Re: What's in a .htmld or .htmd?

Scott Anguish (sanguish@digifix.com)
Thu, 26 Jan 95 23:26:11 -0500


> Actually -- it's not just a feature for FTP. I use command line
> tools [htget and family] to grab html documents all the time --
> often automatically. For part of this process, I often turn all
> the relative URLs into absolute URLs -- this means that an html
> document that may have come from within an htmld is EXTREMELY useful
> without the information encapsulated/associated with[in] it.
>
> When doing this, it is quite valuable to have a somewhat intelligent
> filename (IE: something other than "index.html"). Of course, even
> a non "index.html" filename is not completely useful -- it contains
> nothing indicative of where the hell the information came from [but
> that's always been a problem w/distributed systems of any kind].
>

Yes, but I'm not entirely sure that this is the 'norm'... and its more important to cater to the average web browser/user than the special case like this... after all, you _know_ what to do to get the document, and your ht_get script could be modified to grab the <title> out of the document...


> This may be a can of worms (and, as those that know me have probably
> figured out, i usually open cans of worms with free abandon)
>

I can attest to this... he is HIGHLY skilled at this! :-)

> As an aside, has anyone given thought to inventing various
> NeXTSTEP-centric structured comments? Kinda like the postscript's
> structured comments but geared towards integrating html into the
> NS environment a bit more seamlessly. I doubt if the
> net-community-at-large would be willing to adopt it, but considering
> the enthusiasm with which developers and users are taking on the
> web as an information provision system within the confines of
> relatively isolated LANs, it might make sense to give this some
> thought.
>

OmniWeb already supports an _obscene_ number of types..

HTTP_ACCEPT image/tiff, image/tiff, image/jpeg, image/jpeg, image/jpeg, image/jpeg, image/x-xpixmap, image/x-xpixmap, image/gif, image/gif, image/x-pict, image/x-pict, image/x-portable-bitmap, image/x-portable-bitmap, image/x-portable-graymap, image/x-portable-graymap, image/x-portable-pixmap, image/x-portable-pixmap, image/x-portable-anymap, image/x-portable-anymap, image/cmu-raster, image/cmu-raster, image/x-rgb, image/x-rgb, image/x-xbitmap, image/x-xbitmap, image/x-xwindowdump, image/x-xwindowdump, *; q=0.500, www/source, audio/basic, image/x-tiff, application/postscript, text/html

the application/postscript type that could possibly be exploited this way...