On Wednesday February 12 2003 09:30 am, et wrote:
> I think urpmi uses either wget or curl, and one supports resume and
> one does not as default, and urpmi defaults to the one that does
> not support resume, or something like that
Default is wget, unless curl is installed. 9.1b3 installs curl, so
that's what urpmi will use. I tend to like curl better, but it seems
wget is often more bulletproof. So, if curl is installed,
urpmi.update -a --wget && urpmi --auto-select --wget
Will make urpmi use wget. From the man page,
"Wget has been designed for robustness over slow or unstable
network connections; if a download fails due to a network problem, it
will keep retrying until the whole file has been retrieved. If the
server supports regetting, it will instruct the server to continue
the download from where it left off."
OTOH, it seems curl needs a switch to support resumes,
-C/--continue-at <offset>
"Continue/Resume a previous file transfer at the given
offset. The given offset is the exact number of bytes that
will be skipped counted from the beginning of the source file before
it is transfered to the destination. If used with uploads, the ftp
server command SIZE will not be used by curl."
Thanks for jarring my brain Ed. I didn't think to look at these
man pages. It appears that --wget is a better option with problem
connections. Then urpmi should resume, even after a re-connect.
I've been usin 'cook'
(alias cook='urpmi.update -a --wget && urpmi --auto-select') as I got
tired of hdlist retrieval failures due to "too many users" and such
errors. --wget just keeps hammerin away till they finally let me in.
So currently I'm usin wget to update sources, but curl to get the
updates ;)
--
Tom Brinkman Corpus Christi, Texas
Want to buy your Pack or Services from MandrakeSoft?
Go to http://www.mandrakestore.com
