DISCLAIMER: These notes are from the defunct k8 project which precedes SquirrelJME. The notes for SquirrelJME start on 2016/02/26! The k8 project was effectively a Java SE 8 operating system and as such all of the notes are in the context of that scope. That project is no longer my goal as SquirrelJME is the spiritual successor to it.
Going to attempt routing today.
What I currently have which is a list of inputs which get turned into outputs in the routing stuff is not going to work because inputs could be files or they could be abstract representation of code. Also CodeRouter should not be using binary formats at all and should only handle compilation of code. Having it handle binary formats complicates things greatly.
Having binary formats directly read and write namespaces sounds like a good plan because it simplifies the plugin operation once the needed data is loaded and such.
Many of these warnings, especially with GCJ are really annoying, there are just tons of them. I need to clear them out.
Another thing I have to do is make it so projects do not need to needlessly duplicate dependencies when individually built. That is, all projects get their dependencies determined after simplistic generation so that way it is not required to include them multiple times in every single package.
Less warnings about unchecked stuff is much nicer and much faster.
Recursing into dependencies in the shell scripts should make handling of them easier as not everything has to include the stuff now. Now for my unchecked stuff, I need an instanceof so that GCJ shuts up about that too.
So far the normal stuff with GCJ takes 88 seconds to build, and OpenJDK 6 takes 153 seconds. So compared to GIJ/GCJ, OpenJDK 6 is about two times slower.
Linux TTYs and PTYs are sycnhronous, so if your terminal emulator is slow then everything being piped to the console is also slowed down.
Based on what I wrote in BinFormatRouter, I might not even need the CodeRouter that I created.
I currently cannot know of the redundency between BinFormatRouter and CodeRouter, I can still use a CodeRouter to translate between the internal code representations of of code within Namespaces. Since I will have to convert code from the intermediary format, various machine languages, and Java byte code. I do not want to duplicate some functionality but it would be more specialized.