however, an .htaccess in a child directory is described as “overriding” one in a parent directory, and that description appears to be accurate - i.e., if a directory contains an .htaccess, the content of its parents appears to me to be completely ignored, not “merged” as i have sometimes seen it portrayed.
if the base of one [sub]domain is x and the base of another [sub]domain is y, their relationships to each other in the filesystem do not appear to matter vis-a-vis .htaccess files, given the above, assuming that both x and y contain an .htaccess. so, in what way is this an issue as it relates to .htaccess files? while your suggestion might bear upon filesystem maintenance issues, in this case the physical nesting is a help to maintenance, not a hindrance. (and i quite disagree with your assertion that peer directories are more proper for subdomains. (generally) a subdomain’s raison d’etre disappears if the parent domain disappears; this implies that a subdomain is logically within the context of a domain and its contents are rightly manifest as being subordinate to that parent context; hence, the subordinate directories).
my sig causes an error on some email address collection software that tries to parse out substring removals where instructions for doing so are explicitly provided.
tmike at your dot pants com
(to reply by email, drop your pants)