Wabbajack 2.1.0.0 improvements (Patreon)

Yesterday I released Wabbajack 2.1.0.0 which includes several small, but fairly important improvements:

First of all, Wabbajack now use game files as a "download source" both during compilation and installation. This seems strange at first, until you consider how something like cleaned game ESMs could be implemented. In the past Wabbajack had specialized code for recognizing game ESMs and then would activate specialized code to patch these files. With this new release this code is much more generic, allowing any game file to be used as a source for binary patching in a modlist. This is a rather critical feature for using Wabbajack with Fallout 4 lists that repack game BA2s with optimized textures (Workbase BA2s).

Usage of game files during compilation is also no longer limited to the main game being targeted by the installer. This will be a handy feature for VR lists that occasionally need to pull in DLC or textures from the non-VR version of the game. Of course all this is done with respect for copyrights, it's just that some lists can now be crafted to require the end-user to own more than one game. This should also help us implement support for Tale of Two Wastelands in the future (a game that's really a merger of Fallout 3 and Fallout New Vegas), but there's still a bit more work to be done there.

The second major feature in this release is a refinement of how we handle binary patching. At some times it may be unclear which file should be used as the source during binary patching. Imagine a texture that existed in Skyrim, was overridden by a mod, and re-compressed by the list author. In the past the choice between these two files was completely undefined (or random), with 2.1.0.0 Wabbajack now selects the file that requires the least amount of edits (or the smallest patch file). This should slim down the installer size a bit. 

Over the past month I've sat back and taken a look at where Wabbajack is, and where I want to see it go. It's quickly approaching a full year of development, as I started working on the program at the end of July 2019. In the past 12 months the project has grown at a steady pace, and we've made a ton of improvements and changes. However at some point I have to be careful when adding new features, as after a year it's easy to add "just one more feature" and end up adding bloat that will have to be supported for months or years into the future.

So instead of adding new features of questionable worth, I've taken the past few weeks as a time of bug fixing, and polishing of the existing code base. The server side logic has been improved a lot over the past month or so, and is at a point where it can auto-heal most mod updates and keep lists up and running for weeks at a time without a need for constant recompilation. 

Overall I'm happy with where the project is right now, there's always a long list of features to add, but mod lists like recently released Living Skyrim 2, The Phoenix Flavor, Fallout 4 EE, and the other almost 20 lists we curate, are a testament that the past 12 months have been well worth the effort, and that the problem of "modlist auto installers" for Bethesda games has been solved.