Comments on: Add Some Color to Your Web Site – and the Site Search http://newsletter.sli-systems.com/2009/06/add-some-color-to-your-web-site-and-the-site-search.html Sun, 7 Feb 2010 05:46:42 -0800 http://wordpress.org/?v=2.8.4 hourly 1 By: shaunr http://newsletter.sli-systems.com/2009/06/add-some-color-to-your-web-site-and-the-site-search.html/comment-page-1#comment-2 shaunr Thu, 18 Jun 2009 22:01:45 +0000 http://newsletter.sli-systems.com/2009/06/add-some-color-to-your-web-site-and-the-site-search.html#comment-2 Yes - we have some work arounds we do for when the encoding is not quite right like this. We were in the process of fixing this when the screenshot was taken. Yes – we have some work arounds we do for when the encoding is not quite right like this. We were in the process of fixing this when the screenshot was taken.

]]> By: searchtools_avi http://newsletter.sli-systems.com/2009/06/add-some-color-to-your-web-site-and-the-site-search.html/comment-page-1#comment-1 searchtools_avi Mon, 15 Jun 2009 22:31:01 +0000 http://newsletter.sli-systems.com/2009/06/add-some-color-to-your-web-site-and-the-site-search.html#comment-1 I agree with you on the value of color, it's very positive for people who want those colors and less-negative for people who don't, as they can skim rather than click and come back. But one of my pet peeves shows up: the broken extended character, presumably a curly quote in the product name, which turns into something like this  -  - A with a circumflex. I've found this very common in HTML created by Microsoft Office products, and it's a shame. I think you must have fixed it since then, but it's one more little problem. Thinking it through, any normalization you do in the inverted index should skip that, but how can it tell that it's not an actual word, RiderÂs? Even if you get it right there, adding it to the document store would have to fix it as well. Or maybe just flag it and bug the content owners? I agree with you on the value of color, it’s very positive for people who want those colors and less-negative for people who don’t, as they can skim rather than click and come back.

But one of my pet peeves shows up: the broken extended character, presumably a curly quote in the product name, which turns into something like this  –  – A with a circumflex. I’ve found this very common in HTML created by Microsoft Office products, and it’s a shame. I think you must have fixed it since then, but it’s one more little problem.

Thinking it through, any normalization you do in the inverted index should skip that, but how can it tell that it’s not an actual word, RiderÂs? Even if you get it right there, adding it to the document store would have to fix it as well. Or maybe just flag it and bug the content owners?

]]>