|Currently our Sisulizer projects take around 4 minutes to run, which doesn't sound like a long time, but taken in the context of an end-to-end automated build time of 10-15 minutes, it's the single longest step in the chain. We're using source localization with Visual Studio solutions and most products have at least 4 (sometimes more) Visual Studio projects within the solution. Depending on the product, there are anywhere between 4 and 10 languages being output.
Given that the output from Sisulizer of each Visual Studio project is independent and each output language is independent, this seems like a great opportunity for running things in parallel in order to speed up the overall time. Our build servers have 12 cores and an SSD, but when Sisulizer is running only one of the cores appears to be in use. Is there any chance of exploring this as an avenue for decreasing the time to generate all the satellite resource assemblies?