The attached patch is - IMHO - ready for commit, but I'd very like some review from a subrequest/redirect/filter guru ;-). PR 17629 deals with the problem, that internally redirected subrequests are not compressed (the main request has deflate in place). In particular they are meet the original filterchain too late (r->proto_output_filters) and thus skip the output_filters that are already in place. There are some testcases, that work for me after applying the patch. (standard installation) - the testcase described in the PR (include virtual of a cgi script which sends a Location header to a local file) [internal_redirect] - (mod_dir and type-map) [internal_fast_redirect and internal_redirect] - (mod_dir and multiviews) [internal_fast_redirect and internal_fast_redirect] - (multiviews only) [internal_fast_redirect] As said, please review. If nobody objects, I'm going to commit this fun these days. TIA, nd -- my @japh = (sub{q~Just~},sub{q~Another~},sub{q~Perl~},sub{q~Hacker~}); my $japh = q[sub japh { }]; print join ######################### [ $japh =~ /{(.)}/] -> [0] => map $_ -> () # André Malo # => @japh; # http://www.perlig.de/ #