**.txt, .csv, .ods, .pdf, .mp4, .flac, .tif** Not all file types are created equal. In terms of resistance to going out-of-support, or file type corruptions, these are the kings: - Documents - .txt - *king*, use [[UTF-8]] character encoding - Any offshoot of text that is just text under the hood (like [[Markdown]], like this file) - .pdf - .odt - .docx ← "x" at the end is important, it's [[XML]]-based - Spreadsheets - [[CSV]] - *king* - .ods - .xlsx ← "x" at the end is important, it's [[XML]]-based - Presentations - .pptx ← "x" at the end is important, it's [[XML]]-based - Images - .jpg - *king* lossy - .png - *king* lossless - .svg - *king* vector - .raw - lossless, large - .tif - lossless, large - ...or use PDFs - Video - .mp4 - Audio - .flac - lossless, large - .mp3 - Data - see **data note** below - [[CSV]] - *king, tabular* - [[JSON]] - *probable king of non-tabular* - [[YAML]] - [[XML]] - *arguably king of non-tabular* > [!warning] Data Note Just because something is possible to open in 50 years doesn't mean it's necessarily *interpretable* in 50 years. If you're storing [[CSV]], for example, you may consider [[Data Package (standard)]] or something more lightweight to say simply "this is what's here". [[JSON]], uncompressed [[XML]], and [[YAML]] are both likely to be around for the long haul, but that doesn't mean they will be easy to interpret. ## My Select File Types I created a note about [[Aaron's Select File Types]], which is a subset of the above. **** # More ## Source - [LibGuides: Getting your data organised: Durable file formats](https://latrobe.libguides.com/dataorganisation/fileformats) - [Guides: Data Management Recommended Practices: Sustainable File Types](https://guides.library.upenn.edu/datamgmt/fileformats) ## Related - [[Plain Text Durability]] - [[CSV]] - [[XML]]