There are a lot of other things wrong with GMX as well though, like the fact the tags and attributes are not consistent. You already pointed that out in the past and you have documented it here:
http://enigma-dev.org/docs/Wiki/GMX_FormatSo what I want is a clean, consistent, backwards compatible, future compatible and extendable format. Making EGM with YAML just because GMX uses XML is stupid at best. We already want to support new things like custom resources, includes in EGM, version control with an extracted format (EGX? And I know GM has GMZ here), reordering the resources with several different types in one folder and so on. I cannot do any of that without breaking GMX, which I don't think we need to use at all.
But YAML ain't the only thing - we need EEF as well, because e-YAML cannot save code. But I guess it wouldn't be that pretty in XML either.
Long story short - I still stand by XML because it's more standard then coming up with two of our own formats, even if one of them is a subset, which doesn't mean anything if we are the only ones who implement it.
But in the end I don't really care what you choose (like I mentioned previously), but at least remake the f-ing thing e-YAML+EEF so it at least doesn't break.
We also need a proper specification for EGM, because now it is very ad-hoc. We don't have version information in EGM, so if we do change something we cannot even differentiate. So it is very easy to break it. We also need a lot more delta checks there, so things that are default are not saved. Like every EGM saves an empty .rtf file (still 163 bytes) which in 99.9% cases are not used, because who the hell uses that information thing anyway. It was used during the age of examples back in 2005, but not today.