---- John Coleman <[EMAIL PROTECTED]> schrieb: > Yes filtering is on, but why would the filter want to process binary > assets? It should ignore them.
How should it detect binary files? Should it scan each input file from start to end, looking for non-ascii characters and if none are found, then start again with filtering enabled? What about UTF-8 input files? Is there actually a way to tell the difference between binary input and unicode? It would be nice if this were automatic, but I'm not sure it's possible. It certainly isn't automatic now; when filtering is enabled then every file is filtered. Regards, Simon
