I partially agree. I recently started to heavily use typst for any typesetting where I have enough time to make a template. But as far as I know, typst does not support all the little things LaTeX can do.
For instance hyperref can highlight links so they look different on paper and different in the viewer. Last time I checked, typst backend does not support using those layers.
I also have mixed opinions on math syntax. It is great, but current parser implementation really wants you to put in some spaces that make the typing little awkward.
Great plus is proper Unicode support. I remember having trouble citing chinese document names in bibliography, because half of the names wouldn’t use the correct font.
Overall, it’s great, but I don’t consider it to be direct replacement.
I personally never had a case where Typst couldn’t do what I wanted and usually found something better in Typst instead. The only exception is to do curly snake-like rectangle borders, but it’s probably doable in Cetz.
As for hyperref, you are propably right, I would personally use a boolean flag for that and do two versions. Though the whole point of a pdf is to have the same look between it and the print, other features might fail depending on the viewer.
I think the extra spaces in math make it more readable, and it’s kinda the price to pay to not have to add backslashes everywhere. One instance when I found it weird was that frak("abc") had a different font than frak(ab c). So i just make a #let abc = $frak( a b c)$ rule.
I personally never had a case where Typst couldn’t do what I wanted and usually found something better in Typst instead.
Same. I love the structure and syntax. I did not mean to sound a bit like hater. I just wanted to mention that there are niche usecases where you might still need LaTeX.
There’s no going back from typst
I partially agree. I recently started to heavily use typst for any typesetting where I have enough time to make a template. But as far as I know, typst does not support all the little things LaTeX can do.
For instance
hyperrefcan highlight links so they look different on paper and different in the viewer. Last time I checked, typst backend does not support using those layers.I also have mixed opinions on math syntax. It is great, but current parser implementation really wants you to put in some spaces that make the typing little awkward.
Great plus is proper Unicode support. I remember having trouble citing chinese document names in bibliography, because half of the names wouldn’t use the correct font.
Overall, it’s great, but I don’t consider it to be direct replacement.
I personally never had a case where Typst couldn’t do what I wanted and usually found something better in Typst instead. The only exception is to do curly snake-like rectangle borders, but it’s probably doable in Cetz.
As for hyperref, you are propably right, I would personally use a boolean flag for that and do two versions. Though the whole point of a pdf is to have the same look between it and the print, other features might fail depending on the viewer.
I think the extra spaces in math make it more readable, and it’s kinda the price to pay to not have to add backslashes everywhere. One instance when I found it weird was that
frak("abc")had a different font thanfrak(a b c). So i just make a#let abc = $frak( a b c)$rule.Same. I love the structure and syntax. I did not mean to sound a bit like hater. I just wanted to mention that there are niche usecases where you might still need LaTeX.
No worries, you are right. And even if typst was perfect you have to work with other authors or journals that won’t want to use it