I recently ran into this same issue but after further investigation I learned it was not so much an issue with the calendar as with the CSS layout. Not sure if this is similar behavior as others are seeing, but I will share nonetheless.
Apparently if you have a relative position element and then have a child somewhere further down the tree that is absolutely positioned (ie: the calendar popup), then the position is correctly calculated for the image/button/etc that is associated to that popup in reference to the page. However, setting those values on the style of the absolute position calendar popup align the coordinates to the coordinates of the relative position element rather than the page. Thus, if your relative position element's top left corner sits at x=50,y=70 and your image/button/etc sits at x=90,y=120 in reference to the page, then the position the browser aligns the popup to is: x=50+90,y=70+120 rather than x=90,y=120. My fix was just remove the relative position from the CSS for that element as doing so caused no difference in the visual appearance of my page. For more information on why absolute position has this effect, see: http://www.w3.org/TR/REC-CSS2/visuren.html#q29 Thanks, Nick -----Original Message----- From: news [mailto:[EMAIL PROTECTED] On Behalf Of Werner Punz Sent: Tuesday, March 13, 2007 6:02 AM To: [email protected] Subject: Re: [Solved] inputCalendar position when using renderAsPopup=true cervatz schrieb: > Hi! The answer is 'no', I do not use facelets. What do you suggest? > Thank you very much. Enrico I recently ran into a positioning bug with the input date and pure css layouts, something in the positioning code still is buggy. if you need a fast solution, you could theoretically use facelets in conjunction with the dojo utils to get a nice input date control. I posted the code for the control recently. (I can repost it) Unfortunatly if no one else wants to look into this I wont be able to look into this issue before next week. All I can do for now is to file a jira issue.
