[IPOL discuss] IPOL Software Guidelines, 2nd draft (and last one?)

José Luis Lisani joseluis.lisani at uib.es
Wed Nov 23 11:14:10 CET 2011


Hi Nicolas,

excellent job.

Just a few, minor, comments:


- notations "MiB" and "GiB". I had to surf the internet for a while 
until finding
the definitions of "medibyte" and "gibibyte". Maybe you could add some link
to explain the difference with the most commonly used terms "megabyte"
and "gigabyte"

- 9th item in check list:replace "interactove" by "interactive"

- we SHOULD recommend, in the source code, the separation (different files)
between "algorithm core" (the one described, *and reviewed*, in the IPOL 
web page) and
the rest (input/output and other auxiliary routines, not reviewed by IPOL).
Of course, the interaction of the "core" with the "rest" should also be 
checked for errors and inconsistencies,
but only the "core" will be thoroughly reviewed.
This would facilitate future modifications of the source code, which 
could be easily incorporated,
without thorought review, if they do not affect the code "core".
The authors MUST clearly state which files are submitted for publication 
and this information
should be included in the README.txt file.
(This is related with another topic in the discuss forum:  "Corrections 
of erratas in IPOL articles").

- The meaning of the following sencente (in "Copyright attribution") is 
unclear to me:
"Of course, in case of a file based on a previous work, the copyright 
attribution to the previous authors MUST NOT be removed. "

I understand that by "previous work" you mean an "already existing 
source code".
That is, the authors modify a previously existing source file.
If the algorithm is based on the work of another author but the original 
code is not used, then the name of the original
author must not appear in the copyright attribution (e.g. ASIFT is based 
on D. Lowe's articles, but he is not cited in the
copyright note)


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://tools.ipol.im/mailman/archive/discuss/attachments/20111123/44ca9526/attachment.html>


More information about the discuss mailing list