TeX, LaTeX and Friends

12:39 AM

@AlanMunn Do you have a working example? I mean, obviously, you can do that. But I don't think it will be pretty!

@DavidCarlisle Maybe people understand it?

@AlanMunn You need manual intervention is the problem. For example,

\documentclass[border=10pt]{standalone}
\usepackage[linguistics]{forest}
\begin{document}
\begin{forest}
  for tree={
    nice empty nodes,
    calign=fixed angles
  }
  [C,calign primary angle=45
    [, before computing xy={for children={l'=0pt} }
      [, grow'=west, [D ] [E ]
      ]
      [, grow'=east, [A ] [B ]]
      ]
    ]
  ]
\end{forest}
\end{document}

@AlanMunn I don't think calign primary angle=45 can be doing anything useful. Is it necessary?

If I remove it, it makes no difference to the result at all. @AlanMunn

Alan Munn

@cfr Yeah, that was probably there from a previous version.

cfr

@AlanMunn There doesn't seem to be a nice package for drawing unrooted trees. That is, one which doesn't require the clunky verbosity of the standard TikZ stuff. And the graph drawing stuff I can never really get to grips with.

@AlanMunn The thing is, how does this extend to the other examples? It isn't nice ...

Alan Munn

12:54 AM

@cfr Well my intuitive idea was to make styles that would do the job. Since the basic structure is still a tree, the syntax of forest should still make things simpler.

cfr

@AlanMunn Forest doesn't do cyclical placement the way TikZ does. That makes it tricky. I guess I don't see how you'd handle the others in anything like an intuitive way. How were you thinking of the trees in the edit, for example?

Alan Munn

@cfr Well the intuitive idea was that there's always one node that can be used as a root node, and the rest of the tree simply depends on growing the dependents from empty nodes in the right direction. But maybe that's not practical.

cfr

@AlanMunn You are going to lose all Forest's advantages, I think, and just get the disadvantages.

Alan Munn

@cfr Maybe. :)

@cfr But thanks for your example, that helps a lot. I'll play around to see if what I have in mind could possibly work.

cfr

@AlanMunn Forest won't let you have the angles of growth you want, I think. Otherwise, it would make sense. So you have to move things later. But that means you have to take care to avoid overlapping etc. and ensure there's enough space and not too much and so on. But then it isn't clear why use Forest.

@AlanMunn Not growth. I mean edge angles. What you want is to be able to say is that you want edge angles of 120 or 90 or whatever, but that's not going to work.

Alan Munn

1:12 AM

@cfr I see. My impression from the other trees was that you wouldn't need to do that too much. Most of the trees seem to have straight lines or regular angled pairs.

cfr

@AlanMunn Then maybe I just don't understand your idea.

Alan Munn

@cfr I'll play around and show you what I end up with. Might take a day or two though.

cfr

@AlanMunn Not once you turn them into rooted trees, though. Then you have cases where you need one to go up and end up with y equal to !uu.y.

But its sibling needs to be at -90. So I guess you can with a growth direction of -22.5 or something?

cfr

1:26 AM

@AlanMunn Fixed angles may not be what you want:

Though deleting it only makes things worse ...

Alan Munn

1:45 AM

@cfr Yeah, you can get some crazy stuff.

But I did manage this:

Werner

7:00 AM

@JosephWright: mod (and perhaps other stuff) seems to be excluded from xfp. Is there a reason for this? Or is the documentation just incomplete? It - mod seems to exist when looking at interface3 (in the form \int_mod:nn).

Joseph Wright

7:19 AM

@Werner Don't think it's on the list of keywords which Bruno implemented, I guess as it's not really an fp function (it returns an int ...)

Joseph Wright

7:47 AM

@Werner You can use of course n - m * floor(n/m)

Morning @DavidCarlisle: did you get a chance to look at the driver stuff?

@DavidCarlisle Yes :)

Transcript for

TeX, LaTeX and Friends

$TeX - LaTeX$ TeX, LaTeX and Friends