Ver 3.0.0 - 01
Just a heads up, I think something went awry with section 19.1, it’s kinda a giant wall of text with some odd lines that I haven’t run across elsewhere in the manual.
Thanks! The section of text was enclosed within an invisible table cell, and since I have the lists & tables to MultiMarkdown converter enabled for that project, it was placing MMD syntax around the entire section—which doesn’t work!
That bothers me as well, but I couldn’t figure out why it wasn’t working before launch. I have the configuration code set properly for numbered PDF bookmarks, but they refuse to print. Well I’ll have to keep playing with it. I consider the current state of the ToC a temporary issue.
A few more buglets:
- p.825 footnote NEEDS_LINK text
- p.829 incomplete link [Updating a Document Without Styles to Use Styles]
- p.71 incomplete link [Using Scrivener in a Secure Environment]
Actually I just did a search for [ and there are quite a few more incomplete links (e.g. p.668)
I figured out the bookmark numbering problem. That’ll be fixed for the next revision.
On the links, thanks! I stumbled on that by using an older search pattern. I’ve taken to using single bracket form for links, [like so], rather than [the bulkier way][]. In the past after I opened up the compiled .tex file, I’d always search for “][” as that would find all broken links with only one false positive in the MMD chapter. It’ll be a little more difficult finding […] as a broken link in a LaTeX file, unfortunately. I guess that’s one point for using the bulkier syntax.
…now, if you were using Pandoc you could use the AST filters to find all links without resorting to fragile regexes 
That might happen. I didn’t have time to even think about a switch of that magnitude over the past few months, but MMD6 has removed the ability to freely inject LaTeX inline as best I can tell. It has a block level insertion capability, but I need way more than that. So the sad truth is that like the v2 user manual, which was released on the cusp of MMD3, it may end up being pinned to a legacy version on account of that major change to the syntax—unless I shift the engine it requires.
But first, fixing typos, broken links and missing sections. 
Actually though I haven’t tested it in MMD6, there is a “better” way to inject different types of block and inline raw content in MMD6 (and Pandoc2): \enquote{aperture problem}{=latex}
github.com/jgm/pandoc/issues/3537)
I had a small part to play in trying to get both MMD and Pandoc to use a common syntax. Anyway if that was the only reason to stop you migrating, it has been solved, and you can use the same syntax between the two engines.
Well that’s certainly good news to hear. I hadn’t really had the time to look into it yet, and only came across the fence-quote method from someone else a bit ago. That looks very similar to that approach, so I imagine my source simply wasn’t aware of the inline method.
Thanks! I do think that’s a better approach. I was never a huge fan of the HTML comment method, for the same reason I was never a huge fan of using Scrivener’s inline annotations to fake styles, back in the day. It consumes a valuable tool.
I’ll have to give it some tests. That really is the only thing holding me back right now. The other syntax changes were about things I never made use of.
Ah… and as I say that, my BBCode generator is broken with MMD6. Well—I’ll test that after I generate this post. Ha!
You do a wonderful job with the manuals. I read them linearly and as resources.
A request, if possible: Will you please post version info at the download site (I save them as PDF for tablet)?
Thanks! 
I will try, it something I wanted to do, but the page template doesn’t have a really good place for inserting metadata or annotating the downloads outside of jamming lots of text into the dropdown menu.
There is a word missing in the first sentence of the second paragraph on page 530: “Using a table of contents of <???> with page numbers and pdfs…”.
It might be a good idea to have a easy to find/pinned topic for us to report errata in the Scrivener 3 Manual…
EDIT: The last line of page 530, “Table of Contents : a special compile tab used for setting up eBooks.” looks out of place, as it runs into the left page margin. Maybe it was intended to be a footnote?
Thanks! I reworded that paragraph a bit, it was a rather messy way of saying PDF is best for distributing reader copies.
The second problem was a broken cross-reference. Those show up with brackets in the text, and brackets mess up the bullet environment. There are a few broken xrefs in the public revision. I’ve cleaned up all that I could find at this point, but these in particular are difficult to spot since the square brackets don’t show up in the PDF.
I’ll set up a sticky thread, that’s a good idea.
In “Quickly switching between editors” on page 157:
The second keyboard shortcut should be ⌃⌥⌘ R.
AmberV: it would be nice if that page or link to PDF said what version is currently the latest, so we know if we should redownload it or not…
Building on that, it would be quite lovely in general if any download included the version number in the file name, for precisely that reason.
Hmm, it’s maybe a little messy looking—and perhaps needlessly so, as I don’t know how important the revision number is in the long run. Normally the software has the latest version, and it is as simple as that. It is in fact fairly rare that I’ll need to upload an update in between releases. This is an unusual case right now where 3.0.1 has the 3.0.0 manual.
It’s nice when you’re downloading a file to know what version it is. For those files included in the Scrivener package, of course it would be needlessly messy.