Jared Parsons

> Here's a https://github.com/github/codeql/pull/16581 that supports source generators and another https://github.com/github/codeql/pull/16672 that shamelessly takes your sample. Walked through both of those changes and they look good to me :smile: >...

> So, I decided to continue with the variant that runs source generators Is there a reason why you have to run generators again vs. just looking at the generated...

> You guys should really be injecting /p:CompilerGeneratedFilesOutputPath=Generated as well Curious: why do you all need these files? If you're re-running generators then the generated files will be in the...

> I think we should continue discussing the EmitCompilerGeneratedFiles problem in the already started teams chat, Is this a teams chat that I can join? Always interested in how adding...

Awesome! Thanks for doing this! Going to start trying this out as soon as that release is available to us.

> Did the new CODEQL_EXTRACTOR_CSHARP_OPTION_BINLOG option provide a solution to the performance degradation you were experiencing? Sorry, missed this question. Yes the binary log mode helps significantly here. Still some...

Think the sweet spot here is for `replay` based investigations. It would be really nice for instance to do the following: ```cmd > complog --compiler c:\users\jaredpar\code\roslyn\artifacts\packages\Micosoft.Net.Compilers.Toolset replay ``` Then could...

It's been a year and haven't really found a need for this. Closing for now until I have concrete scenarios that end up driving it.

@paul-schroeder-msft believe that you're running into this issue with editorconfig priorities https://github.com/dotnet/roslyn/issues/72889

Closing as 17.10.2 is now released