[IPOL discuss] revision of the Software Guidelines - second draft 1.01b

[Nicolas Limare]

Hi,

> I would like to suggest a few updates to the current IPOL software
> guidelines. My first draft is available online:
> -> https://tools.ipol.im/wiki/ref/software_guidelines/drafts/
> 
> I will collect all these remarks and propose a second draft after
> two weeks.

I'm a little late, but the second draft (1.01b) is online:
-> https://tools.ipol.im/wiki/ref/software_guidelines/drafts/

As usual, send your comments, and I will use them for the next draft
version, expected Sept. 23. And by "comments", I also mean your
propositions to add anything else to these guidelines.

If you skipped the previous draft because of the hot summer holidays,
it is detailed there: 
-> https://tools.ipol.im/mailman/archive/discuss/2012-August/000887.html

8<----------8<----------8<----------8<----------8<----------8<----------
Changes from the previous draft (1.01a)

* Use the phrase "support code" instead of "foreign code"

Just sounds more appropriate.

* Add a sentence about the size of this "support code"

I used Enric's example.

* Only mention C89, C99 and C++98

"ANSI C" was used in the previous versions because it has been the
informal way people referred to C89 for many years, but this
denomination was not needed in the guidelines, and is becoming more
and more ambiguous since there is a third C standard now (C11).

Links to the standards were already provided in a footnote; I updated
a broken link. I prefer to link to the final draft versions for some
standards: they are available for free, and exactly identical to the
official versions. The reference name of the official documents is
mentioned too.

* Require "#ifdef _OPENMP" for OpenMP

This is what I meant by "must compile without OpenMP". Enric's
comments show that it is better to write it clearly in the guidelines.

* Remove the requirement about the tabs/spaces

As Enric guessed, I first added this requirement to provide a precise
meaning for the 80-column limit, following a comment I read once about
this question. But asking to replace tabs by spaces, or deciding if a
tab is 2, 4 or 8 columns, is purely a matter of style and personal
preferences, and I tried to avoid these matters in the guidelines.

So, I think it's better to let people do as they like, and whatever
size they use for their tabs, it will have no consequence at all. The
"validation tool" will still count the *characters*, not the *colums*,
so 1 tab = 1 character, and no one will be annoyed by this detail.

-- 
Nicolas LIMARE - CMLA - ENS Cachan         http://limare.perso.math.cnrs.fr/
IPOL - image processing on line                          http://www.ipol.im/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <https://tools.ipol.im/mailman/archive/discuss/attachments/20120909/b8d9d341/attachment.pgp>