Closed Ghost-Terms closed 4 months ago
Leaving a note for later, maybe introduce a new arg -reltmp
which will use the old behavior.
@ImportTaste Check the build artifact, add the new flag to your script and you'll get the old behavior.
My curiosity is quite high, can I ask why you didn't just edit the script locally?
Thank you.
My curiosity is quite high, can I ask why you didn't just edit the script locally?
Mainly because I'd have to keep making local edits to every new version, and little things like that add up. I could automate a script edit, but it might sometimes break depending on what code changes are made.
Alright, I'll merge this PR now, I might not make a release at the moment, but you can use the build artifact since that's what's uploaded to a release anyway.
A recent commit (https://github.com/otavepto/gbe_fork/commit/d8c14ab13edb215c3a6cf76ab6c21ba874890f54) made it so generate_emu_config always uses the directory that the executable was ran in. I don't like this change, In a completely separate directory from the executable.
I have a script called
run.ps1
which generatesmy_login.txt
from an encrypted file with my stored credentials..bin/generate_emu_config.exe
is called with the arguments I passed to the script. Then it uses SysInternals's Secure Delete utility to dispose of it afterward.Keeping the generate_emu_config files in a
.bin
subdirectory makes upgrading convenient, I just get rid of the files inside and extract the new files in their place. But since the executable directory now has to be used, it's a messier process.I'd honestly like the old behavior back where it just uses the working directory for everything, but a parameter that makes it use the working directory instead of the executable directory would be fine too. The third option would be a parameter or multiple parameters for manually setting paths; this solution is my least preferred, though.