> If you could actually quote us a URL then we could check this theory but > until then it remains a theory.
Malcolm, below is a link to a search box that will search one of my sites. You can do a query on whatever you would like, but I suggest using "Randy" as one of your searches, since I know that will show you my problem. (some searches show as [filename.html], and others show the actual tittle). http://fox97.com/common/htdig/ssi/wfox-fm.html Please let me know if you have any other ideas/comments :) Thanks again, -daniel Malcolm Austen wrote: > On Tue, 8 Jan 2002, Daniel Escobar wrote: > > + >index the contents just fine and anyone following the hit link with IE > + >will get it displayed just fine (but probably not in NN and in Opera it > + >depends on how you have it configured) ... but ... because htdig didn't > + >analyse the html content, it will just show the filename in the search > + >results. > + > + I don't understand how htdig would be browser dependent, when it runs > + on the server side. > > No Daniel, you've missed a point. HTSEARCH runs server side but HTDIG runs > client side when indexing the pages and the user who is looking at the > search results is also client side. > > + Most likely, I didn't explain my problem well enough, so here it goes > + again: When an htsearch query is performed, the results page contains > + some results with the appropiate tittle, and others with the filename > + btw brackets. [filename.html] Why is that happenning, when all the > + files have the <title>...</title> tags? > > You've convinced me still further with that extra comment. The filename is > .html and yet you have exactly the behaviour that I see when my htdig > indexes a .txt file. I surmise the following: > > You have a page coded in HTML but when htdig (the robot) fetches it the > server hands it over with a MIME content type of text/plain. That being so > htdig does not attempt to decode the HTML structure and just indexes the > words. (The page should show up on searches for 'html', 'title' etc.) > > When the user does a search the page shows up as [filename.html] but, > using IE, when the user clicks on the link the browser decides to ignore > the content type and display it as HTML because Bill thinks he knows > better than to believe the content type. > > This is (one of many situations) where the Opera web browser comes in very > handy. It has configuration options either to believe the MIME content > type or to ignore it and believe the file extention instead. > > If you could actually quote us a URL then we could check this theory but > until then it remains a theory. > > regards, > Malcolm. > > [EMAIL PROTECTED] http://users.ox.ac.uk/~malcolm/ -- Daniel Escobar Software Engineer Cox Radio Interactive 678.860.2749 (cellular) 404.979.7860 (office) _______________________________________________ htdig-general mailing list <[EMAIL PROTECTED]> To unsubscribe, send a message to <[EMAIL PROTECTED]> with a subject of unsubscribe FAQ: http://htdig.sourceforge.net/FAQ.html
