Sometimes it is hard to decide whether to continue maintaining an
atoms module or not.... So I did "snapshots" of downloads on
9/11/2020 21:44 and 14/12/2020 12:20 respectively to see which
modules are in the actual top downloads.
Still, this is not a fair result, such as IPCV having some
"patches" which will increase the download number, and some
other factors as well.
So should IPCV continue to be maintained ...?
Here's the results.
Image Processing and Computer
This module is a small one, containing a single iodelay() public
function, and only functions --mainly overloads -- in Scilab
language, making the module easy to update.
This module is among the ATOMS Best-of noticeably because several
other popular modules require it as a dependency. Then iodelay is
automatically installed when its "parents" are installed.
This means that when iodelay is not up-to-date, this blocks other
modules depending on it to work. This is currently the case : 10
months after the publication of Scilab 6.1.0, the current iodelay
version is only for Scilab 6.0.
All these aspects make me thinking that iodelay should be part of
Scilab, to be included either in the CACSD or in the
signal_processing internal Scilab modules. Then no repackaging and
re-publication tasks of the module would be needed anymore for
each new Scilab 6.x version.
The iodelay module is currently published under the Cecill license.
This looks compatible with the Scilab's license.
To me this inclusion could be proposed in the next Scilab 6.1.1
release. There is no reason to wait more.
Any comment is welcome about this proposal, noticeably from Serge,
the iodelay author.