Software development is fraught with trial and error. Case in point - this URL/data issue I've been blogging about. I wrote a rather cool CGI in C which actually does decode a base64 encoded data (see data.c source). It is fast and very straight forward. It worked beautifully with a binary encoded vCard. When I moved on to jpeg photos I found yet another issue with Internet Explorer (Q208427). Even though RFC 2616 states there is no limit on Uri's, Microsoft put one on anyway (2,083). After a bit of research I found (as usual) no one else has a set limit, unless it's a really old web browser. So my grandiose ideas of delivering a solution with a slick CGI has been ground to dust. Please pause for a moment for the short life of data.c.