Organizational Effects?
Posted Sep 25, 2003 2:39 UTC (Thu) by
stevenj (subscriber, #421)
Parent article:
The Great Package Management Experiment
Thanks for the article; it's nice to see such a systematic comparison. The "system upgrade" step is a bit hard to compare, of course, since the magnitude of the upgrade would vary for the different systems; perhaps the number of packages upgraded should have been included as a metric.
What I would like to see would be an experiment that measures organizational effects: the degree to which different packagers and package sites for a distro are centralized/coordinated. (I suspect that this has a greater practical impact than the technical differences between packaging tools.)
For example, pick 100 randomly selected packages from freshmeat.net (or similar), and see how long it takes you to find packages and install them (verifying that they work) on the different distros. (Of course, the package versions might differ; you could collect statistics based on the age of the packages available from the "default" sites, if any, plus the time to find and install the most recent versions.)
(
Log in to post comments)