Compare the amount of other things to get working correctly to the amount of effort to make the change to a feature that works, but not as nice as everyone wants (and may have ripple effects in the rest of Blender’s interface). Do that comparison and it’s pretty easy to see how low on the priority list that change would be.
Put another way: would you prefer layers stacking up or more features for actually animating in Grease Pencil?
The dev not being aware of that critical issue is what worried me the most.
This is extremely unintuitive, no layers in the rest of the world works like that.
So yeah, this is really important too.
Obvious answer is obvious.
It is, but other things are even more important.
The problem has been acknowledged, but the solution will come later.
I suggest we should start making a ruckus (again) in the future, if it seems to be forgotten then. For now at least the functionality is there.
It’s not that simple.
For the people coming from other 2d apps, this is one of the first things they will notice that is really weird, counter-intuitive and possibly a deal-breaker. Obviously for hardcore blender users this is nothing, they are at home.
So…
If that’s the only thing that they find weird, then I think we’d be doing pretty well.
one of the first things 
The developer is working on getting armatures working at the moment. He’s super fast, works closely with artists, and never seems to take a day off—I’d be surprised if the layer thing doesn’t get fixed before official release.
The GP dev is doing a remarkable job, I just don’t like how separate the interface is.
After 3 years or so i consider myself a hardcore blender user, but i still find a bunch of irritating things (like this one), but the point is (and you know whats coming now) its still in Alpha and people coming new to Blender know this (or should know this).
The argument about new users is a little-bit like the phrase: please think about the children!
Trying to make everything safe for children does not only cost a lot of energy, it turns the world into a padded cell.
Keep on staying patient, it’ll be awesome once its done.
If they mess it up, i’ll join you with my torch and pitchfork.
Since the dawn of Blender, the behavior is the same as a new line of code - this is the way all of blender works, the entire interface and every editor inside. To change it, we have to change the way we think of it so that compositing and texture slots, image layers etc all work this way. Not a simple thing to do at all. If we think of new users, then a new user discovers that GP layers are as they expect but then are confused about the other areas that don’t if we don’t approach all of the program consistently, and then the new user bails out? I think if we are consistent across blender, then it is easier to figure out one area from learning another. If we aren’t, then we are stuck.
I do not think that here there should be consistency necessarily, I think that when developers have the resources to do it, the two systems could coexist in all Blender. For example, for Modifiers Stack I think that the current system is intuitive and this should continue as it is now. But I also think that for any area that involves painting and layers, the ordering that painting programs use in layers is more intuitive. And as I said before, it will also be useful when developers implement real-time PBR painting with Eevee 
Anyway I do not think that this is something urgent to solve, as others have suggested I prefer more features before this.
Exactly.
That’s the perfect and only time to make noise, after that we’re doomed. 
Just a personal thought…
In abstract, for organizing, mind mapping… i personally find it logical.
ie.
identity: Name Surname > to optimize for easy access & association > Surname Name
date: 27.08.2018 > ordered chronologically > 2018-08-28
Of course am more of a math person and vice versa “to pretend” is nonsense, idiocy to me.
I also can’t accept or conform to dogmas (Adobe is hurting humanity).
My only proposal would be (later for everything nodes) that nodes could be organized and used in all directions. No more layers, only collections.
eg
what’s this “black hole”?
Don’t be too pessimistic, i don’t think development on GP will stop after 2.8 is done.
But i have to admit i am not an 2D animator (yet), a voice of someone who is actively working with it should have more weight.
You keep an eye on it, if necessary we make more noise.
I just think atm the noise is enough, better not to make too much or the signal is lost.
My only proposal would be (later for everything nodes) that nodes could
be organized and used in all directions. No more layers, only collections.
Sounds interesting (make another thread about it, since its OT?). Autodesk Flame does it this way.
Does the benefit justify the increased complexity?
I can imagine people who are overwhelmed by a “normal” node graph will get a panic attack if they read this proposal. 
The reason I do not insist that Blender adopt Photoshop’s top to bottom layer stack paradigm is that Adobe Photoshop doesn’t have what Blender has - Nodes. I can redirect a single image texture into dozens of node inputs in a very spaghetti fluid way that surpasses what Photoshop can do, and so I feel like the top to bottom display of layers is eventually not what we will be using 
Maybe we get node interface for Grease Pencil, huh?
I like the idea, it has a lot open to the future.
It has nothing to do with photoshop, it’s way beyond that. Not even talking about 2d apps, but all 3d apps has a layer system and they all work the same.
No point being weird.
Yes, it does - Photoshop is the exact program example that all users point to as why Blender should behave differently - I am not saying you can’t try ot make the change across the program, but it might be better to look for the way forward since these programs like Photoshop are actually slowly moving more toward a nodal system in the background like Substance Painter where there might be an UI layer stack up/down but the actual effects are achieved through something like our compositor or material nodes. Nodes are the way to go IMO, but maybe that isn’t your opinion.
I do think that consistency is key here, and changing one area but not another will just make more problems unless we define it better - like maybe all image handling will be represented as you say, in the inverse relationship similar to celluloid stacking instead of like it is where it emulates lines of code.
I agree that nodes are the way to go, it would be very cool if (or when) that happens.
BUT! In the meantime, please give users a consistent layer stack order (I mean, consistent with a standard 2D workflow, not with Blender’s own inconsistencies…). This is not about copying Photoshop, is about copying the way you work with layers of paper, celluloid, etc in real life.