Loose rivets;
...but this one has me beat. If indeed they were flying south-ish in the valley, they would have been heading towards the highest point towering a thousand feet above the requested. That peak is just a 'stone's throw' from the crash site.
Is it conceivable that for a moment each pilot was comfortable with the rising ground on their side - thinking they were simply to the side of a solid mass?
Having done the Innsbruck run for some years, I know only too well how a wide open space suddenly becomes a claustrophobic gully when one pilot turns the wrong way.
On the German aviation forum cited earlier, there is a Google Earth image suggesting a flight path around Mt. Salak beginning south of the mountain then circling the mountain to the right, and, perhaps in the belief that the "wide open space" to their right was the south end of the mountain again, the aircraft was turned to the right, and up the canyon.
This assumes that views of some sections of the mountain were partially obscured and that either the aircraft was in cloud most of the way up the canyon or the TAWS was, for whatever reason(s), not functioning as intended.