On Thu, Jun 11, 2009 at 9:51 AM, Eric Springer<erikina@xxxxxxxxx> wrote: > Likely for the same reasons we have a bug tracker instead of a 'what > works list'. But I agree, Fedora performed quite favourably especially > taking into consideration the database benchmarks. The problem with knowing that Phoronix got a head scratching result for Apache.. doesn't really tell us anything useful. If I were going to treat this as a bugreport I would have liked to have seen at least a "me too" effort to confirm the result from the person...contributing..the benchmark snippet. I personally have very little faith in the laypress's ability to communicate information developers can actually use to make sense of unexpected problems as its in their own interests to sit on those problems and write about them outside of the affect project's communication and bug reporting processes. If this is going to be taken seriously an actual tester or contributor needs to start trying to confirm it..someone who can be relied on to work with the package maintainers and developers. The laypress reviewers who do things like run benchmarks are as a breed highly unreliable when it comes to actually HELPING diagnose the problem. Numbers for the sake of numbers isn't the point. The benchmarks are only as useful as your commitment to followup on diagnosing potential problems. The laypress continues to miss the point about having an open development process by which users can engage with developers. I live for the day when each and every problem reported in articles written by the technical laypress, comes with cross-referenced links to bug reports opened by the laypress "journalists" who discovered the problem so the discussion on the diagnose of the problem can continue where the right people..the people who can provide and integrate a solution..can actually deal with it. No the technical laypress isn't actually interested in seeing problems solved...they just want to find things talk about. The worst part is there's no obvious place to start looking for a difference for the Apache benchmark. The article doesn't make any suggestions as to where the difference is. Unlike the filesystem tests where there is a clear difference in underlying system configurations and even Phoronix picked up on it as a probable cause. Numbers for the sake of numbers. Do they even understand how to interpret their own test suite as a diagnostic tool? I guess what we really need is the same test run on stock F10 and F11 on the same hardware and see if there is a regression there. If its selinux latency they'll both be impacted and it should be a wash. -jef -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
