It depends on the ftp client you are using. For example, on Mac OSX (the ftp client from BSD), the ftp client uses the ftp client by default, while, for example, on CentOS, the client does not do this by default, and you only need to connect to the host name. Thus, it depends on the taste of linux and the default ftp client installed.
The default ftp client on CentOS (ARPANET):
ftp ftp-trace.ncbi.nih.gov cd 1000genomes/ftp/data
If you want to use the full URL in CentOS 5.9 or Fedora 18 (where I tested it), you can install an additional ftp client. For example, ncftp and lftp have the behavior you are looking for.
ncftp available via yum or your favorite package manager:
ncftp ftp://ftp-trace.ncbi.nih.gov/1000genomes/ftp/data/ NcFTP 3.2.2 (Aug 18, 2008) by Mike Gleason (http://www.NcFTP.com/contact/). Connecting to ... ... Logged in to ftp-trace.ncbi.nih.gov. Current remote directory is /1000genomes/ftp/data
lftp, also available through your favorite package manager:
lftp ftp://ftp-trace.ncbi.nih.gov/1000genomes/ftp/data/ cd ok, cwd=/1000genomes/ftp/data lftp ftp-trace.ncbi.nih.gov:/1000genomes/ftp/data>
Another, more efficient way to get a page is to use wget or curl. They work for http, ftp and other protocols.
Ursula
source share