@Johannes_B Exactly. The implication there is that 'this idea has merit, but I don't have time to work with it right now -- I'll catch up with you later about it'
@JosephWright -- location not a mystery; a proposal for toronto was accepted by the tug board some time ago, and unless something really untoward happens, that's likely not to change. dates ... that's another matter.
@JosephWright -- true. i'll ask karl to take a look, and consider making that much of an update. (he may not be willing; things have happened to change locations in previous years, though in locations much more volatile than toronto. now, if it were quebec, ...)
@SeanAllred @SeanAllred I'm not sure what you mean about LaTeX development being open. As @DavidCarlisle says, all of the code is in a public SVN and LaTeX-L is there for discussion. The team list is mainly useful as the likelihood of a response is a bit higher, and there is a certain amount of freedom.
@SeanAllred -- probably a good idea, but i'm a github luddite, so i'd be at a disadvantage. (for that matter, i still haven't mastered the technique of preserving a link to a particular item in this chat, and i haven't figured out where and how to ask the question how to do it. when i've found it, then i'll add the instructions to my handy-dandy cribsheet, so i'll know where to find it again. for some things trying to google it is too sweeping.)
@barbarabeeton You mean getting URLs for specific messages? Mouseover a message and click the downward pointing arrow on the left side. In the popup you'll see permalink.
@SeanAllred If you can get yourself to Milwaukee, you're welcome to hitch a ride with me from there. I've not decided if I'll drive or fly (there is of course the possibility that I won't be able to attend at all), but I don't mind either way and the drive would be better with company.
@TorbjørnT. -- oh, thankyou, thankyou, thankyou! (obvious, once it's pointed out. there are still some sites/apps where it takes me half an hour to figure out how to log out. i learned about computers with punched cards. i don't grok guis.)
I wonder how LaTeX-L will react when I send messages to it, now. At the moment, I just have everything @seanallred.com redirecting to my gmail (since I no longer use a hosting service, just DNS)
My fault for not wanting my gmail out and about, though :)
@SeanAllred no but might be a good place for formulating ideas on a specific piece of code before taking general issue to latex-l
@SeanAllred you think email is out dated so you can't be helped:-) (must admit I have dropped off c.t.t, although at one time I was posting a reasonable proportion of its messages)
@SeanAllred click on the subscribe button, or send a message containing "subscribe" to the list address, probably easier than signing up here, to be honest.
@SeanAllred new depends on the order, this site is a lot newer and harder to use than a mailing list, you have to learn the markdown syntax, and question/answer/comment conventions. a mailing list you just send mail.
@SeanAllred you can subscribe by mail, or by that web form.
@JosephWright, @DavidCarlisle -- here's an example where \protect isn't good enough, but \string works. \string isn't mentioned anywhere in lamport, and only in very specialized situations in the companion. any chance of fixing the underlying problem so that the "published approaches" work?
I want to have an entry $r_{\infty}$ in the index. When I mention it inside a footnote, there's an added whitespace ($r_{\infty }$). This results in multiple entries in the index.
Here is a minimal working code to reproduce the error:
\documentclass[a4paper,10pt]{article}
\usepackage[utf8x]{inp...
@DavidCarlisle -- is there any likelihood of this being remedied in the upcoming releases? i'm getting back to a tome on indexing i've been working on for a couple of years (it's up to 13 pages, "including index"), and i'd like it to be correct and authoritative when it's ready for release, probably in a couple of months. (it will go onto ctan, and from there into tex live. also recommended from the ams web site, although most likely not one of the "ams-named" packages.)
@barbarabeeton I can ask rest of team but I'd guess not. To make spaces consistent you have to give up on reading the argument verbatim, which would mean anyone with \index{\foobar} would get a very bad experience on an update.
@DavidCarlisle -- okay. i'll just muddle around with examples that have \index in a footnote until i have one that looks reasonably clean, and document that. (this tome contains (almost?) everything i've ever seen in ams production that causes problems in indexes, and is intended as a resource for authors who complain that they're having indexing problems, as well as guidance for ams production editors. maybe i should let you (make you?) read it before i set it free.)
@barbarabeeton the above makes \index not read its argument verbatim so then it will work in footnote just as at top level, but you can only do that if you are indexing "safe" things, to index % or \zzzz you'd have to work a bit harder
The argument of \index is read verbatim as that;s normally what is required. You don't want that, so the easiest thing is to put \index inside any command. \mbox would do:
\documentclass{article}
\usepackage{makeidx}
\makeindex
\begin{document}
\section{First Section}
\mbox{\index{first \ara...
As an alternative mechanism
An alternative to egreg's solution I was going to post
\newcommand\hmm[1]{#1}
Then you can use
8\hmm{\index{....}}
You could redefine \index to always do this, this just always prevents the verbatim nature of \index from activating, so you have to be a bit more...
\index reads its argument semi-verbatim, which is why the basic \index prevents \id expanding, but within a macro the catcode changes have no effect.
These days you can simplify the expansion control using \detokenize something like
\documentclass[a4paper,12pt]{memoir}
\usepackage[x11names]{xco...
@barbarabeeton ^^^
@barbarabeeton actually the detokenize version in that last one is probably generally safe,
@SeanAllred On the Reddit business, what's the archiving situation? With email we can all keep all of the mails (I don't as I treat email as disposable, akin to a phone call, but others keep a lot): there's not dependence on the continuing existence of the server
@SeanAllred yes sure it could have been that, and to index \foobar you could (have) used \protect\foobar but having had \index read its argument verbatim since the beginning of latex, certainly over 30 years, changing it now will break something.
@JosephWright I'm not a mod of any subs (that I know of?) so I'm not sure what their capabilities are
@DavidCarlisle certainly :(
which goes back to why I (personally, philosophically) think breaking changes will always be necessary in new major releases. Most of the rest of the software world realizes this :)
(The software world being a proper subset of the real world.)
@SeanAllred but perhaps this "new" \detokenize thing could offer a way out making the argument inert without using catcode changes, as noted in the answer above.
@SeanAllred most software isn't archival in the way documents are. It's OK to have breaking changes in an editor or operating system, you learn to cope, but documents are supposed to last for thousands of years. papyrus does.
@JosephWright Ok I'll give it a couple of days. (But given that the OP was "desperate" I suspect she's solved her problem by now, since she hasn't commented back at all.)
\input luatexbase-attr.sty %
\newluatexattribute\foo
\attributedef\attributezero=0 %
\directlua{luatexbase.new_attribute("foo")}
\directlua{
for i,j in pairs(luatexbase.attributes) do
texio.write_nl(i .. " " .. j)
end
local nt = newtoken.create("foo")
texio.write_nl(newtoken.create("foo").mode - newtoken.create("attributezero").mode)
}
\show\foo
\bye
@DavidCarlisle Yes, that was for testing
@DavidCarlisle Two attributes, same apparent name, different numbers, different accessibility :-(
@DavidCarlisle I'm thinking they've made a mistake in allowing two routes to define attributes: as they can be set up using \attributedef I think they need to have an allocator written only in TeX or at least not also trying to track them in Lua
@SeanAllred Practically I can do the following: (i) show that the posted code snippet doesn't produce an error and (ii) reproduce the reported error by removing a // from the code (iii) explain why the line reported in a TeX error isn't always accurate. Given that this particular error is likely to be quite common for this package, I think this would be a useful answer.
@DavidCarlisle I'm just working out how to handle this: presumably accept that any attributes defined from Lua are not tracked other than by the defining code
@SeanAllred And the question does make sense to me. "I'm using this package and I got this error. I think this is the code responsible. How do I find the error?"
@JosephWright I know:-) I suspect though given some comments in the manual that luatex will eventually get proper token support allowing you to do \def from lua, in which case the lua allocation could define the tex csname, but as is I don't really see it's possible
@SeanAllred Sure, but the error is thrown by the package in question. Anyway, I'll wait a day or so. I don't think there's any rush to close it right now. My main point was that given the common nature of that error with the package an answer of the sort I outline would be useful.
@JosephWright alternative would be a different syntax on tex side \newattribute{foo}, \setattribute{foo} etc then you could keep the name/number mapping in a toks register that could be updated from lua, but I don't think it's worth it, really.
@DavidCarlisle Which I would be a strong opponent of. If one wants to fix breqn I can agree, but inflicting it to every user would be completely wrong.
@DavidCarlisle he has to wait long time then :) since I never wrote a single package in my life before, and the thought of even looking inside an .sty file and seeing all those catcodes and funny looking characters scares me to no end, I do not think this will happen any time soon.
I want to have an entry $r_{\infty}$ in the index. When I mention it inside a footnote, there's an added whitespace ($r_{\infty }$). This results in multiple entries in the index.
Here is a minimal working code to reproduce the error:
\documentclass[a4paper,10pt]{article}
\usepackage[utf8x]{inp...
