On a related note: Is it even considered "good practice" to take a function which returns a page, and attempt to modify it by (first "unrwapping" it and then) appending another fragment of xml to it (and then "rewrapping" it into a [transaction page] again)?
This is what I have been trying to do in Exercise 7.3 of Gian Perrone's tutorial - append a logout link to all the pages on the website (when the user is already authenticated). Then I realized that my function [pageWithLogout()] might not work very well. For example, given a page [pg], it tries to output another page made up of: an empty <head/> tag, with the "unwrapped" [pg] appended to it. What if [pg] itself had a head tag? I remember some framework I was studying was "smart" enough, when joining <xml> fragments, to "merge" all the the content of their <head> tags into a *single* <head> tag - but I can't remember if that framework was UrWeb (maybe it was Yesod? :-).
_______________________________________________ Ur mailing list [email protected] http://www.impredicative.com/cgi-bin/mailman/listinfo/ur
