@SeanAllred I thought I saw some prototype code from Bruno once about this too, but then he got caught up with other things (as we all do). Maybe he just discussed the idea.
@SeanAllred Can we play around with it for a while before an official CTAN release? This is the sort of thing that might make more sense merged into expl3 proper!
@WillRobertson I remember the conversation came up those few years back and the main argument against expl3 supporting such a thing is its own scope. It was difficult to come up with a typesetting-centric use-case for objects like this.
The template system's objects are similar, but inherently limited to nine fields. It definitely makes things needlessly difficult.
@WillRobertson Related conversation with a link to the LaTeX-L mail: vvv
In chat, @cgnieder just pointed me to xtemplate after poking around my own startup ltx3obj.
In section 3, the documentation states (at the time of writing):
An object type (sometimes just "object") is an abstract idea of a document element that takes a fixed number of arguments corresponding ...
@SeanAllred, @WillRobertson If I remember, Bruno suggests an object approach to allow us to cover variables that we've not built in. For example, we need to cover some kind of two-dimensional data format for e.g. reading CSV files. Of course, performance issues might lead to using different code (see how pgfplots reads data, for example), but at least we should explore a built-in generic type.
I wonder if it is recommended to add \! after \int to make the integrand little closer to the \int sign? Maple generates Latex this way. It add \! after each \int. Like this: \int \!{\frac {-{x}^{2}+2}{{x}^{3}+2\,x}}\,{\rm d}x but Mathematica Latex export does not add the \! and I never seen it before done this way.
@JosephWright: Seeing "xtemplate" reminds me of an issue with the documentation: For some commands (e.g. \DeclareTemplateInterface, \DeclareTemplateCode) the description of the arguments are missing.
@Nasser It depends on the font. I like the approach taken by the "cool" package -- it defines a command called \Integrate{f,x,a,b} which corresponds to \int_a^b f dx , but where you can customise a lot of the spacing and so on without having to edit every formula you write.
@UlrikeFischer That would be my fault I guess — pretty sure I wrote that documentation :)
@JosephWright IIRC there was also cool stuff regarding, say, storing a clist or seq inside a prop, and getting things in and out didn't require cumbersome multiple getting/setting commands for each "layer".
@WillRobertson I'm guessing that the code contains the arguments, but that due to some error they swallowed in the print. (It seems to affect only commands with macros description over more than one line).
@WillRobertson main issue I see, mirrors.ctan.org/macros/latex/contrib/cool/… is that one has to write \Sin(x) instead of \sin(x) and \Infinity instead of \inf and \PI instead of \pi, etc... to take advantage of it. Which means one will be coupled with this package for ever. Without this package, the code will not run unless one changes the code back to lower cases.
^^ that should be \Sin{x} and not \Sin(x), etc....
@Nasser Not hard to define your own commands similarly, though, if you want to break away from using that package. In fact, sometimes I do exactly that when I just need a single function or two.
@JosephWright not really, just pointing out that since the only affected characters (should be) previously undefined unicode slots, existing documents should not be using them so there should not be compatibility issues, and that unlike tex fonts system fonts will change under us anyway to track unicode (in parts) so making the default codes track unicode is the only option really.
@JosephWright yes they have a standard policy of never changing properties (so they can reject any request to do so) but sometimes they decide a bug is a bug and change anyway. Hence the (should be) but still I remain unconvinced that updating the codes is problematic unless someone can show me a real document that would ever have been affected by a unicode update.
670c670
< 029D;LATIN SMALL LETTER J WITH CROSSED-TAIL;Ll;0;L;;;;;N;LATIN SMALL LETTER CR
OSSED-TAIL J;;;;
---
> 029D;LATIN SMALL LETTER J WITH CROSSED-TAIL;Ll;0;L;;;;;N;LATIN SMALL LETTER CR
OSSED-TAIL J;;A7B2;;A7B2
@JosephWright I know but in general it sort of makes sense to aim for tex levels of stability in a macro layer over tex82, but if running over xetex or especially luatex I think you need to let go a bit and accept minor bug fixes of unicode properties as just ignorable changes. Since the system fonts also change get updated so metrics and available characters change without any change to the document, I really can't get too worried if one or two case mappings change.
@JosephWright If you mean the addition of case mappings, this is because the characters are new in Unicode 8; however case mappings are frozen after they’re added (not completely sure about that, can find references if necessary).
@Johannes_B Generally speaking, I agree with @DavidCarlisle that Unicode is much more stable than the Unicode-aware variants of TeX.
@Johannes_B Oh damn it, sorry ;-) Yes, I meant Joseph. I usually just type “@Jo”, and you’re before him in the alphabetical order, so you usually come first.
@Johannes_B Although maybe it’s also connected to the activity level, in which case it’s because you post more than Joseph ;-)
670c670
< 029D;LATIN SMALL LETTER J WITH CROSSED-TAIL;Ll;0;L;;;;;N;LATIN SMALL LETTER CR
OSSED-TAIL J;;;;
---
> 029D;LATIN SMALL LETTER J WITH CROSSED-TAIL;Ll;0;L;;;;;N;LATIN SMALL LETTER CR
OSSED-TAIL J;;A7B2;;A7B2
@ArthurReutenauer There are not many of them, but there are changes in the case mappings
@JosephWright, please see the edited version. — Alexey1 hour ago
Anyone fancy having a go at this? I get a different list of language so presumably a different language.dat file is in use, but I'm not sure why that would affect \@hash@
@JosephWright interesting that's the latex2.09 version of yargdef more or less (with a loop) rather than the 2e one which just truncates #1#2#3#4#5#6#7#8#9 :-)
I would like to have a Cyrillic version of TeX/AMSTeX. I have installed the cyrplain package, but further instructions are too long and technical for me, sometimes referring to README of other packages or suggesting to manually move some files, and the expected result of following them is not qu...
@NicolaTalbot any chance you could change \RequirePackage{etoolbox}\RequirePackage{etex} to \RequirePackage{etoolbox} in datatool? (etoolbox will load etex on old formats, and on new formats it doesn't:-)
@PauloCereda Don't be silly, Oxford colleges have people who arrive when required and serve tea and biscuits:-)
@SeanAllred, well, I didn't even know he used a duck... By the way, the Tufte-Jekyll theme is impressive, but it gives me itches to think that no proper italics were used.
@PauloCereda I'm sitting in my office after a boring department meeting, waiting for going to the station, catching my train and sipping my well deserved glass of wine.
@JosephWright Do you have examples of existing case mappings that were changed, as opposed to new ones? This is a little different.
@Joseph Anyway, the changes within Unicode will always much less consequential than the additions we need to make to unicode-letters.def in order to set up characters from unicameral scripts for hyphenation.
@PauloCereda So are you coming to the UK-TUG meeting or not? I didn’t follow.
@PauloCereda amoebic, amorphous... Well, using LaTeX in Brazilian universities and following supposed rules is like boarding a jet for a walk around the block. Last week I saw someone wanting help to disable hyphenation completely, and using Arial 12 for body type.
@JosephWright Anyway, what I meant was: in TeX we need to say \lccode"0905="0905 for a great number of characters in order for them to be considered for hyphenation, and that introduces an artificial case mapping (in TeX) that is entirely due to an idiosyncrasy of the language.
@JosephWright Some of these characters may gain real case mappings in later versions of Unicode, but most of them won’t (Hebrew, Arabic, Indic scripts - we already have many hyphenation pattern files for the latter). Hence the difference from unicode-letters.def to UnicodeData.txt will always (by my estimate) be greater than the differences from one version of UnicodeData.txt to the other.
@PauloCereda 1 hour drive (plus parking and stuff) more or less, but if I get a chance (like going to Darmstadt) I drive for an hour in the other direction to Birmingham which is a more sensible airport
@Joseph That must unfortunately be very common, I think. Disabling hyphenation is a classic request of newcomers to TeX who want to emulate whatever horrible style they’re used to.
@PauloCereda You’re more than welcome to sleep on our couch for a few days if you want.
@ArthurReutenauer Thanks, Arthur. :) I'm just looking at the prices because I need to see how much things cost in the UK (someone told me it's one of the most expensive countries in Europe).
@PauloCereda As for travel within the country, you can count on £30 for a round-trip train ticket from London to Oxford (that’s €40 or $50, very roughly). Probably less if you book early.
@David, @Arthur: on a serious note: I will try to attend. I will poke Brent and talk to him about some hints on travelling from Brazil to UK (Brent lived for many years in here). :)
@StefanKottwitz If you are there, please get in touch with @barbarabeeton with respect to templates. maybe @ArthurReutenauer wants to join the discussion as well. who did i miss?
@PauloCereda Laid down my weapons. i realized i cannot fight a battle i cannot win (though i tried to multiple times in my life).
Now i decided to stop fighting, and work along with templates. As this is (seems to be) the starting point of many many users, we need to take care of improving this initial state rather than complaining when it's too late.
@barbarabeeton It was Cedarville University in Ohio were I was an Associate Professor of Mathematics, Computer Science, and Physics. I retired a year ago.
@egreg -- european institutions are apparently much more sensible about having their students formally introduced to how to write proper academic documents. in the u.s., except for a requirement (which may not even exist any more) to demonstrate "freshman english composition" competency, it seems to be up to individual instructors to maintain standards. and from what i've seen, not all that many do (or are even capable of doing so, especially in technical disciplines).
@ArthurReutenauer -- having just received and edited a tugboat submission by an austrian secondary school student -- whose command of written english is better than that of some of my young relatives -- i'm not ready to back down on my contention. maybe i've just been privileged to see the "cream"; if so, i consider myself fortunate ... but i'm still mightily impressed.
@JosephWright Yes, I plan it! Not yet registered, as I'm involved in a project at work at that time, hopefully I can stay away some days to go to the meeting.
@clemens I have been discussing this here in chat on a short basis with @barbarabeeton and i also contacted Vel from the template site. There will be a discussion on the topic in Darmstadt, unfortunately i cannot attend. I'll try to write down some feelings and thoughts about it so that at least a few of my thoughts are getting into the discussion. I hope this will be a fruitful endeavour.
@clemens The very first thing to do is to help Vel to update existing templates and get obvious errors to die out.
Urgent! How to use Mac Texmaker to get bibliography? I tried biblatex (\usepackage[sorting=none, backend=bibtex]{biblatex}), \usepackage{bibtex} and biber but they do not work. What do you recommend?
Should I try natbib? I typed out the bibliography myself before without bib. file. My system seems to be installed thru TeX Live with MacPorts, which is bad and seems to lack biber. But I cannot change to MacTex now as the due date is coming. What is the solution?
@kwy just saying "they do not work" is not all that helpful and is unlikely to lead to correct answers. What doesn't work? Did you actually run bibtex or biber on the file? Describe the steps you took and any error messages you encountered.
I've browsed the forums and found a number of posts that have addressed this issue, but none of the solutions seem to work for me. I have the following script that I just copied from the bibtex home page to get familiar with it. Instead of the citation number I get a question mark. I compile usin...
In chat, @cgnieder just pointed me to xtemplate after poking around my own startup ltx3obj.
In section 3, the documentation states (at the time of writing):
An object type (sometimes just "object") is an abstract idea of a document element that takes a fixed number of arguments corresponding ...
