I think an XML file is more practical, let's take a possible example
(only an idea):
<package="{name}">
<version="{version}" />
<replace="{package}-{old-version}" />
<reqdep0="{package}-{version}" />
<reqdep1=[...] />
<[...]/>
<optdep0="{package}-{version}" />
<optdep1=[...] />
<[...]/>
</package>
[reqdep] stands for required dependency
[optdep] stands for optional dependency
Just a simple parse using DOM and using the right way the child elements
should solve everything.
For packages being updated or new could be used some extra child
elements like:
<updated="{yes/no}"/>
<new="{yes/no}"/>
and add the check while parsing using DOM.
Isn't it more practical and efficient rather than using a SQL based
database etc...?
Regards,
Luca
P.S. Sorry for the previous mails but had some errors, hopefully solved.
----- Original Message -----
From: "Tony Nelson" <tonynelson@xxxxxxxxxxxxxxxxx>
To: <fedora-list@xxxxxxxxxx>
Sent: Wednesday, August 01, 2007 5:28 PM
Subject: Re: Back Again
At 9:42 AM +0100 8/1/07, Chris Jones wrote:
From what I see yum is doing, it download the primary, the other
file,
and
possibly filelists, /every/ time a single package gets added to the
repository. Even though 99% of the content is the same as before.
This, in my opinion, does not really such an optimum design to me.
You
should /not/ have to download /everything/ every time a single
package
changes.
Out of interest, how are these files downloaded ? Would it be possible
to use something like rsync to avoid downloading the whole file, and
only get the 'bit' that changed ?
For me this is the only bit of yum I still find a little annoying. The
new sqlite format may be bigger, but it is much faster to parse. If
the
download time could be improved, then for me yum wouldn't be OK.
There is a plan to extend the forthcoming use of delta-rpms to the yum
metadata.
--
____________________________________________________________________
TonyN.:' <mailto:tonynelson@xxxxxxxxxxxxxxxxx>
' <http://www.georgeanelson.com/>
