I was always upset that my PhD advisor would read printouts of my paper drafts and correct my LaTeX by hand in the margins. I mean, not even seeing the code...?

Now I'm reading paper drafts by my PhD students and collaborators as PDFs and also correcting their TeX code because I cannot stand wrong font for maths or wrong spaces between numbers and units. I wish I could stop, but I just can't...

It's all my PhD advisor's fault! *shakes fist in the direction of her old institute*

@vicgrinberg

" - ": πŸ˜΅β€πŸ’«
$km/s$: 😩
"e.g. something": πŸ˜•
"(\cite{REF})": 😱

@vicgrinberg @knud and just keep changing the text until all underfull hboxes are gone!
@baszoetekouw @knud pssst, nobody admits to this loudly...

@vicgrinberg @baszoetekouw

Oh, I've never done that. But I use \-

@vicgrinberg @baszoetekouw @knud although in theory LaTeX should take care of the layout, it can only do so if the input is correctly described. When looking at its rendered form and fixing layout issues I'd only fix one instance of a "bug", but that doesn't make the input correct under all possible scenarios. It is like fixing a bug in a program only for x=42, and having everything break when x=433.

I recently looked at how a paper I wrote looks in arXiv's experimental HTML rendering. The footnotes are now on the right margin (great, it is nearer the text), but they have big whitespace gaps in the middle of a sentence and it looks horrible. Not LaTeX's fault, I explicitly told it to keep the name of a company together using ~, which was needed to fix some weird hyphenation when rendered to PDF. But when displayed on the much narrower right margin that is now wrong. It probably would've been better if I told LaTeX to avoid hyphenating the word, instead of telling it to avoid breaking at a space.

But that is all still too low-level (layout fixes leaking into the text). Would be great if LaTeX had a way to semantically say "this is a company name", and then do whatever is appropriate for that based on the current render target.

@edwintorok @vicgrinberg @baszoetekouw

There's an interesting discussion here

https://tex.stackexchange.com/questions/76132/how-to-typeset-a-small-non-breaking-space

describing how to set a breakable whitespace of minimal and maximal length. That might do the trick.

How to typeset a "small" non-breaking space

I can use ~ to produce a non-breaking space. But it is too large for me, and I would like a non-breaking space of the size of \, for example. Is it possible ?

TeX - LaTeX Stack Exchange
@vicgrinberg @baszoetekouw @knud oh Knuth himself clearly mentions this as a viable strategy, don't worry, it's nothing to be ashamed of. But only do that when the text is FINAL
@oblomov @vicgrinberg @baszoetekouw @knud
Do it when the text is "final_reallyfinal_v2.2_done_corr3_edits.tex"

@jannem @vicgrinberg @baszoetekouw @knud

I guess using git for source tracking and collaboration is a bit too much still 8-D

@oblomov @vicgrinberg @baszoetekouw @knud
I did use git for my papers. It's a godsend! Check in every new revision as you write. Edited versions from your coauthors go into a branch with their name before being merged into main.

Highly recommend.

@oblomov @vicgrinberg @baszoetekouw

Won't help me if my text again gets touched by a publisher. But an interesting approach maybe for preprints.

@vicgrinberg @baszoetekouw @knud Hey, the chasing underfur \hbox warnings is a wonderful excuse for procrastination on writing more actual text.
@vicgrinberg I feel the same way when I spot double spaces or even a single character off in the indentation. It’s frustrating, and I have to step in and correct it, along with explaining that what they wrote isn’t quite right. Plus, even if the fonts are all the same, I notice a difference on the first line, which can be a bit upsetting.
@vicgrinberg teach a person typography and they will suffer from seeing bad typography for life

@mxk @vicgrinberg knowing kerning can be hard on the senses

#typography

@mxk @vicgrinberg I am still recovering from the infamous Higgs presentation.

@ftranschel @mxk @vicgrinberg

Oh, I don't know that one. Please tell me more.

@ftranschel @vicgrinberg @LiseAndreasen CERN has a habit of using questionable design choices including Comic Sans in their presentations and Higgs probably was the time a lot not people first saw this
https://fontsinuse.com/uses/11764/higgs-boson-discovery-announcement
Higgs boson discovery announcement

Fonts In Use
@mxk @vicgrinberg and the font of all knowledge
@mxk @vicgrinberg Me carefully sizing my window to take a screencap of this joke with good linebreaks. 😰
@mxk @vicgrinberg Teaching singers to sightread can have similar consequences.
@mxk …and those around them who have not yet learned typography shall be made to suffer also, via commentary from the former @vicgrinberg
@ellane @mxk well played πŸ€£πŸ‘
@mxk @vicgrinberg Warning: this is also true of welding
@mxk It is true and my life is ruined 😭 @vicgrinberg
@mxk @vicgrinberg Teach a person languages and they will suffer from seeing bad machine translations for life
Urban Dictionary: keming

(n.) The result of improper kerning. (to kern is to adjust space between characters in typeset text).

Urban Dictionary
@mxk @vicgrinberg and they will get a subscription to the New Yorker
@vicgrinberg
Back in the the days when I was proofreading diploma thesis I used two different pen colours. One for the really important stuff and one for typographical corrections. That way they could easily ignore the things that I could not. But that is of course different when ones name is on the paper. Then stuff like that is critical (says the one who made a font to use in his PhD thesis ...).

@vicgrinberg reminds me of the time in Technical Drawing class during my BSc the professor was correcting someoneβ€˜s work in front of the whole class, marking the transparent drawing paper with a red pen. The student, startled, not believing his eyes interrupted him: β€žI think you’re looking at it from the backside πŸ˜§β€œ

without even looking up or flipping the paper, he said in a monotone, calm, emotionless voice: β€žI know.β€œ

We weren’t able to tell if he was lazy or just showing off

@vicgrinberg I seem to have a limit for the frequency of this kind of error above which I cannot take in the content anymore. I only see the errors after that.
@hmwilker ahah, yes 😭 Or at least I get really pissy in my comments - and I have a few collaborators who are really terrible and don't even try 😭
@vicgrinberg we have linting / style guide enforcement tools for programming languages, do we have similar tools for TeX?
@ShadSterling I don't think this would work with a technical solution since this is context dependent on what the text actually days. And there already are rules...
@vicgrinberg @ShadSterling Finally something LLMs might be good at!

@vicgrinberg my t shirt is raising questions already answered by my t shirt

https://discuss.systems/@ricci/110905390639969469

Rob Ricci (@[email protected])

Attached: 2 images Wearing this shirt to the faculty retreat today to trigger my colleagues

discuss.systems
@ricci ahaha, absolutely love this!
@vicgrinberg I hate the empty faces around me whenever I say LaTeX. While they play around with Canva and Word, I use InkScape and LaTeX. And don't get me started on social media: I say Mastodon and get looked at as if I had an illness. To cut it short: Be happy about LaTeX errors, at least you have some healthy nerdy people around you.
@janhelms well, it's not about being (healthy) nerdy - in my research field it's a requirement, all big astrophysics journals require the paper submissions to be in LaTeX, using their respective style files. So very different context.
@vicgrinberg πŸ˜‚ relatable but never a problem I had because before I was a grad student my wife and I met while working for a book typesetting company the used TeX/LaTeX.

This post and comments thread are just πŸ‘Œ

Thanks @vicgrinberg and all who contributed. You made my day. 

@dacmot haha, glad you enjoyed! I'm definitely having fun (mostly) πŸ˜‚

@vicgrinberg

would it be better if your students used Word for their papers?

@echopapa they could not - in my field, astrophysics, all the journals require paper submissions to be in LaTeX, using their respective style files.

@vicgrinberg

oh, torturing poor MS-fans (or was this SM?).

very nice.

@echopapa it's not used in my field at all; too many formulae, even at first year student level. It's more the other way around - at some point you reach a point in your career when you have to deal with administration and then they torture you with Word documents.
@echopapa @vicgrinberg The only way Microsoft Word could make anything better is by imagining your work done in Microsoft Word. Suddenly your biggest problems with your current programs aren’t quite so big.

@vicgrinberg You need to stop. If you do it for them, they won't learn.

Ditto marking up word documents for papers - students just hit that "accept changes" button and learn nothing.