Re: [HACKERS] Red-black trees: why would anyone want preorder or postorder traversal?

2017-09-11 Thread Aleksander Alekseev
Hi Tom, > In short, therefore, I propose we rip out the DirectWalk and InvertedWalk > options along with their support code, and then drop the portions of > test_rbtree that are needed to exercise them. Any objections? Doesn't sound like something that will be used any time soon. When and if it

Re: [HACKERS] Red-black trees: why would anyone want preorder or postorder traversal?

2017-09-10 Thread Alexander Korotkov
On Sun, Sep 10, 2017 at 3:25 AM, Tom Lane wrote: > In short, therefore, I propose we rip out the DirectWalk and InvertedWalk > options along with their support code, and then drop the portions of > test_rbtree that are needed to exercise them. Any objections? > +1, I don't

[HACKERS] Red-black trees: why would anyone want preorder or postorder traversal?

2017-09-09 Thread Tom Lane
There is quite a bit of code in src/backend/lib/rbtree.c that is currently dead according to the code coverage report, but we're hanging onto it with the thought that somebody might use it later. That's fine as long as there is a plausible use-case for it ... but I have to wonder what is the