[J3] [EXTERNAL] Re: Suggested editorial change

Clune, Thomas L. (GSFC-6101) thomas.l.clune at nasa.gov
Mon Jul 20 09:12:22 EDT 2020

On Jul 20, 2020, at 8:54 AM, Reuben D. Budiardja via J3 <j3 at mailman.j3-fortran.org<mailto:j3 at mailman.j3-fortran.org>> wrote:

I agree with (trying out) the three columns format for the table.

On 07/20/2020 07:41 AM, Clune, Thomas L. (GSFC-6101) via J3 wrote:
I agree.   Continuations of the procedure names is esp. problematic.  It makes the process of scanning the table for a given procedure tedious.   SCALE, SCAN, SELECTED_…, KIND (what?), …SELECTED_…,
Another suggestion, probably not desirable, but I thought worth mentioning:   switch to more of a YAML like presentation.
    Arguments:   ([P, R, RADIX])
    Class:       transformational function
    Description: Real kind selection.

The YAML-like presentation would probably make the list takes many more pages since "Arguments", "Class", and "Description" would be repeated. Maybe it can be mitigated by using a two-column format on those pages but I don't know if that's something we want to do ... This would no longer be a Table.

My thought is that pages (and bits) are free.    But some people do still read hardcopy of the standard or in books.   (I actually like to have my hardcopy Fortran books by big/heavy.   It lends them gravitas.  :-)

- Tom


This is too large a change for an on-the-fly-edit of the sort Malcolm proposed,  but if others consider this approach better, I’m happy to volunteer to put together the edits paper.  (And would try to provide Malcolm with the necessary TeX edits to spare him the tedious exercise.)
- Tom
On Jul 19, 2020, at 5:43 AM, John Reid via J3 <j3 at mailman.j3-fortran.org<mailto:j3 at mailman.j3-fortran.org> <mailto:j3 at mailman.j3-fortran.org>> wrote:


Yes, this would be a good improvement. I don't like the long gaps after short procedure names and I don't like long proecedure names being continued.


Malcolm Cohen via J3 wrote:
I was just looking at Table 16.1, as one does, and wincing at the ugly continuation lines sometimes due to a long intrinsic name and other times due to a long argument list, and wondering if it would look better than instead of
 * Four columns "Procedure", "Arguments", "Class", "Description"
we had
 * Three columns "Procedure (arguments)", "Class", "Description"
I tried it out just on the SELECTED_whatever_KIND intrinsics, and indeed it eliminated the ugly continuations, at the (minor?) cost of no longer having the argument lists all line up.
I'm sure it would not eliminate all continuations, but...
...what do people think? Does this sound like a good idea?
For most entries this is easy - it just takes deleting the first ampersand on the line, but entries that would still be continued may need a bit of extra work to make the continuation indented. So maybe an hour or so of editing time, depending on how fussy I get about the look of the result.
Of course we could go back to the original format if people decide after seeing it that they don't like it after all.
Any and all comments welcome.
..............Malcolm Cohen, NAG Oxford/Tokyo.

Reuben D. Budiardja, Ph.D.
reubendb at ornl.gov<mailto:reubendb at ornl.gov> | (865) 576-9519
National Center for Computational Sciences
Oak Ridge National Laboratory

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.j3-fortran.org/pipermail/j3/attachments/20200720/73d0e2d7/attachment-0001.htm>

More information about the J3 mailing list