Startup Ideas Bank
Typst: A Clunky Step in the Wrong Direction for Document Processing.
AI roast score: 55/100 (D)
The idea
Typst 0.15.0
Changelog
0.15.0
Typst 0.15.0 (June 15, 2026)
This section documents all changes to the Typst language and compiler between Typst 0.14.2 and 0.15.0. If you are migrating an existing document to Typst 0.15, make sure to check out the Migration guide . It walks you through changes you may need to make to your existing documents to ensure compatibility with Typst 0.15.
Highlights
Typst now supports variable fonts
HTML export now supports equations out of the box via MathML
With the new, experimental bundle export target, a single Typst project can output multiple files (e.g. a multi-page website)
A single document can now contain multiple bibliographies
Typst can now target multiple PDF standards at once
The new within selector simplifies many introspection use cases
The new divider element represents a thematic break that templates can style
Spot colors enable use of custom pigments in offset printing
With the new file path type, project-relative paths can be passed to packages
The new, more general typst eval CLI subcommand supersedes typst query
Layout convergence issues now result in detailed diagnostics
Two long-standing list layout issues with marker alignment and centering were fixed
Paragraph handling in HTML export is improved, preventing unexpected paragraphs from appearing
This documentation now has a print version
Language
Syntax
File paths (e.g. in imports or image function calls) may not contain backslashes anymore; instead forward slashes must be used (Breaking change)
Added hints for invalid characters in code mode
Added hint when trying to use a unary operator directly in an embedded expression using a hash (e.g. #-30deg )
Fixed potential stack overflow crashes by enforcing a maximum parsing depth
Fixed incremental parsing of unclosed strings
Styling
Text show rules now have tracebacks that include the matched text
Fixed a crash with text show rules that match on multi-character symbols
Scripting
Extended hint when built-in definitions are shadowed to set and show rules
Added hint when trying to spread one or multiple dictionaries into an array
Improved diagnostics for invalid method calls
Improved hint for unknown variables in math that are available in std
Fixed a misleading error message when trying to assign to a temporary return value
Library
Foundations
Added file path type that is now accepted in all places where paths were previously only represented as strings
A path constructed in one file can be used in another file, but will be resolved relative to its original file
Likewise, paths can be passed across package boundaries
The initial path type is very minimal, but additional features like file existence checks or directory walking are planned
Collec
The roast
The updates to Typst might make you feel like you're getting more value, but in reality, they’re just band-aids on a fundamentally flawed product. The focus on features like 'variable fonts' and 'HTML export' suggests a lack of a cohesive vision; users want functional, reliable tools, not experimental novelties. And let's not even start on the 'breaking changes' that force existing users to scramble. It feels more like a half-baked attempt to catch up with competitors than a groundbreaking innovation.
Your profile shows you're technically a solo founder with no funding and a 'speed' moat, but the real question is: will anyone actually pay for this? The revenue model is far too dependent on uncertain consumer interest ('q15=will_pay'). With the abundance of solid alternatives already available, Typst is going to struggle to carve out a niche when it comes to adoption. Without a clear path to market traction, your solo journey is likely to be a lonely one.
Red flags
- No funding to scale development means limited reach and capability.
- Heavy reliance on a single product feature set exposes vulnerability to competition.
- Solo team lacks the bandwidth for rapid iteration and customer support.
Verdict
Reassess your value proposition and focus on core functionalities that genuinely solve user pain points, rather than adding features for the sake of it.
Roast your own startup idea →