mercredi 11 août 2010

The prototype should be online this month

The "Image Zoomer" has been reworked a little bit: there is a small timeout at both the mouseOver and the mouseOut events. Now it also Downloads a bigger image when zooming (higher quality), so its more than a simple zoom that would use the same image but just resized (low quality).
All business cards are now rounded top and bottom, that simple design change makes an enormous difference, so much that I am wondering why did not I try that earlier? It's so much better.
Nearly all the code is now generated from jSon data, but the generators still need more work to make them fully fool-proof for production. Also there is a trick to make the search-engine work: all words (with the possible exception of common stopwords) must be enclosed in html tags: <i>"word"</i>, this should be implemented today.
The new HTMl structure is very strict, this has the advantage of making the precise indexing easier, so much that it can be done with basic javascript, no need for big stuff like jQuery or other behemoths (they are good, but sooo big that I would like to lessen their use to get a even faster page load). It will be possible to index each informations in the business card more precisely, subject by subject, so searches would compare adresses with adresses, summaries with summaries... Now each search is global, a search on the summary also searches in the adresses at the same time it looks at the title, the phone number. That will be implemented later since the actual search method is quite good and there is more need for making pages for other cities than Montreal.

lundi 9 août 2010

New Zoom javascript design

It could have been be done with simple CSS code but you get a smoother user experience using a javascript timeout to init the automatic zoomer.




Plus the new background of each business card (vCard or hCard) is a simple gif that should work better in obsolete or/and old or/and buggy web browsers. It's OK to me because the new design is way simpler visually, the HTML is also new and very very simple and is completely generated from a jSon dataBase that is very easy to write and read. For that I have writen a 3 javaScript functions that write HTML from jSon. The main idea was to find a way to represent an HTML structure in a function, so it would be very easy to modify and that I could write like :

          htmlTag('html',
               htmlTag('h1',data1),
               htmlTag('p',data2),
               htmlTag('div',data3)
          )