tarsa / lossless-benchmark Goto Github PK

Make raw JSON dump downloadable using a button like in CSV case. That will unclutter UI.

Guide should explain:

• how benchmark is built
• how it works
• what features it have
• how to contribute results (by raising issue, pull request, writing on encode.su, etc)
• how to prepare results to submit (e.g. how to measure memory usage)

Also fill README.md (in main directory) with project description.

If someone has US locale (or a comparable one) then thousands separator is a comma which matches current fixed thousands separator in decompressor sizes. But when someone has incompatible locale, e.g. Polish then thousands spearator is different (space in case of Polish locale). Splitting decompressor size to two columns in CSV, one numeric and one string will allow to solve the problem in a generic way.