What happened? Due to the recent XZ-Utils drama I checked the code and I'm appalled. There are more BLOBS than source code. https://github.com/ventoy/Ventoy/tree/3f65f0ef03e4aebcd14f233ca808a4f8946...
Ventoy is a tool to make a USB with multiple ISOs bootable, letting you select which ISO to use on boot. Another newly-created account claims to be the dev's friend and translator and has received no contact from the maintainer.
so ive deep dived into as much information as i can find. the TLDR is the main dev LongPanda supposedly went on a vacation to china (most likely his home country?) to which there is conflicting information on his return. one path is he make a lemmy account 9hours ago and made a message that doesn't describe the blob and sound like a gpt response. to which his "irl friend made an account 3 hours ago to comment that he hasnt heard from LongPanda in months. both were removed from lemmy.ml because of suspected impersonation. the other side of the coin is the LongPanda is still gone and hasn't addressed the blobs.
after looking thought the documentation, you can build from source. in the instructions it says "5. Binaries
There some binaries in Ventoy install package. These files are downloaded from other open source project's website, such as busybox."
i am not a programer but in the source build it lists the blobs and were there from supposedly from other FOSS projects with sha256's. so theoretically you should be able to verify the blobs, with the sha256.
I like Ventoy, it's handy but I don't think it's indispensable so probably what I'll do is go back to using Etcher (which is open source AFAIK) until this resolves itself one way or another. I assume either the dev will respond properly with an explanation and everything will be fine, or someone will get fed up enough to fork it. I feel like it's probably nothing nefarious, but it doesn't really hurt to be overly cautious in this case IMO.
Imho it will be much easier to replace blobs with verifiably correct blobs or add the source to build them than to retroactively find the original builds from whence they came.
Searching for some of those binaries looks like it would require comparing the hash against a large set of candidates which would need to first be unpacked from releases (fedora mostly???) and hashed unless the hashes already exist somewhere.