Pandoc writing app
Anyone here invested in a Pandoc-centric workflow?
Tried PanWriter?
Would love to hear opinions. I've never used Pandoc for large writing projects, just small conversions, so I can't judge the capabilities of the editor.
Author at Zettelkasten.de • https://christiantietze.de/
Howdy, Stranger!
Comments
I have a Pandoc-centric workflow and have known about PanWriter for a long time but have never used it. Its developer Mauro Bieg ("About me") also has a cool blog and has made many code contributions to Pandoc.
new member.
interested in Pandoc / Markdown compatible system of notetaking and retrieval.
long-time user of iA writer; tried Panwriter briefly - looks good.
even more stripped back than iA Writer - which i like - but with no collapsible folder view it presses my need for improved organisation in my notes and filing - which is why i'm here....
my particular interest is moving all research studies, notes, thoughts, and writings in markdown to R Studio.
it's a logical progression - i use R for statistical analysis and graphical output , and markdown for drafting and writing.
dynamic documents makes sense, in this case.
eventually the thesis will be output from R Studio using R Markdown.
i am interested in recommended essential readings around the subject of zettelkasten / best practice / pitfalls / preferred filing notation.
i understand it may take some work for initial setup but may bring large benefits in the long run.
or should i start and adjust / evolve as i go along?
i will look at other forum postings, but thought it may be useful to add R Studio and R Markdown into this conversation.
i have not offered a solution to the original question, but share a common interest?
what does a 'Pandoc-centric workflow' look like?
thank you in advance!
a good day to all.
jonathan
@digbyphonic said:
Are you familiar with Quarto? I don't use it (yet), but you may want to look into it. "Quarto is a multi-language, next generation version of R Markdown from RStudio, with many new features and capabilities" ("Quarto – Using R"). Further reading:
"Quarto is based on Pandoc and uses its variation of markdown as its underlying document syntax" ("Quarto – Markdown Basics").
I would say: Just start and evolve as you go. It sounds like you have some programming skills; keep in mind when you're building your system that it's smart to use markup that you can easily convert to something else in the future. Markdown, Pandoc, and standard Unix scripting tools help with this.
Pandoc converts Markdown documents into almost any other format, so you can write in Markdown and then convert it into some other format for publishing. Pandoc Markdown can include LaTeX equations, which makes it a nice solution for scientific writing without heavy LaTeX markup. (A different software package called MultiMarkdown has a similar function but fewer options, but some people prefer the MultiMarkdown syntax over Pandoc Markdown syntax, but Pandoc can also convert MultiMarkdown syntax.) Pandoc also has a great filter system that you can use to customize it easily; the pandoc-crossref filter, for example, permits dynamic numbering of equations, tables, and figures. (Something like this latter functionality may be built in to Quarto—I don't know.)
Since you need to do R computations, Quarto (mentioned above) can do that for you while using Pandoc. Pandoc by itself won't do it.
Further reading: https://forum.zettelkasten.de/discussion/2572
A Pandoc-centric workflow has no opinion about how a document should be created. It couldn't care less about PDF, Markdown, HTML, Docx or any other format.
Pandoc is probably what you end up with after porting your documents from one system to another. It advocates strict separation of structure and style.
However, Pandoc cannot do much more than that, so your projects are going to reach a limit on complexity you can handle, eventually.
my first Zettel uid: 202008120915
If there were support for WikiLinks and the option to view the H1 header or a YAML title variable as the filename or in addition to the filename when the file is selected, I would consider using these alternatives to Zettlr, for which I have a workflow that includes Zotero, MikTeX, and WinEDT. Since they don't seem to accommodate a Zettelkasten workflow out of the box, or with straightforward modification, they are more useful to me for notes in isolation. I wanted to work with VS Code and started writing extensions, but this was a diversion.
GitHub. Erdős #2. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Alter ego: Erel Dogg (not the first). CC BY-SA 4.0.
@zk_1000 said:
This is true: Pandoc can read and write many different formats, with Pandoc's AST as the internal intermediate data structure (so conversion is "lossy" if you use any file format's features that don't have a direct translation into Pandoc's AST). When I said "Pandoc converts Markdown documents" I was thinking of Markdown because that's what @digbyphonic wanted to use, but this statement by @zk_1000 should be considered an important correction to what I said.
@zk_1000, could you give an example of what you are thinking here? I don't doubt that it's true but I have difficulty thinking of practical implications. What would be an example of something that @digbyphonic might want to do that would exceed the limit?
@ZettelDistraction said:
Pandoc supports WikiLink syntax when you enable the relevant extension.
But I wonder if @digbyphonic was thinking of using only RStudio as a Zettelkasten app. I don't know if RStudio has the functionality to easily create and browse WikiLinks. @digbyphonic may need to find a Zettelkasten app to use alongside RStudio—I don't know.
@digbyphonic asked:
Here's a simple graph of my Pandoc-centric workflow (it says "Pandoc or Quarto" but I don't use Quarto):
In Graphviz syntax:
digraph workflow { Zettelkasten -> Draft [ label="Select & Organize Zettels" ]; Draft -> Draft [ label="Edit Draft" ]; Draft -> Output [ label="Compile via Pandoc or Quarto"]; }That's helpful, but the "title after pipe" syntax isn't compatible with Zettlr--a tragedy of unimaginable proportions. Behold, without the pipe, a mouse hover over a WikiLink causes Zettlr to pipe up with a helpful file dialog box:
But add a pipe and some text, Zettlr chokes on the pipe:
I nevertheless seek the holy grail of software compatibility.
GitHub. Erdős #2. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein. Alter ego: Erel Dogg (not the first). CC BY-SA 4.0.
thank you, everybody!
much to get into here, which tells me i'm on the right track: this site is a wonderful collection of knowledge and insight.
-
yes, i have seen Quarto. thank you.
it may have been missing one citation feature i need (i'll try to look back into what that might be, and post if relevant. maybe linked to Zotero?).
Quarto can fully open an R Markdown file, but Quarto functions do not exist in R Markdown.
i may proceed with the established software and move over once Quarto is more established?
i agree with a software-agnostic approach.
Markdown and R fit this need.
-
i need to absorb the concepts mentioned, and those within the suggested books and articles, and then envision how to use the software tools available.
i now realise that i use a clumsy long-hand version of note-taking with similarities to atomised 'Zettels' already. it's not ideal - glad i'm here.
others have 'been there, done that' over many lifetimes, and figured out a better way for themselves.
the personalised and evolving nature of a Zettelkasten is very interesting!
once i can visualise how this might work 'in action' on a PC screen i'll be ok - not there yet!
thank you.
jonathan
@digbyphonic
Mine is a bit more 'technical', but all my writing is in markdown (including Zettelkasten). This allows me to simply insert a yaml frontmatter block of pandoc settings at the top of it and run
pandoc source.md -o dest.pdf. For documents where I need rock solid citations, figure refs, etc, I write directly in a .tex document (LaTeX).The Skim pdf reader for mac also allows automatic reloading, so I just setup an autocommand in neovim that executes the pandoc / pdflatex command on save, which Skim then sees and reloads the document. I can then stay in neovim but have the same immediate visual feedback of the final document.
Eg, this yaml frontmatter makes for a nice 'modern' style of pdf with the title, subtitle, author and date centered.