Hello. I traced the url given at command line, and it looks like there is no difference if one gives "~" or "%7E". Is this true? The urls end up to url_parse() which switches "~" (as unsafe) to "%7E". If the original url is not used at all as it looks like, then there is no difference. But mysteriously the url with "~" and the url with "%7E" downloaded files differently!
I also added new log outputs and while testing them with the problem sites, surprise, there seemed to be no problems. So, the fact that urls are not downloaded, could be just some code bug in wget. But why this problem appears when "~" is in the download url? Have I just missed the other cases? Or is the bug in the code which expands the unsafe characters? Juhana
