cocalc VS pandoc

I have some unit tests for billing and subscription code for my company that started breaking in CI today due to the leap day: https://github.com/sagemathinc/cocalc/commit/8575029c2b76787...

1. I'm integrating ChatGPT extensively into https://CoCalc.com. This integration makes a lot of sense, because cocalc is a platform in which relatively inexperienced students use Jupyter notebooks, linux terminals and Latex. So far, the most popular feature by far is a "Help me fix this" button that appears above stacktraces in Jupyter notebooks.

2. One software engineering challenges is that ChatGPT often outputs code in markdown blocks. I've had to emphasize in prompts that it should explicitly mark the language. I then got inspired to make it possible to evaluate in place the code that appears in these blocks using a Jupyter kernel, and spent a week making that work (so, e.g., if you type a question into the chatgpt box on the landing page at https://cocalc.com, and code appears in the output, often you can just evaluate it right there). There seem to be endless surprises and challenges though. For example, a few minutes ago I realized that sometimes the giant tracebacks one gets when using Python in Jupyter notebooks are so big (even doing simple things with matplotlib) that they end up resulting in too much truncation: https://github.com/sagemathinc/cocalc/issues/6634

3. I'm mostly using GPT-3.5-turbo rather than GPT4, even though I have a GPT4 api key. Aside from costs, GPT4 takes about 4x as long, which often just feels too long for my use case. The average time for a complete response from GPT-3.5 for my application is about 8 seconds, versus over 30s for GPT4.

Github's implementation is really lazy. There are many much better approaches to precisely this problem. E.g., Jupyter notebooks implement one that has matured in the wild over a decade. There's this very flexible markdown-it plugin that implements anther https://github.com/goessner/markdown-it-texmath, and my version of it here https://github.com/sagemathinc/cocalc/blob/master/src/packag... which I rewrote in typescript with a focus on the same semantics as Jupyter has, but for CoCalc, and I've been working on using unifiedjs to provide more general latex for Markdown (not just formulas) here https://github.com/sagemathinc/cocalc/pull/5982 Parsing math is much easier if you use a plugin to an existing markdown parser, rather than trying to do some hack outside of that (which is what Github probably does, and also what Jupyter does).

My main authoring tool is then Emacs Markdown Mode (https://jblevins.org/projects/markdown-mode/). For data entry, it comes with some bells and whistles similar to org-mode, like C-c C-l for inserting links etc.

I seldom export my notes for external usage, but if it is the case, I use lowdown (https://kristaps.bsd.lv/lowdown/) which also comes with some nice output targets (among the more unusual are Groff and Terminal). Of cource pandoc (https://pandoc.org/) does a very good job here, too.

This is one of those things that the ever-amazing pandoc (https://pandoc.org/) does very well, on top of supporting virtually every other document format.

Folks feel the same way about Markdown versus LaTeX: why use something significantly more complicated where a looser, human-readable grammar works better?

For any other situations, I use https://pandoc.org/, or, generate a Word doc scriptomatically.

pandoc toolchain pour builder une version confortable/imprimable en phase de travail (ePub, pdf, docx, html)

Congrats on the launch, I guess, but there are so many free options that I can't think of a situation where paying $0.25 per document would be justified...? Just to name a few:

Back in the days, I used to use XSL-FO [0] and it was okay. It was not very precise but it rarely if ever broke, and was perfectly integrated with an XML/XSLT solution. Yeah, this was a long time ago.

Last month I used html-to-pdfmake [1] and it's also not very precise and more fragile, but very efficient and fast.

Yet another approach would be to pro grammatically generate .rtf files (for example) and use Pandoc [2] to produce PDFs (I have not tried this in production but don't see why it wouldn't work).

[0] https://en.wikipedia.org/wiki/XSL_Formatting_Objects

[1] https://www.npmjs.com/package/html-to-pdfmake

[2] https://pandoc.org/

Others have mentioned static site generators. I like Hakyll [1] because it can tightly integrate with Pandoc [2] and allows you to develop custom solutions if your needs ever grow.

[1]: https://jaspervdj.be/hakyll/

[2]: https://pandoc.org/

Have you compared it with a conversion by pandoc (https://pandoc.org/)?

I have used it to kickstart a blogging project that I wish to come back to soon. The Lua inter-op for custom readers, writers and filters is great but I wish there was more editor integration and even perhaps an official IDE/editor with built-in debugging features (probably something already do-able with Emacs but I haven't checked). The only blocker for my project is no support for "ChunkedDoc" for Lua filters [1] which forces me to write more code and a complicated Makefile.

[1]: https://github.com/jgm/pandoc/issues/9061