hi, could someone please tell me why this does not work: pastebin.com/Ck2cfARZ It's nearly exactly like the code on page 140 of the TikZ manual but I somehow don't see the forest for the trees.
(I don't think this warrants its own question since it's too specific.)
@egreg Hats off to the improved welcome messages (your most recent one on the Quantum Circuit question being most examplary). ;-) Is there a thread on TeX.meta on this or is this all improvised?
LaTeX3 question: This might not be the place, but I just saw in the L3 list the new \tl_upper_case:n(n)… I'm not at the level of the developers, but since I've read many times those “layers” Frank distinguish (and, in a similar way, so do I), but it certainly shocked me a little: shouldn't \tl_upper_case:n(n) be in another layer? I think that's a different approach, more related with the end user and I'm surprised it's just put together with other \tl_… functions.
@Manuel yes well probably for text level uppercasing there will probably need to be functions at a different level but you often need a basic level of case change awareness even at the lowest level for normalising tokens or filenames or whatever,theese are expandable and Unicode aware versions of \lowercase so designed to fit somewhere near the bottom of the stack (but these things are never absolute:-)
@PauloCereda By the way, I enjoyed my glass of wine on the train. Some of the waiters that come to serve passengers don't even ask, when they meet me on the evening train. ;-)
@Manuel We need something to implement a higher-level function with: beyond some bootstrapping the aim is to use documented interfaces as far as possible. It's quite tricky, but I'm yet to find a 'better' place for this. (We've talked about \text_..., at least for thought experiments. Still doesn't quite seem right: suggestions quite welcome.)
@Manuel Functionality seems required, so the question then becomes how do we do such things. One problem is that 'case changing' is a 'text' operation but also does seem 'natural' for doing something like \tl_set:Nx { \some_case_function:n { tokens } } and to end up with stored case-changed data.
@Manuel Also note that case changing isn't really 'design layer' or similar: if we are talking uTF-8 input then most languages have a well-defined mapping here. (Of course, input encoding is another thorny issue!)
@Manuel One problem at present is we expose \lowercase and \uppercase, and they are really not the idea tools for case changing. What we need is a way to separate out case changing from case folding from 'weird token' creation. That's again tricky.
@DavidCarlisle Any view on whether we need some separate module for 'text manipulation'? Might be a better long-term approach.
@Manuel Note in particular that this stuff is currently in the part of expl3 where we say 'experimental': may yet turn out not to be the best 'home' for this code
@DavidCarlisle @JosephWright Okey. I see it's still in l3experimental.
@JosephWright I saw just now that there was another message talking about that. By the way, how does that list work? It's the first time I ever see something like that (a list), and I don't understand what it is exactly.
@DavidCarlisle It's older than the web, and I happen to be younger. That might be the reason we (lists and me) never met before :) Register at listserv
@Manuel gmane and google groups are (as far as I recall) a mail interface to newsgroups such as comp.text.tex which were originally (and still with the right client) accessed via netnews (nntp rather than smtp) but basically same idea.
@JosephWright not sure:-) tempting to wait until more higher level things sit on top to take a more informed view of how to split up the functionality but by then it may be too late to change the names...
After upgrading to TeXLive 2014 compiling my TikZ pictures prepared with the hobby library stopped working.
Consider the following MWE:
\documentclass{article}
\usepackage{tikz}
\usetikzlibrary{hobby,positioning}
\begin{document}
\begin{tikzpicture}[%
scale=1,
% circlenod...
@percusse I was just looking:-) but replacing on e ifundefined is probably easy but if the whole package needs updating may be harder I may look a bit more but Joseph's a lot closer to that stuff
@DavidCarlisle My take too: my feeling is we can't leave 'all this stuff' until 'later', but we do have to be ready to change stuff. Provided we give notice and have proper documentation, that should be OK.
@Manuel I guess @DavidCarlisle covered this :-)
@percusse Some code wasn't actually working, as we didn't have tests. The feeling was as it was almost all bust we could just drop it: perhaps this wasn't quite right
@JosephWright @egreg is the honourable thing to tell Peter he can generate a list of m's with \romannmeral or to give a lecture about l3 levels and these being all wrong? tex.stackexchange.com/questions/187362/…
@JosephWright isn't it good that xparse frees the programmer from csname and expandafter wrangling:-) \expandafter\NewDocumentCommand\csname#1\expandafter\endcsname \expandafter{\romannumeral#2000}{%
After upgrading to TeXLive 2014 compiling my TikZ pictures prepared with the hobby library stopped working.
Consider the following MWE:
\documentclass{article}
\usepackage{tikz}
\usetikzlibrary{hobby,positioning}
\begin{document}
\begin{tikzpicture}[%
scale=1,
% circlenod...
