Reading the results

The results screen is built around four numbers. They are not just a summary: each card is a tab, and clicking one changes what the table below it shows.

Two of them mean less than they appear to. Read this page once and the numbers stop being alarming.


#The four cards

Card

What it counts

Pages changed

How many pages the run actually rewrote, with a caption showing how many were checked in total. Far smaller than "checked", because the search that finds work is deliberately wider than the conversion itself.

Macros migrated

Macros converted cleanly.

Macros failed

Macros that could not be converted, counted per macro (not per page). The caption reads "N unexplained" or "all explained", where explained means Capable recognised the cause and can offer a hint. A failed macro is left exactly as it was on the page.

Unsupported

Macros left as they were. This is not only "we cannot convert this", it also counts pure do-nothing cases, where a selected migrator had nothing to do on a page. On a wide run those can dominate the number, which looks alarming but is not a data problem.


#What each card opens

Card

Shows

Nothing selected

A per-migrator breakdown, with a percentage-clean bar for each migrator

Pages changed

The page-by-page activity log

Macros migrated

The same log, filtered to successful conversions

Macros failed

Issues grouped by cause, with a count and, where we recognise it, a plain-English hint

Unsupported

The same grouped view for macros that were left as they were


#Grouped issues

Identical problems collapse into one row with a count. Where Capable recognises the cause, the row carries a hint saying what to do about it. Everything drills down to the exact pages affected.

The Macros failed caption reads either "N unexplained" or "all explained". Explained means we recognised the cause and can offer a hint, not that it is harmless.


#A few things that catch people out

  • Macros failed counts macros, not pages. One page with twenty broken diagrams is twenty failures.

  • Recognised-versus-unexpected is based on matching the wording of the error message. A genuinely new failure whose wording resembles a known one will be shown as explained.

  • Drill-down from an issue group lists at most 2,000 pages, so on a very large run the group count and the list will disagree. The counts are the accurate ones.



Unsupported is mostly nothing. Failed is the number to read.