[ho-tex] vertical spacing with amsmath+fleqn+hyperref

[David Carlisle]

On 14 May 2016 at 23:38, Reinhard Kotucha <reinhard.kotucha at web.de> wrote:
> Hi,
> when I use hyperref and amsmath with the fleqn option, extra vertical
> space is added.
>
>    \documentclass{article}
>    \usepackage[fleqn]{amsmath}
>    \usepackage{hyperref}
>
>    \begin{document}
>       \begin{equation}
>          a+b=c
>       \end{equation}
>       \begin{equation}
>          a+b=c
>       \end{equation}
>       \begin{equation}
>          a+b=c
>       \end{equation}
>       \begin{equation}
>          a+b=c
>       \end{equation}
>    \end{document}
>
>
> Everything is fine without the amsmath package
>
>    \documentclass[fleqn]{article}
>    \usepackage{hyperref}
>
> and everything is fine without the fleqn option
>
>    \documentclass{article}
>    \usepackage{amsmath}
>    \usepackage{hyperref}
>
>
> I already discussed the problem with Heiko at Dante-2014 but Heiko
> told me that he probably hasn't enough time to investigate.  Could
> anybody else look into it?
>
> Of course, whatsits have an impact on linebreaks and hyperref takes it
> into account.  It seems that hyperref simply doesn't support
> amsmath+fleqn (yet).
>
> BTW, even without hyperref, vertical spacing doesn't look correct.
> When amsmath is used *and* the fleqn option is set, there is more
> vertical space between the first and second equation than between
> others.  But this is certainly an amsmath issue and unrelated here.
>
> Regards,
>   Reinhard
>
> Karl, thank you very much for establishing this mailing list.
>
> --

Thanks I guess this is for me, I'll try first to at least work out
whether the issue lies with hyperref or amsmath (in case it's still
with me, but wearing a different hat:-)

Note however that it is always wrong to have a display math at the
start of a paragraph or immediately after another you get spurious
"white" paragraphs in that case, but it would be good if hyperref
didn't change spacing anyway.

Unlikely to get to this before the for first TL2016 cutoff (code too
risky to attempt to change in that timeframe!)

David