© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
67d34a1
(Get the code!)
<jamespage> georgelorch: right that's what I guessed but wanted to make sure - that's reflected in the release notes for 5.6 say
<georgelorch> Honestly I have never seen those guidelines documented anywhere, but that is what we have lived by for my 5 years or so here at Percona. Release notes though are detailed and accurate, at no time will we ever try to 'sneak something through' by not documenting, if it happens, it is a mistake and not intentional.
<rbasak> georgelorch: thanks. So for the feature changes, are there any changes that may change behaviour to users in a way that the user doesn't want?
georgelorch: for example, 5.6.34-26.19 deprecated some options. Is that going to cause users any grief if they want to continue using them? Or for example does the change to wsrep_desync_count affect any behaviour from users' perspectives in a way they may not want?
<georgelorch> rbasak, what do you mean, on minor/point update, the general rule is that users should never even know that an update happened unless absolutely unavoidable to say fix some security issue.
that one I can not answer, I do not work on PXC so really have no idea what it does.
generally though deprecating an option mid series means the option should still be there, just not responsive in the same way and will go away next major release. Some of that is us, some we inherit from our upstreams.
<rbasak> Thanks. "users should never even know..." is what I was looking for :)
<georgelorch> +1
<jamespage> James Page georgelorch: thankyou for clarifying that!
<georgelorch> there are always exceptions, but that is the rule that we go by