| Summary: | benchmark results comparing to other ditributions | ||
|---|---|---|---|
| Product: | [OLD] Docs-user | Reporter: | Jerome <jerome.bouat> |
| Component: | Gentoo Linux FAQ | Assignee: | Docs Team <docs-team> |
| Status: | RESOLVED WONTFIX | ||
| Severity: | enhancement | ||
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
|
Description
Jerome
2005-05-01 02:52:25 UTC
it would be far too hard to implement imho. you would have to get someone willing to try at least 3-4 distros, set them up properly, configure the kernel to the same specifications, figure out which cflags gentoo should use, run the tests... not to mention that this would have to be done for every arc that gentoo supports (take a look, we support alot) even within an arc, you should/would have to run it on multiple setups. just imagine the difference in a 486 and a new blazing fast dual-core! they are both x86... i say we leave this up to 3rd parties (resolved: won'tfix?) 1. I would like some pointer to such test 2. I did not though to performing the benchmark but in this case maybe you could use a virtual machine as hardware reference (for example QEmu). 1) you can google 2) this shouldn't be assigned to doc-team if you want gentoo _iself_ to be responsible for running these tests. ask the infra team... you have to realize, however, that asking gentoo to do it is rather pointless. you are virtually assured that and person/group which has a direct interest in the outcome of the tests will make the tests such that they always win. thus, even if we publish such things, noone would take them seriously. this is why you have a 3rd party, who has no vested interest, to perform them besides, even if we were to do these tests, they would go in /main/en, not /doc/en, which, once again, shows that this is not a doc-team bug Jerome: The performance depends on your ability to build a system. And the big advantage of Gentoo is the configurability (which also allows to build a very fast system). I agree that this should be a won't fix. We'd need to build systems with exact usage profiles and rerun benchmark suites every half year or so, to get somewhat valuable results. Not that it would harm, but I think we have better things to do. If you know how to optimize your system, Gentoo will make it easier for you. If you need the flexibility offered by Portage, use Gentoo. You're asking for the performance gap on your hardware, the only way to know is to try for yourself. I have a hunch it will be slightly in favour of Gentoo. Performance would not be the main reason to switch to Gentoo. Its flexibility, community and documentation should be enough to convince you. This is obviously a very biased opinion. |
