Hyperlink export regression

[Pavel Sanda]

On Sun, Jul 07, 2024 at 01:08:38PM +0200, Pavel Sanda wrote:
> in 2.3 we xhtml-export hyperlink "aaa.pdf" of type "file" as "aaa.pdf".
> 
> in 2.4 we xhtml-export hyperlink of type "file" via "file:aaa.pdf".
> 
> I think this goes back to 144cf4bb9afaa.
> 
> Now this breaks html links in old documents without any warning
> and it's difficult to remeber to review all hyperlinks accross
> many document which I use for xhtml. Also I have scattered
> such documentes all over the place and do not see any easy way
> how to detect all of them in one step...
> 
> What shall we do with it? Is conversion in lyx2lyx in 2.3 - > 2.4
> to move "file" -> "web" type reasonable? (Did not check what will
> happen with pdf output though).

Alas, I checked the pdf output now and unlike xhtml output the distinction
between file and web output exists in 2.3. So fixing the situation via lyx2lyx 
is not an option because it will break pdf exports.

At the moment I do not see better option to clearly state this in release
notes and suggest there manual fixes of the files doing xhtml export.
(Which might be pretty ugly for more lengthy documents...)


Pavel