All good points - it all just comes down to availability vs. quality of information. For the vast majority of applications, I'm thinking that a somewhat curated repository of XDFs would meet most people's needs. The vast majority of people need access to VE tables, spark advance, and VATS flags, all of which are pretty easy to get good XDFs for. The addition of an XDF comparator would help narrow down the herd and knock out the obvious inferior or incorrect XDFs. I would likely choose stickied XDFs that are known to be up-to-date and maintained over older legacy XDFs with a mysterious origin. The idea I have is people post new XDFs in the form of a pull request, maintainers of the repo test and vet, then it is integrated into the repository. Initial population will require some legwork but ideally it is a self-sustaining machine. That way issues can also be discussed in one location, instead of me going to the BoredTruckOwner repo and grabbing some random XDF, and having no way to correct issues in the posted file for other people. I support the idea of an XDF area on these forums as well - it would help reduce noise.antus wrote:I think you will need to track sources somehow, and keeping consensus on what is the newest/best/most accurate will be key. Properly maintained xdfs are much more valuable than a 3 year out of date xdf someone with good intentions picked up, made one change then posted as newer. I think community will be a big part knowing who is/has done what. That is where the forums are useful. A comparison tool will be great, but encrypted XDFs can cause a problem there, and it'll require testing to know which is correct when there are differences. The XDFs in stickied threads here are the best currently available. Git is a good idea for trackability but tunerpros slightly changing format over versions will create noise. I also dont want to break the 1 thread / 1 maintainer community side of it here. If we need a GM definitions area only for xdf threads with OSID in the title we can do that. Keep xdfs posted here for feedback/comments and link out to github for trackability. I imagine we'd share access to the repo among a couple of active members to keep it up to date.
On the note of password-locked XDFs, would someone mind explaining what they are? The definition itself is encrypted, or the usage of it is password locked? What is the ethical reason for doing this - just intellectual property?