Formatting fixes

[wgropp]

Problem

There are numerous minor formatting issues in the RC. Here are some of them. Pages refer to page numbers in the RC.

1. xx-xxvii: Make section number space wider in the list of figures/tables/examples. Probably a komascript option

2. xliii-xliv: move two names from the list of people to the bottom of the first list (Hugo Taboada and Keita Teranishi)

3. p101 line 31: MPI_RECV needs mpifuncshort

4. Correct use of \code (language specific, e.g., C or Fortran) and \mpicode (language neutral). (many places) Deferred

5. p 146-7, the f08 _c binding is split across pages, with only the first line (MPI_Type_create_darray...) on p 146. Made an adjustment to an optional skip with penalty before the bindingboy which worked, but also sometimes added a lot of space at the bottom of a page. I'm not going to fix this one.

6. p147. Should the HPF code be \code or \mpicode? Already uses \code. No change.

7. p147. Should "cyclic()" (line 47) be \mpicode? (Yes)

8. p155, line 44. max should be \max? Yes. There were also a few min and max that also needed to be made \min and \max (there was inconsistent use in some of the equations).

9. p 166. The left and right columns in the table look like the names are in slightly different font sizes. This is odd. Should consts and functions have the same size? Should the size of function names in this table be smaller? See also p 230, line 15. Deferred

10. p 194, line 34. Is there too much space in "Chapter 7"? Yes - it is a "Chapter ~\ref..." - that space before the tie causes the problem. Need to add that as a check. Note : used in a few places in math mode to add space around the colon. Note - added a check in getlatex which found another problem in appLang-Const.

11. p 238. The spacing after "..." is inconsistent - see lines 43, 44, and 46. - Fixed in this location. Note, however, that the document is inconsistent - sometimes there is a space and sometimes there is not after the ... (more precisely, some use "$\ldots$," and some use "$\ldots$ ,")

12. p 248, line 41. The spacing around the "=" is inconsistent. Note there is inconsistent formatting of a=b when used as text for "a has the value b" or "a is given the value b". - Defer until agreement on conventions around use of =.

13. p 299, line 39. Inconsistent spacing in the sets.

14. p 317, line 5 and line 36. Spacing around "=" is uneven. Defer until convention on using =.

15. p 339, line 39. Check spacing before \MPI/ in caption for figure 7.3 (on review of the LaTeX, it looks ok - nothing to do).

16. p 341, line 42. Unfortunate line break at line 42. This was caused by an explicit flushline, which has been removed.

17. p 375, line 7. The MPI_XXX_CREATE_KEYVAL should not cause a line break

18. p 378, line 40. comm_name should use short version (mpiargshort)

19. p 391, lines 26-29. This would be better as an eqnalign (line the = up). Deferred.

20. p 405-6. The description for the argument "weighted" is split across lines. Can we make the description either unbreakable, or unbreakable is very short, as in this case?

21. p 448. Uses an en-dash instead of an em-dash in two places. Need a test for this (tricky, because ok in code). Fixed in the text; deferred adding a test.

22. p 461. There is a \noindent before "With the Fortran mpi_f08 module". This is incorrect. Either remove the \noindent or start a new paragraph (and also remove the \noindent). Deferred

23. p 470, line 6. Use short for errorcode.

24. p 491ff. There are code without the grey listing background. Is this intended? Deferred.

25. p 513. Should there be periods at the end of the bullet items in lines 5-8? Deferred.

26. p 557, line 41. There is a true that should be \infoval{true}. Also line 36, page 558 line 6

27. p 565. The constlists have inconsistent case - the first block is lowercase and the second block is capitalized. Also update convention for these. Deferred.

28. p 565, line 23. base should be mpiargshort.

29. p 579, line 27. MPI_NO_OP should be short

30. p590. Should MPI_WIN_POST be short?

31. p594 lines 6 and 6. win should be short.

32. p617-8. All three figures (12.6-8) should fit on the same page. Fix puts two on one page and the last on the next page, but also allows text on the page.

33. p 619, lines 32. The BSEND-RECV-DETACH should use mpifuncskip for the correct font.

34. p 642, the "/tmp/foo" should either be a string (so the correct quotes are used) or not use quotes. Deferred.

35. p 649, line 5. Section 10 should be Chapter 10

36. p 649, line 6, there is a bare MPI. Why wasn't this caught in getlatex? Fixed in getlatex; additional bare MPIs found and fixed.

37. p 650, line 38 No quotes around info. Deferred

38. p 656. Table 14.1 is alone on a page. Why doesn't the text on p 657 start on this page?

39. p 696-698, tables 14.2-4 interrupt the C binding MPI_Register_datarep_c. Deferred

40. p738, line 29. There are too many spaces before "In general". This is because index macros used without a trailing %:
    Some performance variables and classes refer to \mpitermdefni{events}.
    \mpitermdefindex{event!tool information interface}
    \mpitermdefindex{tool information interface!event}
    Could add a check for ^\mpitermdefindex{.*} *$ to getlatex. (This check has been added)

41. p 795. Table 19.1 is alone on a page. Text from p 796 should be on that page. Fixed

42. p 818. Paragraphs are missing their indents. Are \noindent being used where it shouldn't? No, this is an artifact of paragraphs within an itemize not being indented. Correct as is.

43. p 835. It would be good for the paragraph indent and the hanging indent on description to be the same. The top of this page shows the ugly misalignment of these two offsets.

44. p 838, line 20. Consider moving the definition into a display (lstlisting block) to avoid the awful linebreak. Deferred

45. p 861. Better to use [h] rather than [t] for the tables (so the last line isn't separated by several tables). There is also room for at least table 20.5 on this page; why isn't it there? Attempted but no joy.

46. p 887. There is a bad page break at the end of the page. Is there a good page break option at the beginning of this definition? Deferred

47. p 892. "chunked" should be listed before "chunked_item"

48. p897. In a perfect world, the "flag=TRUE" and "flag=FALSE" would be in \mpicode.

Proposal

Correct these issues

Changes to the Text

Minor changes to the formatting, no changes to the text

Impact on Implementations

None

Impact on Users

Improves the quality of the document

References and Pull Requests

See mpi-forum/mpi-standard/pull/1119 , which handles the above that are not marked "Deferred"