I'm quite fond of KDL and intend to use it in a future project. Quick said-contras: Q: "Another markup/config language?" A: Yes. We actually do get (slowly)

I'm quite fond of KDL and intend to use it in a future project. Quick said-contras: Q: "Another markup/config language?" A: Yes. We actually do get (slowly) better at solving already-solved problems. Evolution of spec does happen and is important. Steady refinement and reflection on past work just is progress. Q: "OK, but is it worth the mental clutter and cognitive load of an additional standard/dependency/format?" A: That's for the market and community to decide together. Example: Time was, Markdown was a bizarre hobby format and most work happened in XML or LaTeX or Wiki (or, gasp, Perldoc). But Markdown proved better, more sensible; so Markdown won. Q: "Why not YAML?" A: Am I the only old here who remembers the era of YAML security bugs? Like, it was like the Sendmail Bug Era, but worse, because it was in everything. YAML, like Perl, is easy to read, but hard to reason about; and that makes it a hazard. These days, we have YAML libs and practices we regard as known-good, but I still trust it about as much as I trust \.doc(x?) or \.pdf, or some weird execv() involving ImageMagick. It's unclean. Sorry, not sorry. KDL is solid and I'm a fan.