@seco
Please don't get me wrong. I'm not making any judgement about whether this change should be included. I usually try to keep my nose out of such discussions. I just happened to be browsing this thread and saw your comment about recompiling. Again, please feel free to ignore my comment if it is off topic.
No, I did not get you wrong I was also suggesting that we should maybe postpone this to 1.8.0 but then again we might encounter merge hell so the sooner we get this out for testing and released the better. But it's not my decision.
1.6.0 broke plugins also.
I fear that if we don't merge this for 1.7.0, it will get so much harder to merge to the point where we lose the will to do it, and it becomes one of many branches we have that would be great to include but too much effort.
If we do merge this, there is absolutely no chance that a plugin will be able to support 1.6.0 and 1.7.0 in a single binary version, right?
Personally I don't see any reason for users not to run the latest version. Did we break anything so badly in 1.6.0 that a significant number of users don't want to upgrade? Even if this is the case (which I don't think it is), presumably those users would also be happy with the older versions of the plugins, so we would encourage community plugin developers to focus on the latest and greatest. It seems that many people are spending a lot of effort supporting older versions of software (which we do to some extent with the OS, including XP support which we have now ditched).