Friends, can I ask for some TikZ help? Consider this code:

\documentclass[12pt]{article}
\usepackage[T1]{fontenc}
\usepackage[utf8]{inputenc}
\usepackage{tikz}
\usetikzlibrary{fit}
\usepackage{forest}

\begin{document}
\begin{minipage}{0.3\textwidth}
\begin{forest}
[a
  [b
    [c]
    [d]
  ]
   [e,draw,dashed]
]
\end{forest}
\end{minipage}%
\begin{minipage}{0.3\textwidth}
\begin{forest}
[f
  [g]
  [,phantom]
  [h,tikz={\node [draw,dashed,remember picture,fit=() (!1) (!l)] {};}
    [i]
    [j]
  ]
]
\end{forest}
\end{minipage}
\end{document}

@PauloCereda Considering.

I was trying to make e and h connect with a line, but I tried some answers in the site using remember picture and even tikzmark but failed...

@AlanMunn ooh. :)

@PauloCereda Hang on.

@AlanMunn Oh thank you! I am really lost.

I am not sure if some update broke or deprecated something... :(

How about:

\documentclass[12pt]{article}
\usepackage[T1]{fontenc}
\usepackage[utf8]{inputenc}
\usepackage{tikz}
\usetikzlibrary{fit}
\usepackage{forest}

\begin{document}
\begin{minipage}{0.3\textwidth}
\begin{forest}
[,phantom,s sep=1in
[a
  [b
    [c]
    [d]
  ]
   [e,draw,dashed,name=e]
]
[f
  [g]
  [,phantom]
  [h,name=h
    [i,name=i]
    [j,name=j]
  ]
]
]
\node (H) [draw,dashed,fit=(h) (i) (j)] {};
\draw (e.east) -- (H.west);
\end{forest}
\end{minipage}
\end{document}

Instead of two minipages and remember picture, just make phantom node to make the two trees be one.

There might be simpler ways to do the fitting in forest but I don't have time to look them up.

@PauloCereda But our resident forest super-guru is online... although she hasn't been in chat it seems so I don't think @cfr will have any effect.

@AlanMunn Hm interesting approach! I will take a closer look! Thank you!

@AlanMunn To be honest, I was curious about using a remembered node. :)

@PauloCereda Hmm. Well I took it to be an XY problem... :)

@AlanMunn :D

@PauloCereda You mean Cardinal Fangs Quack then? ;-)

@ChristianHupfer "Nobody expects the Spanish Inquisition" LOL (I searched for it on YouTube)

@CarLaTeX I will poke you with the cushions ;-)

@ChristianHupfer ROFL

@ChristianHupfer you are mean. :)

@PauloCereda \let\mean\christianhupfer :-P

@ChristianHupfer ooh

@AlanMunn: I wrote a proper question. :) For some reason tikzmark does not work anymore. :)

So the Playboy guy died.

@samcarter: we could have a Playboy bunny duck. :)

@TorbjørnT.: thanks for your answer. :)

Anyone know what's changed recently in (u)pTeX?

@JosephWright That's the first time I hear this contraption. :)

@JosephWright The changelog lists some points in july concerning unicode blocks and otf support...

@TeXnician Saw that, but it must be more recent: shows up in our test suite

@JosephWright Ah, okay. I didn't know that there's such an active development for that branch of TeX...

@TeXnician Difficult to say as I can't read most of the docs ;)

@JosephWright Doesn't Google Translate speak Japanese?

@TeXnician I know that Japanese users do use pTeX/upTeX a lot: we had various issues with expl3

@JosephWright For me, Japanese is the natural, linguistic approach to xii.tex :)

@PauloCereda Sure, no problem. Perhaps Sašo (forest author), cfr, or someone else can shed light on what causes the issue with old code.

@TorbjørnT. Indeed. :) Or even Andrew Stacey. :)

@TorbjørnT.: see @UlrikeFischer's update!

@UlrikeFischer: very impressive, thank you! Never heard of \subnode...

@PauloCereda I'm one of the people that looks in documentations ;-). Actually a bit searching found similar answers: tex.stackexchange.com/questions/255472/…

@UlrikeFischer Oh my, now I feel silly! I looked for tikzmark, but haven't seen this \subnode idea.

@UlrikeFischer: do you think my question is a duplicate of that one?

@PauloCereda philosophical question. You are asking about two trees and wants to connect them, none of the other question asks about this, but naturally "use \subnode" is the answer to allmost all questions about "how to do something with a forest node from the outside"

@UlrikeFischer Now I don't know what to do! :)

@JosephWright that reminds me, not run the test suite for a while...

@DavidCarlisle Changes seem a bit odd but they are 'not us' so at least at present I've simply changed the expl3 outcomes

Just a quick question: Is there a way of having both, parskip=half and parindent in KOMA?

@PauloCereda both compact, expressive and readable?

@DavidCarlisle Exactly :)

@UlrikeFischer, @TorbjørnT.: my thesis thanks you. :)

@DavidCarlisle ^^ you might recognize the syntax. :)

@PauloCereda Buridan's Duck?

@UlrikeFischer ooh

@TeXnician not with an option but you can naturally set the values directly.

@UlrikeFischer Thanks. So I'll stay with the "hacky" approach.

@TeXnician You could naturally ask a question about it in some sensible place and then check the sources if an (undocumented) option texnicianwantsparskipandparindent appears.

@PauloCereda A bunny duck is already included - you'll just have to "dress" it and choose the colours.

@samcarter ooh

@AlanMunn ^^ the acknowledgement section of my thesis will be similar :)

@PauloCereda so at least there will be some text in it.

@DavidCarlisle you are mean

@DavidCarlisle ooh

@DavidCarlisle oh no

@PauloCereda sadly I couldn't find an "oh yes" :(

@DavidCarlisle :)

@DavidCarlisle do you thank me, too?

@Skillmon did you mean to ask @PauloCereda ? :-)

@DavidCarlisle yes, I did. Sorry for mispinging.

@Skillmon not disturbing @PauloCereda is a good thing, because if you ping @PauloCereda too often then @PauloCereda will never get his thesis finished and @PauloCereda's supervisor will complain

@DavidCarlisle Did you see? There is a new tabular package: ctan.org/tex-archive/macros/latex/contrib/xltabular

@UlrikeFischer I hadn't seen, interesting, might have fixed some of the "issues" I had with ltablex judging by the ctan description.

@DavidCarlisle But the code looks like you can get quite confusing results if you load the package not as first "tabular" package.

@UlrikeFischer I didn't dare look at the code (have head in hyperref, as you may have noticed)

@DavidCarlisle Yes, thanks. I have been looking at github.com/ho-tex/hyperref/issues/37. I think Ross is right, and hyperref shouldn't redefine \noboundary.

@DavidCarlisle 'ello

@UlrikeFischer I agree

@egreg get @UlrikeFischer to make a pull request:-)