I have only a short time ago downloaded and installed DEVONthink 4.3 and have been reviewing the extensive markdown enhancements. I have come across one problem I seem unable to resolve.
I have a number of markdown documents that include images. Each image is the subject of a DEVONthink Item Link and always followed the form:
![alt text] (URL)
where alt text was the name of the relevant JPEG image minus the extension. The result was that alt text displayed as the caption of the image.
That still works if in Settings → Files → Markdown I select, under “Preview”, MultiMarkdown 6. However, no caption appears if I select DEVONthink markdown. (I have experimented switching Image Reference between Item Link and Automatic but that makes no difference. The caption is still absent when using DEVONthink markdown.)
In essence, I’d like to be able to use DEVONthink markdown and retain the hundreds of captioned images I have in markdown documents without working through all of them and revising the links in some way.
It is entirely possible that I have missed something obvious, of course. If so, I apologise for wasting everyone’s time.
That is a peculiarity of MMD (see here: Images) IMO, the alt text should not be visible in the generated HTML unless the image is missing. The MD title should be displayed as caption for the image, if it is available.
Regardless, the image rendering in DT 4.3 MD is weird.
# Markdown title
first paragraph followed by an image
and some more text in the next paragraph.
Trivial example. With a title in the image, the image is not rendered at all. Which is explained by the HTML generated incorrectly like this (in MMD6):
# Markdown title
first paragraph followed by an image
and some more text in the next paragraph.
the image is rendered correctly both for DT MD and MMD6. DT’s MD produces a
figure, as does MMD6. The difference here being that MMD6 adds a figcaption element containing the alt text (@Stephen_C: sorry for blabbing so extensively, but I wanted to clarify what is happening).
Now, how to fix it? First, IMO,  should be rendered as <img src="URL" alt="alt" title="title">. That is a safe bet because it works inline and in stand-alone images.
For stand-alone images, I’d suggest this: <figure><img src="URL" alt="alt"><figcaption>title</figcaption></figure>.
The alt attribute is intended as a replacement for the image if it can’t be loaded and (even if it can be loaded!) as a text alternative for visually impaired users. Putting it in the caption is duplicating the text and not helpful for accessibility.
If that seems to strict a view, an alternative might be to put the alt text in the figcaption, if the title is missing from the MD. That’s what @Stephen_C was experiencing with MMD6. And I still see it here with DT4.3 and MMD6.
either DEVONtechnologies (after fixing the obvious problem with images and titles) follows MMD6 and shows the alt text in a figcaption element. Then you could simply use the new renderer.
or they decide to put the title in the figcaption element instead. Then you could use a script to change you MD documents so that the current alt text becomes a title.
In any case, we’ll have to wait for the next DT release.