The question I have deals with error detection within Arbortext.
Hello - The question I have deals with error detection within Arbortext. Background: I work in a group of 6 engineers that write service documentation using the SGML verion of Arbortext. This group includes a single book-boss, who is the most expert in SGML editor tools, including Arbortext.
I typically work on small books (100 - 200 pages) that are combined into larger books that amount to in excess of 4,000 pages. When the junior writers submit smaller files (IDE) to the book-boss, one task performed by the book-boss is to remove formatting problems and things that might cause problems with the book to build without error. I recently submitted a small book (<100 pages) IDE file to the book boss that had a syntax error that was not detectable using the "Completeness Check" or "Green Checkmark" or the publishing tool used to build.
It was discovered after three days of debugging at the larger book level (>4,000 pages) that the error I introduced was an extraneous "newline" command at the end of an ordered list (within a proper division). In this case, nothing was entered after the "newline" command, so essentially a "newline" to nothing was the issue. With the small book (<100 pages) there was no indication whatsoever an error existed. Several attempts to break the 4,000 page book up into smaller 1,000 page books and publish indicated the same: No Error. However, when the larger book went to publish (a 2.5 hour build time, each time) a fatal error occurred. The Arbortext tool didn't issue any clue or error description of the "newline" to nothing fatal error before or after the build.
The process to find the error was the fine tooth comb method that was trial and error as the culprit IDE file was initially unknown. This process took roughly 3 days of the book-boss's time.
Question: Are there error detection tools within Arbortext besides Completeness Check (Green Checkmark) or the ultimate publish the book and hope for the best method.
Suggestion-Question: If not, is PTC working on improving error detection so that hidden errors in small books go detected early in the documentation process before they make it into the larger books, making it painstaking to resolve errors, in the case above, a minor syntax error.