Resuming 'reposync' File with 'wget'?

[das]

On Sun, 2007-10-14 at 08:25 +0530, das wrote:
> If only reposync had an inbuilt wget: that it can resume a big file from
> till where it was done. It would be so helpful for us who work over slow
> connections to the Net like 32kbps. Just now I got up from sleep and
> this was the output in the reposync window, the last few lines:
> 

In a particular local directory (/mnt/data/fedora.repo) I am doing a
reposync with this command:

reposync -n -a i386 --repoid=freshrpms --repoid=updates 

This is creating a repository of the newest packages (-n) for
architecture i386 (-a) for both freshrpms and updates by taking the
repository details from my /etc/yum.repos.d/. 

Now they are getting created. 

During doing this I got a problem. For a big file
(nexuiz-data-2.3-2.fc7.noarch.rpm) of size around 260MB, reposync going
incomplete downloads of the file for eleven times, and still incomplete.
So, what I did was wget the file individually and keep it there in
the /mnt/data/fedora.repo/updates/ directory. But, this was a brute
force method. Is there any better way of making reposync use wget in the
backend. 

The next problem is, once the repo is done and complete, how I am going
to use it in updating a system after the installation of F7 is done with
a dvd? I have an external HD of forty GB for this purpose. Can any of
you please provide a step by step guide? See, I am absolutely new in the
F7 way of doing things. 

--
das