Dag Wieers wrote:
We have a document for users were we urge them to harass the integrator already. But I think it makes more sense what we expect from integrators than to proxy-by-user without any guidelines.
yes. Also, the page we point people to at the moment is focused around VPS's - which are a large part of the culprits, but might not be the whole set. I wonder if there is an easy way to identify the other 'culprits' ( for the lack of a better term ), and maybe have specific wiki pages that address their own 'special' case.
Right, and then we get complaining users asking us why they can't do this or that and then this is because their yum is broken by design. Or they have no security updates. And what happens on the forums/IRC, they get told they are NOT using CentOS.
'using CentOS' or 'using something derived from CentOS' or 'using something that includes portions of CentOS' could potentially be very different from each other.
You get 2 parties hating each other and the guilty party earns the cash and maybe doesn't care :)
ok, so a two pronged approach - one targetting the providers, and focus on best practises and another set of wiki pages, working on a 'recovery path' that users of these setups might be able to walk through to bring their machine upto a stable status.
It can, ofcourse only be a best effort from the CentOS side of things - eg, you cant 'fix' a OpenVZ install, its not going to work if you fix it all the way. Neither is a Plesk or cPanel install. If you 'fix that' the product itself stops working. What we could do is make the situation clear to both sides, and give them both material to work with.
my 2c worth.