RFC: either rename "kind" and "level" into some thing more descriptive about duecredit HOT 4 CLOSED

I like tags better than kind. I think that the tricky part might be to come up with a complete semantic that can cover all possible cases in the future.

For tags, what about changing canonical to ref?

And for level maybe allow only

• module
• function
• class
• class_method
• data_structure (?)

what else?

from duecredit.

yeah -- let it be 'tags' indeed. We will not figure out possible cases until we start covering them ;-)
I don't mind 'ref' -- let's stick to it

level -- what would be the meaning for 'class'? if no methods invoked, it means not used, we wouldn't care... or it might somehow be the same as "module" in terms of grouping -- but indeed we would need to see a use-case to see how valuable that is. But that makes me think that may be instead of "level" it should be named more descriptively as a "group" or even better a "namespace" or ...?

data_structure -- what would be a use-case and how would be different/and triggered?

from duecredit.

On Fri, Jul 10, 2015 at 11:23 PM, Yaroslav Halchenko <
[email protected]> wrote:

level -- what would be the meaning for 'class'? if no methods invoked, it
means not used, we wouldn't care... or it might somehow be the same as
"module" in terms of grouping -- but indeed we would need to see a use-case
to see how valuable that is. But that makes me think that may be instead of
"level" it should be named more descriptively as a "group" or even better a
"namespace" or ...

Hmm, if we step back, we want to give due credit to certain methods
(algorithms) and software packages. But packages can implement modules
inspired from elsewhere. Probably then we just need to specify these
levels: functions (or methods), modules, and packages, to follow python's
structure.

data_structure -- what would be a use-case and how would be different/and
triggered?

I was just thinking that if a new method implements a new data structure
(which I don't know how likely it is :-P), then that wouldn't fall into the
other categories. But it's very unlikely that is going to be used on second
thought.

from duecredit.

On Sat, 11 Jul 2015, Matteo Visconti di Oleggio Castello wrote:

to see how valuable that is. But that makes me think that may be instead
of
"level" it should be named more descriptively as a "group" or even
better a
"namespace" or ...

Hmm, if we step back, we want to give due credit to certain methods
(algorithms) and software packages. But packages can implement modules
inspired from elsewhere. Probably then we just need to specify these
levels: functions (or methods), modules, and packages, to follow python's
structure.

rright, so pretty much again, either "namespace" or "level" which would
give us the hierarchy determining to which higher level (module or
package) cited functionality belongs

your comment also inspired a possible additional tag, which could be
"original", or reincarnated "canonical" (in addition to "ref"), which
would signal that it is the "original" implementation by the method's
authors. Unlike someone else implementing the method (e.g. in our case
implementation of Seltzer's stat assessment in PyMVPA is "ref" but not
"original")?

Yaroslav O. Halchenko, Ph.D.
http://neuro.debian.net http://www.pymvpa.org http://www.fail2ban.org
Research Scientist, Psychological and Brain Sciences Dept.
Dartmouth College, 419 Moore Hall, Hinman Box 6207, Hanover, NH 03755
Phone: +1 (603) 646-9834 Fax: +1 (603) 646-1419
WWW: http://www.linkedin.com/in/yarik

from duecredit.