I’m still reading John Burnside’s The Music of Time: Poetry in the Twentieth Century (see this post), and I’ve gotten to the chapter “La razón poética,” about the Generation of ’27. Burnside has an admirable desire to focus on women poets, and this chapter has a section on Ernestina de Champourcín (the name is apparently from Provence, and the -ou- is pronounced /u/). On p. 196 we find the following:
To live, and to welcome, not just the unknown but that which no mortal being (nadie) can know. It is, perhaps, tempting to see in such work the language of exile, where the love of home-place, and of those who remain there, is tainted with bitterness and longing for what is forever lost (and, as exiles soon learn, once the home-place has been corrupted, the only remaining option is to carry on, as a very different writer puts it, ‘boats against the current, borne back ceaselessly into the past’. That past, however, can never be regained.
I don’t know whether the average reader notices these things, but as a copyeditor I can’t help seeing that the parenthesis before “and” is never closed; what’s especially surprising is that there’s no place to put an end-parenthesis, since “That past” refers to the immediately preceding “the past,” and you can’t really do that across the barrier of a parenthesis. I wouldn’t go so far as to say I’m outraged or disgusted (not being from Tunbridge Wells), but it doesn’t sit well with me. This is what happens when you get lost in the weeds of a rambling bit of prose, and the publisher doesn’t pay somebody to notice and fix it.
That would be a good title for a collection of poetry.
A child practiced a piano piece, but got bored and ran out of the room without striking the last note. The father, who was sitting in a comfy chair reading, waited uneasily for a minute or two, then sighed, heaved himself up, and played it.
Seems only a step away from the frequent misplaced parentheses. Okay, maybe only frequent in unedited prose, but misplaced parentheses do happen in edited prose too. Though usually they do at least get closed.
I’d at least replace the parenthesis with a period. And then break up the long, wandering sentence into smaller pieces.
What’s the Tunbridge Wells reference?
I wonder why the compiler didn’t throw an error…
What’s the Tunbridge Wells reference?
Disgusted of Tunbridge Wells.
There will be a sequel to the book, in which the parenthesis will be closed.
(Or maybe not. There’s always the chance that they’ll make it a trilogy.)
I suggest this emendation:
…forever lost and, (as exiles soon learn) once the home-place has been corrupted …
It’s still a rambling sentence, but at least it is a sentence, and the back-pointer from the next sentence will not be blowing in the wind.
An open parenthesis activates some kind of program in my mind that pokes me intermittently until it’s resolved. I had fun with this tension on Twitter once by posting something with an unclosed parenthesis. Multiple rapid replies supplying the closing parenthesis, some accompanied by an ironic rebuke, suggest how much it bothers people. My favourite reply was: “Close all parentheses or you’ll let in the ghosts.”
Lewis Thomas’s essay “Notes on Punctuation” plays on the tension of stacking open parentheses:
)
I am very sensitive to proper bracketing, both in reading and especially in my own writing. Having matched opening and closing delimiters is most typically an issue in mathematical expressions, rather than in running text, but I have basically the same reaction to seeing mismatches in either case. With the math, there is more to matching the delimiters than just having the same number of opening and closing parentheses, since there can typically also be square brackets and sometimes French brackets.
A lot of physics journals actually have the same conventions for parentheticals in text as for grouping in mathematical expressions. For example, the Physical Review journals call for nested delimiters to follow the pattern { [ ( ) ] } in either case,* and I follow that convention in own work. I actually have a tendency, when writing an expression that turns out to have more nested more levels of nested delimiters than I had initially anticipated, to go back and adjust any delimiters that I have already written, so that the { [ ( ) ] } pattern is there—even if an expression might be just as readable with ( [ ( ) ] ) nesting. Of course, many people who write mathematics do things differently. A colleague (who does a lot of work in Mathematica—where parentheses are used for order-of-operations grouping, square brackets for arguments of functions, and French brackets for lists), uses parentheses for grouping almost exclusively.
* In math, if more than three delimiters are used, the American Physical Society style says the three kinds of brackets are supposed to repeat, but with the outer versions in boldface. I have avoided ever using more than three nested expressions in text,** so I have never needed to grapple with the possible necessity of boldface parentheses in that situation. Nor have I ever needed more than six levels of delimiters in a math expression, which is good, because the style guide is silent on what to do in such a situation.
** By far the most common reason that I might need nested delimiters in text is actually when I have short mathematical expressions in the running text. For example
As professional copy editor, you might be interested in typesetting languages such as
https://en.wikipedia.org/wiki/TeX
(created by Donald Knuth in the late 1970s), or its more recent descendant
https://en.wikipedia.org/wiki/LaTeX
which are used almost universally now in mathematics and physics — at least, in languages using the roman alphabet. They check parentheses closures automatically, and have revolutionized large parts of the scientific publication industry. In the opinion of many they may be the greatest advance in written communication since the Phoenicians; it is likely that all further advances will be built on this foundation. It is for example now possible to insert
https://en.wikibooks.org/wiki/LaTeX/Hyperlinks
in .pdf documents, which send you directly to online references. Something in these fields not written in one of these languages might as well be hand-written on lined paper in purple crayon; cf Scott Aaronson’s
blog `Shtetl-Optimized’:
Ten Signs a Claimed Mathematical Breakthrough is Wrong.
1. The authors don’t use TeX. This simple test … already catches at least 60% of wrong mathematical breakthroughs….
In the opinion of many they may be the greatest advance in written communication since the Phoenicians; it is likely that all further advances will be built on this foundation.
As they say at Wikipedia: “of many [who?]” and [citation needed]. If I were a mathematician, I presume I’d use LaTeX (despite the pretentious name and stupid suggested pronunciation), but nobody who’s not a mathematician need concern themselves with it, and their lives will be none the worse if they never hear of it. It’s possible, and even productive, to promote a good thing without this kind of hyperbole.
Also Computer Modern is butt ugly {says me} but alternatives like Lucida New Math seem to cost money. That’s not Leslie Lamport’s fault, of course, but it does keep me from using LaTeX for everyday writing even though it would be compatible with my habits in general.
but nobody who’s not a mathematician need concern themselves with it
Or people who need mathematics in their job, e.g. scientists whose work involves statistics. Still, it’s a relatively small subset of humanity. 🙂
Well, TeX is useful if you want to write poetry in Quenya for example. I realize that what I said sounds hyperbolic, and I don’t mean to be contentious, but I suspect these tools will grow in importance with time, as populations become more numerate, cf CP Snow’s `two cultures’ thesis. Even nontechnical prose can gain from hyperlinks, see for ex
https://eliotswasteland.tripod.com/
In any case this is a big deal in the scientific publication industry: university libraries, for example, struggle with the cost of academic journals
https://www.theguardian.com/science/2017/jun/27/profitable-business-scientific-publishing-bad-for-science
which has led to the creation of much less rapacious online scientific journals edited by the academics themselves. The current battlefront has moved from math/physics to biology and medicine (in which peer review is economically important, for example, for drug licensing), where old guard publishers struggle to maintain their stranglehold. Law may be another example; these fields don’t need the technical advantage of something like TeX.
@jm
I would have thought law could benefit from hyperlinks to precedents and to legislation, if these have been digitised and made available (even with a membrrship or consultation fee). Maybe JC has an opinion.
wld sy tht th grtst dvnc n wrttn cmmnctn snc th Phncns rmns th nvntn f vwl symbls (bt yr mlg my vry.)
Spcsbtwnwrdsrprttysfl, t.
pncttn nd cptlztn prhps nt s mch nlss yr th nd wnt t wrt bt scnc
@DE I stand corrected, thankee kindly!
You don’t need LaTex to put live hyperlinks in pdfs — you can even do it with good ol’ MS Word.
LaTex is indeed very handy for math stuff but it’s far more than most people need. Again, MS Word’s equation-writing thingummy has many limitations but it will deal with most ‘everyday’ case (whatever than means).
During my brief and abortive attempt to learn computer programming, one thing I did learn (repeatedly) was that in programming a failure to close something you’ve opened isn’t just clumsy or inelegant or illiterate, it often makes your whole program unreadable and essentially non-existent.
I meant to write something about how useful LaTeX is for tracking delimiters. Actually, LaTeX itself does not do this, except for delimiters that are supposed to be sized vertically to match the expressions they surround. The LaTeX parser looks at the height of everything between a \left( and \right( and makes the parentheses slightly taller. If the \left( and \right( commands are unmatched, it gives an error when compiled. This does not do anything to make sure the delimiters are of the same type though. A \left[ can be closed with a \right|, but this is actually a desirable feature (and a practically necessary one when dealing with multi-line formulas).
Most people writing in LaTeX now use front end word processors, and these can generally track bracket matching for you everywhere in the text. This is one of the ways these front ends make the LaTeX source easier to read; they can also do things like color code the LaTeX commands in the editing window, which makes it easier to follow what the code is doing and to spot typos. The front ends also make running the LaTeX executables just a matter of a button press, as well as installing any add-in packages that you want to use automatically.
The ease with which LaTeX manuscripts can be processed is one thing that helps keep publishing costs down in mathematics and physics. The American Physical Society journals stopped charging publication costs for manuscripts submitted using LaTeX long ago. More recently, they have started making all article about particles and fields free open access. Using tools for tracking changes to LaTeX files, they also make it really easy to see what corrections the production staff has made to an article at the proof stage.
I don’t expect everyone to use LaTeX. I am a LaTeX wizard now (like the friend I mentioned above being a Mathematica wizard), and, having learned how to use LaTeX for its mathematical capabilities, it comes fairly naturally to me to use it for any written document that requires nontrivial layout elements. If I have to write something that isn’t just plain text, and there is no chance that I will need to share the source with anyone else, using LaTeX just makes sense.
For simple equations in PowerPoint presentations, I used to think that I could just use the built-in Microsoft Equation Writer (mentioned by David L), saving LaTeX for things that were more complicated. However, Microsoft’s proprietary equation format has a long history of compatibility failures. Any change in operating system, font package, or PowerPoint version is likely to kill your embedded equations. I quickly reverted to making equations in LaTeX and inserting them into PowerPoint as images.
@JE: yes, that’s why TeX automatically checks parentheses closure, which brings us back to the original post.
@PP: I gather online access to legal documents is also a big economic deal these days.
@LM: `it does keep me from using LaTeX for everyday writing’ : Understood. But if you have access to it, it’s quite easy to use for civilian purposes, \eg letters of rec.
Brett, you can create powerpoint in latex. You need a specialized package or two (never done it myself, but looked at someone else’s work). I am not sure how people working with complicated diagrams do it in latex (they do it, certainly). The language is fundamentally one-dimensional. And when I actively worked with it, inserting graphics was a bit of pain. In other words, there is a room for improvement.
For some HTML academic publications, each reference is followed by a link to a corresponding Google Scholar search, which is often Good Enough.
FWIW one social science data analyst writes “these days I almost never write anything directly in LaTeX. Instead, [my custom LaTeX] templates form part of a pipeline that stars with a Markdown or RMarkdown file and ends up as a HTML, Word, or PDF document.”
I got a job once by knowing what LaTeX was – I mean, probably not entirely, but I suspect I was the only person they interviewed that day (teaching admin job) who could tell them anything about it, because it made the physicists on the panel look quite pleased!
And all I did know was that it was something like HTML for maths, which a mathematician friend had told me.
Or with Acrobat itself.
Every PDF of a scientific paper contains links these days. Practically none in biology have ever touched (La)TeX.
@D.O.: There are specialized packages for producing virtually any kind of output in LaTeX now; that doesn’t mean that using LaTeX to mimic something like PowerPoint (which is for producing primarily graphical output) is a good idea. Nor do I have any idea what you mean about the language being “fundamentally one-dimensional”—unless you mean the same thing, that it is not a good native format for graphics. Note that in my comment above, I specified that LaTeX would be good for a “written document,” not absolutely any kind of output file.
Moreover, you are right about there having been problems with graphics in the past. As I said, LaTeX is not designed for producing graphics; although it can, in principle, be used for that purpose, essentially no scientist makes their figures in LaTeX. (Tables are another matter entirely, however; there, LaTeX is much easier to use than any Microsoft product once you have learned it.)
The early developers of TeX and LaTeX made some bad guesses and decisions about how to handle the incorporation of external graphics. LaTeX output was early on tied to PostScript standards, which made sense at the time. Unfortunately, the way the PS files were handled was not optimal. The formal requirements of the PostScript (and, even more, the Encapsulated PostScript) formats are quite stringent, but in practice it turned out that adhering to the strict standards were not really necessary, especially if you just wanted to view a PS file on a computer, rather than send the file to a printer. A lot of undefined behavior occurred, without it usually being a problem, and a lot of software packages produced sloppy PS and EPS files. However, the old LaTeX engine and graphics extension packages assumed that they would be dealing with graphics files that hewed perfectly to the official standards, and thus they could not handle many files that other software packages had no problems with. I once literally had to edit the text of an EPS file to remove some extraneous characters at the beginning that were impairing the ability of LaTeX to insert the file into a document.
LaTeX was unfortunately also slow to adapt to the use of Portable Document Format that Adobe introduced in 1993. It took until 2000 for a Vietnamese graduate student in Czechia to produce PDFTeX and a bit longer for the actually useful PDFLaTeX, which can handle and produce PDF format files natively. The high value of the Portable part of the PDF specification was significantly undervalued. Even as PDF files became the most important standard for sharing mixed text and graphics files on the internet, there was significant resistance to supporting it among many LaTeX developers, since it was a proprietary Adobe format (which, for better or worse, the PS format was not really seen as, in spite of its similar origins).
TeX actually did, from very early days, make an effort to create its own DeVice Independent standard, but it was a dead end; so far as I know, there is literally nothing one can do with a DVI output file except to postprocess it into a more useful format like PS. That was just one of the ways in which (La)TeX development turned out to go in ultimately non-useful directions. Donald Knuth also thought that font development and embedding—something that has virtually disappeared from consumer-level software over the last few decades—would be a major need for TeX users, leading to the development of Metafont, which as been essentially moribund since long before the last updates in 2014.
METAFONT was fun, though. I spent some time down a rabbit hole of trying to make a proper italic æ for Computer Modern, but I don’t remember why I didn’t like the pre-existing one. (Actually I do remember now, it looked too much like œ).
But yeah, Knuth should have spent more time on actual science. TeX and METAFONT are tours de force, but I’m pretty sure someone else would have made something just as useful. Even the literate programming paradigm seems to have died. (Knuth wrote the TeXbook, manual and theory of TeX, as a single [1007kB] source file that could be ‘tangled’ into Pascal source code for the TeX program [essentially reordering stuff so related code and declarations could be next to each other in the input], and ‘woven’ into TeX source). Maybe an IDE enforcing the use of java doc comment tags can be seen as a remote inheritance from that, though.
Insofar as we’re more into languages around here: I remember looking into LaTeX back when I was briefly attempting majoring in mathematics, about 10 years ago by now, and quickly discovered that its Unicode compatibility is horrible. By which I mean, any Unicode that you might actually type yourself using an advanced keyboard layout seemed to have no way of being directly accepted by the editors available to me at the time, forcing clunky workarounds like an ASCII coding of IPA; additionally in a scheme gratuitously distinct in various smaller details from X-SAMPA and any other standards that were around for this before Unicode support grew to tolerable levels. E.g. the general consensus was that ‹C R› stand for /ç ʁ/, but TIPA uses these for /ɕ ɾ/ instead. (Granted, all other systems had a few idiosyncracies like this too, e.g. most had ‹&› for /æ/ but standard SAMPA uses ‹{› for this and defines ‹&› as /ɶ/).
It does not come up in any of my work, but I have heard there is a quite handy syntax trees typesetting module though.
Latex is fully Unicode compatible now, but sometimes figuring out how to turn on a relatively simple setting can be a lot of work. I used to do most of my writing in Latex because I’d been told as an undergrad that it was so much better at producing longer/structured documents than Word. I’ve since found that Word does everything I’ve ever wanted Latex to do at a fraction of the time investment. My guess is that what my well-meaning teachers meant was that Latex makes it almost impossible to format a document ad hoc, so you can’t trap yourself in cases where you’ve written a 100-page thesis with section headings and parenthetical citations done by hand and now you have to manually update your bibliography and table of contents the night before you turn it in.
@D.O. That tale about the child practising the piano reminds me of the story of a pianist so sensitive to the tensions created by chords that, if his wife needed to get him to get out of bed, she’d go over to the piano and play a dominant seventh,… Unfortunately my Google-fu isn’t strong enough for me to locate it on the web.
Thank you for clarifying the Tunbridge Wells reference for Y. I feel that it’s all too often that someone writing on some web site makes a reference I don’t understand — and it’s probably American.
@Lars Mathiesen There are also freely-available Tex packages for other mathematical fonts. For example
\usepackage{txfontsb}
for Times for both text and maths;
\usepackage{mathpazo}
for Palatino for both;
\usepackage{fourier}
for fonts Utopia for text & Fourier for maths.
@jack morava You say “I suspect these tools will grow in importance with time”. For how long must people have used a tool for this growth to be complete? Tex was released in 1978, and Latex in 1984.
I think I need to install LaTeX on this machine and see what the world has come to. I did try one of those instant gratification programs long ago, but it was slow and clunky and flickered a lot and my boss wanted Word files.
@Brett, I assume D.O.’s one-dimensional is what I would call ignores vertical whitespace — there is no way to use separation of input lines or tabulators or similar to delimit items in the input, you have to use printing characters to indicate what is what; on the other hand, you can always break and indent lines the way it makes most sense to you.
That said, I vaguely recall that a LaTeX macro could actually get at the literal next character of the input stream, it was just bad style; so by now somebody has probably built a visual table input doohickey.
BTW, Lilypad for musical scores is the same sort of idea: flat Unicode file to score sheet with an editor that displays the result in almost real time. I do have that installed.
@rosie: ” For how long must people have used a tool for this growth to be complete?”
In 1623 or so Galileo said
La filosofia è scritta in questo grandissimo libro, che continuamente ci sta aperto innanzi agli occhi (io dico l’Universo), ma non si può intendere, se prima non il sapere a intender la lingua, e conoscer i caratteri ne quali è scritto. Egli è scritto in lingua matematica, e i caratteri son triangoli, cerchi ed altre figure geometriche, senza i quali mezzi è impossibile intenderne umanamente parola; senza questi è un aggirarsi vanamente per un oscuro labirinto.
Philosophy is written in this grand book, which stands continually open before our eyes (I say the ‘Universe’), but can not be understood without first learning to comprehend the language and know the characters as it is written. It is written in mathematical language, and its characters are triangles, circles and other geometric figures, without which it is impossible to humanly understand a word; without these one is wandering in a dark labyrinth.
We’re working on it…
PS: According to Stephen Hawking\begin{footnote}{teatime conversation with M Atiyah, I have no better reference}\end{footnote}, “Every time you quote an equation, you lose half your audience’. This seems sad to me.
Quotations
Stephen Hawking
“Someone told me that each equation I included in the book [A Brief History of Time] would halve its sales.”
Now, I’d like to shake my cane and clack my dentures at TeX and its ilk (as I have done all my life). This discussion started with Hat, a skilled copy editor (he has never published a typo in this blog that I saw, and I apologize for having thought so once, a little while back), noticing an unclosed parenthesis. And what do people do? Segue into a discussion of a typesetting program that can notice these things for you. Since I like a hyperbole, it’s like segueing from Monet to spectrophotometers.
It takes a programmer’s heart to feel that writing in markup codes is liberating. I have, for my sins, programmed in many languages and I daresay I’m good at it, so it’s not like I’m a luddite. However writing, or even typesetting a text in a human language with a programmer’s mind feels to me ugly and perverse. I’ll give it to Knuth that he has thought about the minutiae of text and typesetting long before most computer types have, but the finest LaTeX can offer feels like programming. I shudder at writing like that.
Eh. Topicality is overrated (or not rated at all, at the Hattery). Comments on yesterday’s post about Dostoevsky’s Writer’s Diary instantly turned it into a discussion of Mussorgsky, which I think may be more interesting anyway.
I’m not complaining about the change of subject. I’m griping about the enthusiasm for LaTeX.
I brought up LaTeX as an example, had no intention of dissing belles lett or consciousness of the breadth of human language. [My old drinking buddy Myles na gCopaleen supports me on this.] I just wanted to put in a word for a broader awareness of subtle thinking.
@Y: In that case, you appear to have a bizarrely negative view of spectrophotometry.
I think spectrophotometry is swell, but I am not in awe of it the way I am of Monet’s understanding of color.
Well, it’s your loss.
If anyone keeps account, by “one-dimensional” I meant viewing text as a single line, occasionally broken for the sake of convenience. Text as a frozen speech, if you wish. Strictly speaking, this is not true, there are footnotes and endnotes and tables. (And yes, I know, it is possible to create tricky layouts in tex/latex). But it is not visualising a text as a layout in 2D.
And I am with Brett here about spectrophotometers. I don’t see how they are in competition with Monet rather than in complement.
Conceivably the Book of Nature and the Book of the Human Soul are written in different dialects of the same language? Just sayin’.
It occurred to me that
https://en.wikipedia.org/wiki/Vincenzo_Galilei
was a musician \dots
The Lindsey / van der Meulen book Informal Introduction to Algol 68 (2e 1977) has a rectangular table of contents. Why, you ask. This being the Hattery, I am allowed to answer my own rhetorical questions.
The book begins with Chapter 0, “Very Informal Introduction to A68” (also published separately), which begins with this sentence: “Since ALGOL 68 is a highly recursively structured language, it is quite impossible to describe it until it has been described.” So Chapter 0 gives you a 60-page once-over-lightly, and then the rest of the 358-page book is divided into eight chapters on the major syntactic units of the language, each of which is divided into seven subchapters on the major semantic units.
Unlike the usual case, the numbered subchapters all mean the same thing, and thus one can use the rectangular title of contents row-wise or column-wise as desired. There are holes, however; subchapter 4.6 does not exist, because there is nothing to say about routines (procedures) in connection with union types (which first appeared in A68).
A bit like category theory: You need to understand all the concepts, because they are all special cases of each other. (Looking at you, Kan).
And unlike C and its ilk, A68 did unions right.
Noone mentioned that tex/latex use the most unusual (for the outsiders) brackets, dollar signes. And even double dollar signs.
Oh come on. Van Gogh understanding of color was far superior to Monet’s, so LaTeX is JUST FINE.
@Y, and LaTeX needing a programmer’s mindset — I agree to a large extent. But if you are writing a novel you shouldn’t be sprinkling \thinspace all over your text anyway, your publisher should have given you a stylesheet implementation so the most you need to do is stuff like \chapter{Sex scenes at last!}.
Of course then you don’t need all the braces and so on, the old line-based runoff style will suffice:
.CH Sex scenes at last!
If you are a tech writer or writing stuff with equations in it, it’s different. (Look at the NROFF source for Unix manual pages and weep). In that case a WYSIWYG editor is likely to get in your way more than it helps, but you will probably have the skills to wrangle LaTeX macros.
(The HTML tags we use on here are the worst of both worlds — very limited functionality, and totally unforgiving syntax).
@David Marjanović, re hyperlinks:
This is happening much more rapidly than I had realized: see
https://www.newyorker.com/culture/annals-of-gastronomy/just-buy-your-loved-ones-eleven-feet-of-salami-a-food-themed-gift-guide
(but this may be paywalled?)
I am unclear on what we are supposed to see here; just the fact that the article provides an abundance of inline links?
I am unclear on what we are supposed to see here
Amiable hyperlinked décadence. The New Yorker and its readers are a bit more self-indulgent than usual at Christmas time.
It’s not – but it’s not a PDF file…?
I was just surprised that the New Yorker was already so into hyperlinks, be they html (as you pointed out), or whatever. [I really hadn’t intended to advocate for LaTeX in particular.]
Ah, that’s pretty common.
Two of my colleagues (both more mathematically oriented than most biochemists, as I am) told me for years that I should move to LaTeX. For years I resisted, but around 2005 I took up LaTeX, and have never subsequently wanted to use anything else. It just does a far better job than anything else. You don’t have to use Computer Modern, which is horrible, as Times and Mathpazo have been available for years, and cost nothing. I use Mathpazo unless I’m writing something for a journal that demands Times.
I think the reason it took me so long was both of my colleagues were PC users whereas I was in love with the Macintosh from about 1987 onwards, and the Mac at the beginning had an understanding of typography far ahead of anything on PCs. I used a wonderful program called WriteNow, which offered everything useful (and did it better) than Word ever did with a footprint about a tenth of that of Word. (People who used MacWrite said it was just as good.) WriteNow partnered seamlessly with a mathematical application called Expressionist, which was actually a front-end for TeX, but you didn’t need to know that, and I suspect most of its users didn’t.
Unfortunately WriteNow was driven out by far inferior products like Word, and Expressionist by abortions like Word’s equation editor. Ultimately both were killed by successive changes to the Mac OS — increasingly difficult to use even with Classic after Mac OS 10 appeared, and impossible when Classic was abandoned.
All this meant that I had to move to LaTeX by about 2005. I’m happy to have done so, but I’d probably still be using WriteNow and Expressionist if I could.