> Hi Valery, > > your patch looks more than reasonable. If the association > between .tcl/.rvt and the 'content_type' header is correctly > done by the server (through the module conf), I don't see why > we should check the file type in the empirical way Rivet has > done so far. The question is, as usual, for the people who > know the history of the module: is there any tricky reason > (that we may be overlooking) for adopting the basic method > of scanning the whole file name in order to extract the > file type? > > -- Massimo > Hello, Massimo.
It seems like it was just a quick hack. The single reason why it played with extensions (IMHO !) was that the Rivet_IsRivetFile function has been used in Rivet_ParseUri (it is very questionable as well), at this point content-type is not available, i see no another reason. Regards Valery --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
