>
> On Oct 29, 2013, at 9:13 AM, Matthew Turk <matth...@gmail.com> wrote:
>
>> Hi Jacob,
>>
>> On Tue, Oct 29, 2013 at 10:01 AM, Jacob Barhak <jacob....@gmail.com> wrote:
>>> Hello to all SciPy organizers.
>>>
>>> This is submitted here after an email conversation with some of the organizers pointing towards an ineffective journal publication venue in 2013. Andy invited me to send the conversation here to address a larger pool of opinions in SciPy.
>>>
>>> The traditional journal publication system is quite broken and cannot keep up with technological changes. Here are some examples:
>>> 1. The review processes are cumbersome blind and long
>>> 2. Journal publications are not geared towards code publication
>>> 3. Version control and sharing are not embedded in most of those systems
>>>
>>> The changing landscape of technology may call for other publication alternatives for the SciPy proceedings that do not need to rely on old journal type publication.
>>>
>>> Journal publications are still used for promotion and other recognition within the scientific community, yet if the traditional system is so broken, then it is time for a better alternative. SciPy is a good base for forming such an alternative.
>>
>> I find this to be an extremely interesting avenue, and SciPy is indeed
>> a good venue for opening up these discussions.  Last year, Will
>> Schroeder's keynote touched upon the work being done through the
>> Insight Journal, which also attempts to address many of these
>> shortcomings.  The WSSSPE workshop at SC13 this year has several
>> contributions that discuss publishing models, too:
>> http://wssspe.researchcomputing.org.uk/contributions/ .
>>
>>>
>>> I really liked the path taken in 2012 where reviews were being asked and openly stored with the paper - a non blind review. I would like to see more of this approach. This is more similar to testing software where someone has to sign on a product.
>>>
>>> I would suggest some elements that make sense to me to keep publication effective:
>>>
>>> 1. Use github or a similar repository or a wiki to publish SciPy proceedings - this will allow linking to code, video, slides, etc.
>>>
>>> 2. Emphasize electronic publication over traditional paper formatting. Which can be accomplished using simple RST or MD or similar non demanding non time consuming formatting.
>>>
>>> 3. Ensure high quality that is accountable by using open non blind review process.
>>>
>>> Note that the latter review process can continue even after publication and paper submitters may be asked to participate in open review as part of participating in SciPy.
>>>
>>> There are just a few ideas. I would appreciate a discussion on those issues to help improve SciPy in the future and use its innovative spirit to influence the scientific community in better directions.
>>
>> Attempting to move the proceedings to a non-traditional journal, or
>> even start one, could be a very beneficial both for SciPy the
>> conference and the community.  My main reaction to this is that there
>> are so many possible partners out there, both within the python/scipy
>> community as well as in the broader "open science" or even
>> computational science communities, that we would really need to ensure
>> we have as many partners in this as possible, which might make it
>> broader than we can pull off for 2014 proceedings.
>>
>> -Matt
>>
>>>
>>>        Jacob
>>>
>>> Sent from my iPhone
>>> _______________________________________________
>>> Scipy-organizers mailing list
>>> Scipy-or...@scipy.org
>>> http://mail.scipy.org/mailman/listinfo/scipy-organizers
> _______________________________________________
> Scipy-organizers mailing list
> Scipy-or...@scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-organizers
_______________________________________________
Scipy-organizers mailing list
Scipy-or...@scipy.org
http://mail.scipy.org/mailman/listinfo/scipy-organizers

Reply all
Reply to author
Forward
0 new messages