Re: Refactoring the tracker
On Tue, 04 May 2010 20:34:38 +0200, Florian Weimer wrote:
> I've decided that it's necessary to clean up the mishmash between SQL
> and Python in the tracker code base. As I've mentioned before, there
> are three or four different ways for deciding if a vulnerability is
> fixed in a particular package version. This makes changing the
> processing logic and creating new statistics rather difficult.
> Another issue which has gained some significance lately is that the
> package and CVE lists have grown quite a bit, leading to longer and
> longer processing times on soler. I've removed a few unused features
> to speed things up a bit, but it seems that we're at the bare minimum
> right now.
> Are there any suggestions for programming languages and web toolkits?
> (A more typical approach to HTML generation would probably make it
> easier to copy the official Debian web site design.) We can certainly
> stick to Python; it's a bit on the slow side, but perhaps we can reuse
> some of the code.
> Constraints are: it has to run on soler eventually, and I don't think
How about making use of a more standardized set of python features such
as dictionaries for the database, and possibly storing those to disk
using pickles or yaml (via the python-yaml package)? I agree,
stick with python.