[IPOL discuss] Code Ocean and IPOL

[Miguel Colom]

Dear all,

Recently we have discovered the "Code Ocean" platform associated to  
IEEE which encourages to publish the source codes associated to the  
articles and to allow users to execute online the algorithms:  
https://codeocean.com

It think it's interesting that you let us known your opinions about  
this platform.

There are some similarities and differences between Code Ocean and  
IPOL (see https://codeocean.com/learn/faq):

- Code Ocean's users need to pay to use the platform according to the  
use (say, per CPU time and stored data). IPOL is free to use.

- Code Ocean's users can close the browser and the algorithm will keep  
running in background. They the open again the page they'll get the  
result if the algorithm has finished. IPOL is similar.

- Their source code is not reviewed, only very superficially. IPOL  
requires the reviewers and editors to make a complete review of the  
source code.

- They don't allow any changes once one algorithm has been published.  
Neither IPOL. IPOL with very special exceptions (erratas, bugs  
discovered after publication)

- They allow to use GPUs, and we don't. However, with the new demo  
system we could attach a GPU to any machine we want and declare in the  
demo that it needs to be run under specific hardware. We can have it  
easily.

- They use several MATLAB boxes and they can study adding more upon  
request. In IPOL we do the same.

- Their demos run in a machine which has at least 8 cores. We have in  
IPOL dedicated servers with 32 cores (16 HT).

- Both Code Ocean and IPOL run their demos in several servers to share  
the computational load.

It seems that Code Ocean is a new player in this game!

Another topic which we could discuss is the Impact Factor of IPOL.  
IPOL is a well-known journal at this moment, but still we don't have  
an Impact Factor (IF). The IF is a metric which Thomson-Reuters gives  
to the journals after they have passed positively an evaluation.

I've found this article by Clarivate Analytics (a company that was  
bought by Thomson-Reuters in 2016) where they explain some of the  
criteria for obtaining an:  
http://wokinfo.com/essays/journal-selection-process/

This PDF lists the publications which obtained an IF in 2016:  
http://scientific.thomsonreuters.com/imgblast/JCRFullCovlist-2016.pdf

If you have any more information about the IF, or ideas, comments, or  
whatever you'd like the share, please go ahead! Also your opinions on  
Code Ocean.

Best,
Miguel