there are still many to-do’s on the list, but development is currently stalled.
Snapping & precision modeling improvements
I can’t say exactly why,
maybe there are other priorities, or maybe the developer who is developing them has other things to do … who knows
For my part I can say that there was a change in what I was told to do.
Due to the large increase in the number of reports created per day, I and others were moved to focus almost exclusively on the triage of these reports.
(But that may change over the next 3 months).
thanks for the clarifications 
So face center snapping was never part of the plans? 
I do agree that this is super important for architectural projects and would be really happy to continue seeing progress with snapping
Ok, just making sure it didn’t die on the vine- I worry about development fund stretch goals being promised and then killed off and potentially scaring off donors, but it seems like that is not the case here. Carry on!
Wish there was snapping for Slide operations, like Max snap tools work with Edge Constraints so I can slide and snap to what ever my snapping tool is set, while Blender snap for slide snaps it to factor value of slide tool?
good point. I dont know how blender users dont feel the need of a face constraint (slide) as well…
If you had a contributor or two who wanted to help lay the groundwork, what would your ideal road map look like?
What if the the global transform settings (snapping, moving the origin, etc) were mirrored in the operator properties, and a ‘use global settings’ property was added to maintain the current behavior? The default user experience wouldn’t change, but it’d make it easier to create design mockups.
A combination of behind-the-scenes tweaks, experimental features (preview features? I’d heard rumblings that the tab will be re-named), and dev add-ons might make incremental improvements more feasible.
Cross posting as suggested (click the link for GIF):
Germano has explained this before.
Snapping in Blender is very basic at the moment, it simply calculates a bounding box of the transformed elements and snaps the closest boundary to the target.
That is why you can’t easily snap say the side corner of an hexagon, or the opposing sides of contiguous geometries.
I don’t care about how it is. I care about how it should be.
This is one of the reasons behind PDT, here is how we set the offset between the two verts, using our Set A/D 2D button:
After the Move, using Delta button:
I know it doesn’t solve the problem of snapping directly by nearest vertex, but it gets over the issue of awkward movements.
Cheers, Clock.
I swear I didn’t start this, but apparently more people think this is poorly done (including a prominent add-on developer):
Also, there’s a “mandela effect” going on in a previous thread where people thought that 2.7 behaved differently, because this way of doing things is so illogical (and also because the UI doesn’t speak the truth).
I jumped into the discussion in Machine’s bug report just to realize the planned “base point snapping” should solve exactly this problem… I hope
Hey look. Another one:
@mano-wii
I think this particular type of snapping should have a certain priority in the roadmap, not because blender does not have its own method, but because it is the most common type of snap, and because many people who have migrated from other software consider it a bug.
It was already proposed.
Including direct conversation with devs.
A year ago.
After decade of waiting.
Addons are written.
Developer found.
Closest explained
Rightclickselect got top.
Task is already created.
That’s how “I care how it should be” actually looks like.
when these functions are by default on blender, i think i will have an orgasm 
I threw together a small patch that adds a ‘live link’ transform orientation. It isn’t as flexible as the custom orientation (where the custom orientation lets you create orientations based on geometry selections, this one’s strictly based on object orientations) but it comes with the benefit of not needing to create/update the orientation after you rotate the source object.
This orientation works with parenting, too, which means you can use it to create object-relative orientations with armature-like hierarchies of parented empties. The UI lacks polish, but I’ve had a ball playing with it.
