On Thursday 29 March 2001 09:38, Andy Stoffel wrote:
Ned Konz wrote:
- I don't need to know your file extension (.html) (that's what MIME
Content-type headers are for!)
[... continuing off-topic ...] Unfortunately, there is a well known web browser that is known to use file extensions INSTEAD OF Content-type headers. (This isn't just urban legend. I work with an operating system on a daily basis (OpenVMS [the 'Open' is silent]) that runs headlong into this issue. (e.g. Assorted versions of Internet Explorer ASSUME the anything ending in .com is a binary image. Oops.... on VMS they are ASCII DCL command files (the DOS equivalent being a .BAT) and if I attempted to use IE to look at one of these it attempts to download it. I don't use Scamper on a regular basis so I'm not aware of how IT handles such things.)
This is an excellent reason not to use file extensions AT ALL: it forces the browser to use the MIME types!
It's really just programmer laziness that makes us see URL's as a mirror of file systems, anyway... I'd rather not know how you've organized your web server's file system. Just give me a clean, unchanging URI space that I can bookmark. Free of implementation details (.html/.gif/.jpg extensions, server names, .cgi/.jsp, etc.).
squeak-dev@lists.squeakfoundation.org