Thank you. I have noticed what wget has the error in the downloading. I didn't make statistics it's errors: servers, files, time. I must download files two time then. And compared it. If I didn't have checksum. Is in new wget other algorithm continuation downloading (option -c)? I downloaded much staff from sourceforge.net so I used wget. Wget can get direct link from soureforge.net. Other downloader get html-page with link.
I downloaded from other place with the errors. For example: http://cdimage.debian.org/cdimage/ ftp://ftp.yandex.ru/ http://files.uk.pfsense.org/mirror/downloads/ ftp://ftp.heanet.ie/pub/arklinux/ ftp://ftp.cc.uoc.gr/mirrors/ ftp://ftp.linux.kiev.ua/ http://ftp.cc.uoc.gr/mirrors/ ftp://ftp.uni-ulm.de/mirrors/ http://distro.ibiblio.org/pub/ I say maybe wget has bad algorithm continuation downloading.
