[Mpi-forum] Questions on MPI Standard formatting

Sun Sep 5 03:23:18 CDT 2021

Hi Bill, all,

I also agree with what has been said so far and I also agree with the statement that if we don’t find a good solution for tables, leaving it off there is not a big problem (lines should be easily identified there.

As for the syntax highlighting – I personally don’t mind it sticking out, but I can also see Joseph’s point – perhaps something between the two version would work.

One question, though: in the example that goes over two pages, the “MPI_Waitsome” call is only bold for “MPI_Wait” and not bold for “some” – is this an annotation bug or a problem with the package of not detecting prefixes correctly?

Thanks!

Martin

-- 
Prof. Dr. Martin Schulz, Chair of Computer Architecture and Parallel Systems
Department of Informatics, TU-Munich, Boltzmannstraße 3, D-85748 Garching
Member of the Board of Directors at the Leibniz Supercomputing Centre (LRZ)
Email: schulzm at in.tum.de

From: mpi-forum <mpi-forum-bounces at lists.mpi-forum.org> on behalf of "Jeff Squyres (jsquyres) via mpi-forum" <mpi-forum at lists.mpi-forum.org>
Reply-To: Main MPI Forum mailing list <mpi-forum at lists.mpi-forum.org>
Date: Saturday, 4. September 2021 at 19:33
To: MPI Forum list <mpi-forum at lists.mpi-forum.org>
Cc: "Jeff Squyres (jsquyres)" <jsquyres at cisco.com>
Subject: Re: [Mpi-forum] Questions on MPI Standard formatting

I am in agreement with all that has been said so far -- generally, this is great.  Minor tweaks (already cited by George and Joseph) would be nice improvements. 

On Sep 3, 2021, at 5:46 PM, Gropp, William D via mpi-forum <mpi-forum at lists.mpi-forum.org> wrote:

One option is to use a different font; the one I used in the first examples has a strong bold.  Here’s the same page with a different font with a less bold look. In some ways, this may go too far in the other direction - the bold text is a little heavier, but the difference is slight. 

One advantage of this choice is that the code text is not quite as wide as with the font choice for the previous example, which works better with the current code examples.

Let me know your preference.  It is easy to change, modulo the issue of needing to wrap a few lines of code.

Bill

William Gropp
Director, NCSA
Thomas M. Siebel Chair in Computer Science

University of Illinois Urbana-Champaign

IEEE-CS President-Elect

On Sep 3, 2021, at 1:40 PM, Joseph Schuchart via mpi-forum <mpi-forum at lists.mpi-forum.org> wrote:

I am OK with leaving out line numbers in tables. Maybe that will be solved in the future.

One comment I had was on the code formatting. While I definitely prefer the lstlistings formatting over the verbatim we have so far, I find the bold highlighting somewhat distracting. I find it hard to focus on the surrounding code pieces. Would it be possible to tone down the highlighting somewhat (slightly thinner, maybe grey)?

Thanks
Joseph

On 9/3/21 2:36 PM, William Gropp via mpi-forum wrote:

Thanks, George.  On the line numbering, unfortunately, this is very difficult to fix (and is a known problem). My proposal instead is to not use the line numbers in these cases; in the rare cases where we need to identify the location, we can count rows in the table or otherwise identify the line. There are some alternatives, but they are awkward and IMHO ugly, and don’t add much.

Bill

William Gropp
Director, NCSA
Thomas M. Siebel Chair in Computer Science
University of Illinois Urbana-Champaign
IEEE-CS President-Elect

On Sep 3, 2021, at 12:19 PM, George Bosilca <bosilca at icl.utk.edu <mailto:bosilca at icl.utk.edu>> wrote:

Bill,

These PDFs looks great, a clear leap forward in improving the readability and cleanness of the MPI standard.

The only issue I noticed (and that was pinpointed in your email) was about the multi-page tabular where the numbering is done consecutively for the entire tabular environment despite the fact that the table is split across multiple pages, and the numbers are placed on the same side as the first page on all pages. If we can fix this to have the line numbers per page and on the correct side for all environments, that would be great.

Thanks,
  George.

On Thu, Sep 2, 2021 at 11:03 Gropp, William D via mpi-forum <mpi-forum at lists.mpi-forum.org <mailto:mpi-forum at lists.mpi-forum.org>> wrote:

   I’ve been experimenting with updating the MPI Forum document
   formatting, and in particular, modernizing the use of LaTeX and
   reducing the custom LaTeX in the document. Attached are a few
   pages from my tests.  These show:

   1) Use of the lineno package to label each line of text - this
   replaces the ruler currently used, and means that text is more
   accurately labeled, but “boxes”, including tables, figures, and
   captions may not have line numbers. All of the attached pages
   show the line numbers in the margins.
   2) Use of the lstlisting package, with syntax and MPI routine
   highlighting. This also sets a background to gray to help the
   code examples standout. The approach used in MPI 4.0 creates a
   box with a border; because of the box, in that approach, the
   individual lines of code are not numbered.  Hence this change in
   how code examples are marked off. The attached examples show both
   one page and a code example that spans two pages.
   3) There is a way to include numbering for tabular environments
   (and some others).  However, this can lead to some odd labeling,
   because of how the labeling is implemented within the various
   LaTeX packages.  I’ve included some examples which show both the
   expected behavior (consecutive numbering) and the unexpected (and
   nearly impossible to fix) non-consecutive numbering.
   4) I’ve also switched from the book to the srcbook document
   class. This is a more modern and powerful document class, and has
   allowed me to remove some of the custom LaTeX code.

   I’d like feedback on these.  I’d like to adopt some of these for
   the next version. In particular:

   1) Use srcbook instead of book
   2) Use lstlisting in code examples as shown here
   3) For tabular environments, do NOT use line numbers for tabular
   environments used within a table (or other floating) environment.
   This is where the weird line numbering can happen (e.g., see the
   attached page 32). For inline tabular environments, use edtable
   to get line numbers (this is what was used in these examples, and
   it has worked well in other parts of the document)

   If you have suggestions for improvements or alternatives, please
   let me know.

   Bill

   William Gropp
   Director, NCSA
   Thomas M. Siebel Chair in Computer Science
   University of Illinois Urbana-Champaign
   IEEE-CS President-Elect

   _______________________________________________
   mpi-forum mailing list
   mpi-forum at lists.mpi-forum.org <mailto:mpi-forum at lists.mpi-forum.org>
   https://urldefense.com/v3/__https://lists.mpi-forum.org/mailman/listinfo/mpi-forum__;!!DZ3fjg!rX65ncURBYvC7HZUZt7_GihcXnt489vlz7HBN630c59pdHc7EaItJ-sh7U7fkRDrnw$     <https://urldefense.com/v3/__https://lists.mpi-forum.org/mailman/listinfo/mpi-forum__;!!DZ3fjg!ocG0YkYjZCIg26wHqY9ROpnb8nHCQTM8tYRtSx_Yw7RsjsMy5TL85BlU7F6wBdHXnA$>

_______________________________________________
mpi-forum mailing list
mpi-forum at lists.mpi-forum.org
https://urldefense.com/v3/__https://lists.mpi-forum.org/mailman/listinfo/mpi-forum__;!!DZ3fjg!rX65ncURBYvC7HZUZt7_GihcXnt489vlz7HBN630c59pdHc7EaItJ-sh7U7fkRDrnw$ 

_______________________________________________
mpi-forum mailing list
mpi-forum at lists.mpi-forum.org
https://urldefense.com/v3/__https://lists.mpi-forum.org/mailman/listinfo/mpi-forum__;!!DZ3fjg!rX65ncURBYvC7HZUZt7_GihcXnt489vlz7HBN630c59pdHc7EaItJ-sh7U7fkRDrnw$ 

<mpi40-report-9-3-code.pdf>_______________________________________________
mpi-forum mailing list
mpi-forum at lists.mpi-forum.org
https://lists.mpi-forum.org/mailman/listinfo/mpi-forum

-- 

Jeff Squyres

jsquyres at cisco.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mpi-forum.org/pipermail/mpi-forum/attachments/20210905/6487fdeb/attachment.html>