On Fri, 2007-09-07 at 10:19 -0400, James Antill wrote:
>  But then it'll drop all but one of the base URLs from the "upstream"
> repo. ... this seems like a bad thing to promote. For instance say
> Fedora derivatives started doing it, that would be a very bad thing.
>  I can sort of understand the desire for the feature, but it seems like
> it should be implemented differently to expose the other repos. in some
> way.
>  For instance this means you'd have to pick specific values of
> $arch/$basearch etc. ... and proxy/proxy_username/etc. could only be
> specified once.
>  Another way to implement the same kind of thing might be to use
> something like the "include=url://" config. option for "subrepo".
> 

I don't have a problem with the idea of being able to refer to other
repositories in some abstract way w/i the metadata.

>  Also one of the things I'd thought about for a future feature would be
> to do multiple range requests on different servers to download a package
> (I'm thinking about things like 100MB+ open office/kernel debuginfo
> packages mainly).

Not so keen on this. This typically means adding in threading to the
mechanism and the last time we took on threading w/i any yum program we
had more bug reports from a frozen application than anything else - the
old yum-updatesd w/it's umm, interesting threading, screwed up more
things than I can mention. I think any threading added to yum will
require an assload of justification.

-sv


_______________________________________________
Yum-devel mailing list
[email protected]
https://lists.dulug.duke.edu/mailman/listinfo/yum-devel

Reply via email to