Thanks, to both of you that was fast. I get the difference in file size between the provided minified source and my minified sources using Google compiler down to within 44 bytes of each other (ignoring comments).
I do have one concern. There are still a lot of differences in the compressed code sometimes it's just variable naming but sometimes it's structure. I also found that this string "parentNode.nodeType" appears twice in my compiled version and only once in the minified file from the jQuery site. If I compare the files and color the differences
Does anybody believe these differences are normal or the result of using compressor with different options?
I have attached the minified file that I created as will as the difference report using Beyond Compare 2. The report is an HTML file. To make the "Report2.html.txt" difference report more usefull I added a line break after every semi-colon in both files. For some reason I can not delete the other report from the post.
Resolved: The minified version is the same if I use the google closure compiler from the jquery repo (google-compiler-20091218.jar). Note: there is not a download version that matches this date on http://code.google.com/p/closure-compiler/downloads/list . Even if I picked the one dated one day earlier (20091217) there were, possibly significant, differences in the minified files. Do we know where the compiler in the repo came from? I think some of these differences are significant.
I think its worth noting in the FAQ that the compiler distributed in the repo should be used. ... or maybe some other disclaimer.
