Perfect! To keep things simple to begin with, I exported using Pure Typst, and it worked fine. Many thanks for sharing this and for keeping this thread up-to-date.
1 Like
Over on the Typst forum, a user has posted that he had just submitted his PhD thesis written in Typst. He gives a link to a blog post he wrote looking at the pros and cons of using Typst to produce a PhD thesis: PhD Thesis in Typst
Mark
1 Like
Late to the party here,
The guy writes:
My biggest complaint with LaTeX as a language is that nothing is consistent. Every package defines their own little utilities for even basic things like if-statements. It feels like you don’t learn LaTeX, you learn each package individually.
He simply does not know how to write a if statment in pure tex and now says that the language is not consistent… Also, the confusion between LaTeX and TeX says a lot.
The problem with this project is that it is trying to reinvent the wheel and bring no true advantage at all. Their compiler can handle trivial cases, but their core design decisions are unbelievable terrible for some type system in the 2020’s. The biggest selling point for their start-up is the compilation time, due to incremental compiller. But an incremental compiler is a computational concept. Someone also can write an incremental compiler for TeX too. There was a project in this direction, called tecnotic, but seems to be alsmot dead now, here.
The rest of the article is just bullshit that the script language can do a lot of computational tasks, something totally trivial done by other means.
My view is that the guys behind typst put to much confidence on their performance and did not studied well how the process of producing documents works. The choose for markdown as a base is a joke, also, the sole fact that they did not work natively with an intermatiade serialized format is also a joke. The fact that the guys eliminated the concept of glue from Knuth algorithm is the worse joke here.
They missed a huge opportunity here, by not learning from the actual tecnologies and trying to reinvent the wheel and not integrated battle-tested concepts and tools. For sure they will find a niche for their start-up, maybe generating invoices and manuals for less demanding documents due to the compiling time. On the other hand, for serious scholars works and typography, Typst, by design, will be far far behind from being useful.
Talking of confusion, since Typst does not require Markdown…
1 Like
Meanwhile, Vim is the only true text editor.
(Any other tech wars we can get out of the way, tabs or spaces?)
5 Likes
markdown as a base model for the syntax, I should have completed the sentence.
A markdown flavour plus weird commands # to call functions and etc. TeX was created in 70’s and has a better parser approach for the markup than a software in 2020’s. This is what we would call terrible design decision. I am curious to see the workarounds for demanding cases.
If you want a concrete example, rst is years-light superior than the markdown mess, but they prefer the markdown flavour. Maybe becuase it is fun and I enjoyable.
Meanwhile, Vim in Emacs is the only true text editor.
An Evil amendment, I know…[1]
Yes, yes, this joke is fairly niche ↩︎
1 Like
The release candidate 0.14 is pretty awesome. Importantly, they have really pushed their PDF accessibility, more details are here:
They support every PDF version and accessibility standard apart from one, and all in ~13MB download, amazing! Lots of other little tweaks, I think they are well on their way to a super V1.0 release: Roadmap – Typst Documentation
The problem with writing raw Typst remains: what if you need to reformat for Word (many editors or journals demand that) or a format other than PDF (they are aiming for HTML as an intermediate, but that pipeline is rough and ready)? This is why always writing in markdown, then converting to Typst/ODT/TeX/DOCX/Text is the best way to use Typst at present, a workflow that remains ideally suited to Scrivener users.
3 Likes
A nice article on Typst from a mathematician (who also happens to be an organiser for the totally wonderful Manim project: Made with Manim)
From the file.
-
PDF/UA-2: There is also the more recent part PDF/UA-2 that targets PDF 2.0 files. It improves accessibility for mathematics and some semantic elements. Support for PDF/UA-2 not yet available in Typst, but planned.
-
Well Tagged PDF (WTPDF): This is an industry standard that is very similar to PDF/UA-2. Like PDF/UA-2, it is not currently supported by Typst. Originally, it was drafted because both parts of the PDF/UA specification were only available at a high cost from the International Standards Organization. Hence, WTPDF was designed so that all conforming files can also declare conformance with PDF/UA-2. By now, both parts of the PDF/UA specification are available free of charge, decreasing the relevance of WTPDF.
Four years from a software in 2020’s that still struggle a lot in html export. Also, trying to use html as medium…
Well… Do I need to say something more? A lot of hype for such poor results. I had some hope when sile appeared (a kind of TeX alternative in pure lua script). But the project did not get traction and was focused too much in humanities. Now it is more in manteinance mode.
All the comparasion that I saw (Typst vs TeX) so far are only acurated when talking about ‘compilation speed’ (for some reasons there are people that really need live preview or they will die in dust). All the other arguments are simply bullshit of feeling, fun and nice from low-level users, like the guy said here in the linked article.
Now, I don’t really mean that I know and understand the inner workings of TeX down to its inner core, I haven’t even written any CTAN-distributed LaTeX packages (nor do I maintain any); just a bunch of local-ish templates for theses and some others for slides, letters, CVs, name tags, lecture notes, and articles; pretty standard stuff.
Do not get me wrong, I would love to see a good typsting system well integrated witth modern technologies and taking advantages of them. TeX was written in 70’s and for sure, Knuth, being the legend as he is, would make a totally different implemantation in face of modern technologies. The guy basically created the field of digital typestting at that time.
But now the dilema, I do not know if it is all merits of prof. Knuth to write a so good piece of software, or if it is the incompetence of people trying to be contenders, that results in the fact that extending a 70’s software still better than any today’s alternatives by design, not even by features.
Aaaaand final release is live:
I hadn’t really played with the new Character-level justification feature, and trying some sample text quickly it looks really good overall!
2 Likes
Yes… installed using Homebrew, though I haven’t tried it locally yet.
My projects on the web app have all updated without problem. I created a new one (basically just text and headings, with a few footnotes and section titles in headers using the “hydra” package).
For that, I tried using character level justification to tighten the text (Libertinus Serif) slightly and really liked the outcome. However, it required some thought, as the blog refers to “Character-level justification”, though the documentation gives “Tracking”.
Mark
1 Like