« first day (2860 days earlier)      last day (2072 days later) » 

12:02 AM
@cfr There is a forest question which already has a nice answer, but perhaps it can be made even nicer by avoiding \strut and using some baseline magic instead.
 
cfr
12:25 AM
@marmot You need \strut or something else, anyhow. baseline aligns the tree relative to the baseline. It doesn't align the nodes in the tree. So even if you use baseline to align the roots, you'll get things misaligned below. However, there is no need to abuse font like that and every reason to avoid such an antiquated solution :-). Certainly it is worth an upvote. After all, it works without doing anything too heinous. But there are much more elegant ways. Especially, using edges...
... is not just easier. It is also less fragile and more flexible. folder behaves (mostly) well with other Foresty things, whereas the old approaches certainly did not. (They were the best anyone came up with at the time, but now we have more anchors and more robust options.)
@marmot Also, the font abuse will obviously break if a user does font=\bfseries or something for a node. People seem not to do this, though.
 
12:58 AM
@cfr That's why I said we need your magic. ;-) (\strut is also not a good idea if the nodes should run over several lines, I guess... but again that doesn't seem to be the case here.) However, I did name the top nodes by hand, alias=X, and then said baseline=(X.base) in the tikzpicture code, that also seemed to work.
 
@samcarter I guess one needs also something like a tikzlings.meta. ;-) Just realized that I do not have precisely the same 3D switch as you have. The difference are just square brackets. I'll be happy to change my code to your conventions (but perhaps the one I am using could be also OK...) ;-)
@CarLaTeX See, these are the side effects of @samcarter's packages. ;-)
 
cfr
1:29 AM
@marmot You could use a relative name, to avoid naming them. However, if you do it that way. the level 1 nodes won't be aligned because the depths don't match. If it were a single tree, it would be different, but, with different trees I can't see any way to avoid forcing a common height and depth if they're meant to align.
 
1:47 AM
@cfr Thanks! I already saw and upvoted your nice answer...
 
 
2 hours later…
4:12 AM
I do not understand your question but if you want to draw growing dots you could do something like this
\documentclass[border=3.14mm,tikz]{standalone}
\usetikzlibrary{decorations.markings}
\tikzset{growing dots/.style={decorate,decoration={markings,
mark=between positions 0 and 1 step 6pt
with {
\pgfmathsetmacro{\myradius}{0.1+1.5*\pgfdecoratedcompleteddistance/\pgfdecoratedpathlength}
\draw [fill] (0,0) circle (\myradius*1pt);}}}}

\begin{document}

\begin{tikzpicture}
\foreach \X in {-90,-86,...,90}
 
 
2 hours later…
6:29 AM
@StevenB.Segletes @samcarter vvvv
All that is cute does not glitter,
Not all the sweet ones are lost.
Ducks who are kind do not wither
And marmots are not reached by the frost.
A Koala and maybe a coati shall be woken
And more out of tikzducks shall spring.
Patched up will be code that is broken,
And the cute ones can always be King!
5
@marmot ^^^ marmots and koalas are mentioned too ...
 
 
2 hours later…
8:37 AM
@marmot that's not a koala it's a marmot with grey facepaint
 
9:32 AM
@UlrikeFischer Bravo! I tip my hat.
 
 
1 hour later…
10:47 AM
@marmot I would like to note that I'm personally not a big fan of the 3D versions.
 
11:02 AM
@UlrikeFischer You got mail :)
 
@CarLaTeX answered. Did you saw the poem above?
 
@barbarabeeton I'll put the sloth on my list of ideas!
@marmot Please don't worry about the definition of the 3D switch, I'll unify this - probably over the weekend, don't have time right now.
@UlrikeFischer Great poem! Would be a great addition to the duckuments package :)
 
11:36 AM
@UlrikeFischer hmm, can almost get that to work. I have a tree with a lot of fonts, and one map file. The I have the map file in web2c/updmap.cfg in the tree. ` kpsewhich -all updmap.cfg` finds it just fine. But I still have to run updmap --sys --enable Map=file.map to access the fonts it covers, that was what I'm trying to avoid. It is rather strange also, when I enable it it is written to a web2c/updmap.cfg in another tree I've added
Never mind, if I manually remove the Map line from the "wrong" cfg, updmap --sys --listavailablemaps | grep file.map now lists it as being enabled, and from the right updmap.cfg
 
 
2 hours later…
1:40 PM
@UlrikeFischer Oh the duck of the ring
 
1:50 PM
I have a plain bibliography (\begin{thebibliography}) set in \small type. The spacing between entries is too large for my needs. What is a simple way to reduce it?
It seems like something people must have asked before. I don't know why I can't seem to find it.
 
82
A: Condense the space between bibliographic entries

boycott.se - yo'As Andrew says, you can use natbib and then reset \bibsep: \setlength{\bibsep}{0pt plus 0.3ex} Without natbib, you can set \parskip and \itemsep to zero (they are responsible for the vertical space). It is advised to add some glue-stretch that can be used if necessary (this is what plus 0.3e...

 
@boycott.se-yo' Excellent, thank you!
If I boycott SE, I don't know where else I could go though.
 
@Szabolcs well, I haven't been asking many questions, especially not recently.
 
 
3 hours later…
4:28 PM
Hello all. Quick question (maybe for the main site, I don't know): Is there a way to define commands in a style file, using utf-8 coding (for human readability), in such a way that it will work from a main file using a different encoding?
E.g., \def\euclid{Εὐκλείδης} (contrived example)
 
4:43 PM
@HaraldHanche-Olsen {\inputencoding{utf8}\gdef\euclid{Εὐκλείδης}} but I'm not sure I'd do that. or safer perhaps {\inputencoding{utf8}\protected@xdef\euclid{Εὐκλείδης}}
@HaraldHanche-Olsen actually you'd need the second version, first wouldn't work
 
5:00 PM
@DavidCarlisle Ah yes, I too had tried something along the lines of the first one. Didn't work. I might experiment with the second one; thanks.
@DavidCarlisle Ah yes, I too had tried something along the lines of the first one. Didn't work. I might experiment with the second one; thanks.
@DavidCarlisle Did my response get posted twice? I see it twice here; it came up with something about timeout with three choices. I hit the “retry” one.
 
5:47 PM
@HaraldHanche-Olsen deja vu is normal here, don't worry:-)
 
 
1 hour later…
6:58 PM
Experts: Is there something other that listing package, which allows one to have verbatim inside a table or a longtable cell, with fixed width columns, which will automatically wrap verbatim to fit the column width such as |p{3in}|p{2.5in}|? I am using listings now and it does that, but it is causing problem with tex4ht. What else can I try? \verb|| does not work, since it does not wrap verbatim to fit.
 
7:46 PM
@marmot Your code it is perfect=+1000 points!!!!! You have understood my request. Thanks a lot :-). Best regards.
 
 
2 hours later…
9:21 PM
Why doesn't my log file show any contents if I use \setbox0\hbox{Foobar}\showbox0? Only output is \box0= [].
 
@Skillmon \showboxdepth and \showboxbreadth are -1 by default in latex
 
@DavidCarlisle Why that? That doesn't seem reasonable to me.
 
@Skillmon why not (it makes some user-level overfull box messages a lot more reasonable) if you are coding and want to debug add \showoutput for example which sets them both to maximumn and also echo to terminal not just the log
@Skillmon one of the earlier changes made in 2e from 2.09:
%    In \LaTeX{} we don't want box information in the transcript
%    unless we do a full tracing.
%  \changes{v1.0g}{1994/04/28}{Turn off overfull box tracing in log}
%
 
@DavidCarlisle Can you also tell me why a \write is not executed in a \setbox (I tried to call \label in \setbox0\vbox{...} but it didn't make its way to the aux file)?
 
@Skillmon because a non-immediate write doesn't write anything it just leaves a whatsit node in the node list being built which will cause a write when (if) that list is shipped out. So if you \copy (\usebox) that box 5 times it will write 5 times but if you never use it the write never happens.
 
9:33 PM
@DavidCarlisle Is there anyway to smuggle a non-immediate write into happening without outputting the box?
 
@Skillmon no
 
@DavidCarlisle sometimes TeX really tries to make things harder than necessary:(
 
@Skillmon the only reason to use a non-immediate write is to get the page number and a non-used box doesn't have a meaning for page number
 
@DavidCarlisle one might think so. But it would've been useful for tex.stackexchange.com/questions/448521/… (though the answer by Werner does work, I think it would've been more elegant in another way, except DEK didn't think so).
 
@Skillmon for example what page number would you want \label to have written?
 
9:38 PM
@DavidCarlisle The current page. The same problem does exist in Werner's answer. Of course the page number is non-sense in a situation like the one in that question, but that would've been the case anyway.
 
@Skillmon but the whole point of a non immediate write is that "current page" means the page on which this is shipped out.
@Skillmon if you want to base an answer around boxing the figure, inside the box do \let\zz\write\protected\def\write{\immediate\zz} so that the \label still happens.
 
@DavidCarlisle I forgot the \protected, I tried it without and lets say the result was non-optimal :)
 
@JosephWright would be proud of me for being so modern
 
@DavidCarlisle Thank you very much for your help.
 
10:05 PM
@Skillmon psfrag solves a similar issue by writing the box offset outside the page area \rlap{\smash{\kern100cm\usebox0}}} or some such
 
10:33 PM
@DavidCarlisle Can't be used because of unforeseeable user input. What if it is used like text\par\hiddenfigure\par text? It would give a blank line.
 
@Skillmon that's just details:-) have a look at (for example) how showkeys tries to add the key printing to the margin without affecting the layout even though \label could be just about anywhere.
 

« first day (2860 days earlier)      last day (2072 days later) »