The command 'equo repo mirrorsort' is a great idea that does not always produce desirable results. It seems to sort the list of mirrors solely on ping responses, which does not take into account bandwidth. Often times, mirrors with great pings will have very poor bandwidth and the user would be better off going somewhere else.
I'm bringing this up on the forums as opposed to a filing a bug for now because I think some brainstorming might be required as to what the correct step is.
From a user's perspective, they would want to have bandwidth tested from all of the mirrors and have them sorted accordingly. However, that sort of mechanism might not correctly balance the load for the mirrors, not to mention it would take much longer to run for the user. I don't think mirrorsort is run that often that using extra bandwidth should be a concern, but perhaps it is as well.
I'm not really sure how practical it is to implement something serverside/mirrorside as far as gathering statistics so that the mirror could give some sort of useful information for someone who does a mirrorsort is. A quick thought I came up with is if a mirror could somehow rate it's performance and the mirrorsort could use a combination of the server's score and ping/bandwidth results to sort. To be honest, I don't know what sort of information would be useful for having a server rate its performance, as I said it was just a quick thought that came to mind.
That said, it is also entirely possible that using pings is the best system to use as far as mirrors are concerned, regardless of whether it is not the best for users. And if this is the case, then it would be better to keep the mirror system happy than break it in an effort to make users happier. But perhaps some sort of compromise could be struck.
Also, I think it would be appropriate if an 'equo repo mirrorsort' ran as a part of the installation or first boot or first time entropy is run. This might already be the case though, I haven't have to install Sabayon for years now.

