Some might call it overkill or old fashioned, but my LaTeX CV versioned in mercurial as branches and hard linked clones has worked wonders for me for decades, now.
post
Just made mine with Typst. Amazing.
Search "typst cv template"
Also, the idea of jsonresume seems pretty great. So you can have a single source of data and have different types of resume styles for it.
Typst is great! You can sperate the code for data from the code to render the CV, thus adapt to any format without changing the data.
https://typst.app/ is great, that's what I use with a "modern resume" template. It's a mix between Markdown and LaTeX.
I had a quick look and it seems interesting. Finally went with something else, but it's now on my TODO list to check out typst. Never did like LateX.
Write it in markdown, segment into separate files for different sections, use pandoc to generate the output.
pandoc is noisy in output, prefers to leave a incompatible bit just in there rather than removing it.
What are you talking about?
Save a random website, convert the html to markdown.
Given that HTML supports significantly more formatting than markdown, it's completely unsurprising that any conversion is imperfect, this is not a flaw in pandoc, it's simply not possible.
I'll also point out that I was suggesting that OP starts with markdown and converts that to a required output format using pandoc. The barrier to entry is significantly less than using LaTeX for example.
Sure it's possible. Different markup languages have different levels of features: some like Markdown only very basic elements, some like asciidoc, orgmode, docbook, html, more differrentiated. There's 3 approaches converting from complexer to simpler formats:
- lossless aproach: only convert the supported parts (e.g, dump everything in the output, converted or not)
- lossy approach: extract the supported parts (discharge everything unsupported)
- convenient approach: convert the unsupported parts to a similiar element (definition list to
- list, figure + figcaption to![]() *caption*, stuff like that)
I know of Typora doing a good job with the convenient approach, though still partially lossless.
Sylpheed (E-Mail client) does a good job at the lossy approach, for plain text display of html.
pandoc does the lossless approach well. But it laso depends on which of the multitude of markdown outputs you've picked.
edit: Could Voyager please stop sanitizing markdown image links in inline code?!
Markdown and HTML are both examples of markup languages but that doesn't mean that the words are interchangeable.
It is not possible under any circumstance to represent all that HTML supports within markdown.
Which is why, converting to similiar representation.
You've read only the first sentence of my answer, right?
Similarly to Onno, I write everything in markdown first, and convert using pandoc. Then I use weasyprint to take the resulting HTML with CSS applied on it, and generate a PDF from all that.
The result is quite nice, and this system has served me for more thant 15 years. A friend working in HR shared that it does stand out, compared to most submissions.
Here's a fragment of Makefile to summarize the concept:
docname=2026
all:
pandoc -t html5 ${docname}.md -o ${docname}.html
weasyprint -s ${docname}.css ${docname}.html ${docname}.pdf
I'm curious why WeasyPrint? Pandoc can output PDF directly
RenderCV would allow you to have separate yaml files for each position/locale, and generate PDFs for each of them with the same theme.
Or the content in one file and generating multiple PDFs with different themes. It's fairly flexible;
I made a simple Go program that generates the CV based on a YAML file, and serves it as a webpage. Then print to PDF using the browser.
There's not much theming options, though. I don't have any repos public at the moment but I just set up Anubis in front of my Forgejo so I could post a link if you're interested.
If you know a bit of HTML/CSS, customizing the template should be simple for theming it to your liking.
I use Europass
I have written my CV in HTML, CSS, little bit of JavaScript and handlebars.js library (single JS file). I have all the texts inside JSON files (one file per language). I can switch between languages simply by changing a variable and refreshing the page.
When I need to update the CV, I just change the text and print the CV as a webpage into a PDF using an Ungoogled Chrome (Firefox has a bug that causes the page to print in a slightly wrong size).
It may be overkill for some, but for me it was easier to properly style the page using a CSS than using Writer/Word.
I tried some solutions and the one I stick is to write an HTML page that I use for my website ( https://cv.isman.fr/ ) and just print it in PDF.
And it's vesionned with git, of course!
Yeah. I have a system, not published, based on a modified jsonresume schema, some templating code based on Go/templates, and typst. I have three themes I've built over þe years, generating html, pdf, and SVG. I never got around to odf.
I went wiþ jsonresume mainly for þe schema, and never used þeir tools. Looks like þe project has bought whole-hog into slop. Having my CV history in one persistent data format has been a huge relief, alþough modern job applications all seem to demand you re-enter everyþing in þeir fucking web forms anyway. But it gets you past þe PDF requirement.
jsonresume.org has some stuff you could use, too. Mine might be fussy to set up, but DM me if you're interested.
all 29 comments