NP Station

The refactoring of the code is in the plans, just need to find time to do it. The expansion / porting also, as a follow-up. In any case thanks for the support guys, i appreciate it.

2 Likes

Sorry but I can’t seem to figure out how to use the pie menu, I don’t see the shorcut anywhere

Hi @bloox64, i believe it is ctrl+shift+space.

Any update on the 2.8 porting status? It would be nice to have at least point move and instance ported (since you can make them unique but not the other way around as easily afaik)

Hi @bloox64,

unfortunately, no defined timeframe for porting yet, still using 2.79.

I kow the NP Station version in addons_contrib has been at least partially updated, but I don’t think it’s working yet in 2.80.

A breaking change was done today in transform.translate so it is very unlikely that NP station will work unless fixed.

IIRC, the main blocker is still updating the OpenGL bits. I think Cambell pushed an update to take care of the transform.translate change (rBACcb83f98b593ffebce273258f8473fce87fae958f).

It’s a pity, because there are a couple other paid addons doing some of what NP does, but they’re paid and they’re quite expensive for what they do, the thing is these functionalities should be on vainilla blender, if I tell my friends coming from Max to use blender (and they will want to use 2.8, and they’re students as I am as well, and haven’t purchased max or any software yet) there’s no free addon for doing this.

EDIT: Idk what did you reply, but don’t get me wrong, I don’t wanna start a debate, but the most I’ve ever paid for an addon was 30€ and I considered it a good deal. The thing is that those other addons (only have what equals to point move here and a couple other things) and they’re being sold at 30€ which I consider overpriced. Your addon isn’t uploaded to a site like gumroad where we could pay a fair price we could consider either. I think if everyone who downloaded this paid 5€ it would be a quie profitable. But maybe that money should go to blender development, since this should be a built-in functionality.

1 Like

Sorry for the confusion, a debate wasn’t my intention. This is what i wrote:

It is for you and guys like yourself that i made the addon for.

Unfortunately, i am not a coder - the addon was my one-time fling with coding and it is an unoptimized jungle of formulas. Porting the addon to 2.80 would require a complete rewrite beforehand, in order to rationalize the code. I know how much it took to make it, i do not have that time right now. Like yourself, i have more basic priorities at the moment and i equipped myself with awesome 2.79 and the NP in it.

Quite frankly, i think it is time the developers who are actually paid for this finally make a step towards reality.

1 Like

i wish someone update NP ALIGN for 2.8
or create similar tool ( it will be insta buy for me )

there are other addon similar to this but i find NP ALIGN
more intuitive and faster because of helper ( A B C )

5 Likes

Lovely^^

We made some redesign of move and copy tools some time ago in order to optimize it for deadly hard scenes we are working with (up to 50000 objects), where creating/deleting empties takes minutes, and immediate copying takes hours.

Can be interesting

Also we are working on CAD snap proposal

2 Likes

Hey, thanks. Interesting rework :slight_smile:

You managed to compress my code line amount to abut a third, realy good skils! Having some problems with the copy command though, it doesn’t seem to duplicate the objects…?

P.s. nice effort on the FreeCAD icon set :wink:

1 Like

Thank you)
Yes, FreeCAD is a pretty much huge research I’ve started in order to figure out problems about blender 2.8 monochrome icons.
I made entire FreeCAD COIL icons pack to feel what can be done with Blender’s one)

About copying - it’s not a bug, it’s a hack.
Feature is that it doesnot makes copies immediately, but after release (pressing Enter).
The problem with huge scenes (over 30k objects) is that creating every single copy freezes blender for minutes, so copy tool was remastered in carpet bombing style.

You marking set of aims fist, then function makes copies on market targets in single operation.
That allows to not to wait through freezes on immediate copy creating in massive scenes during marking targets.

For the same purpose both functions reuses snapping empty, so it is created only once, when function launches first time and leaves in scene after operation complete to be reused on next launch.

Also added special naming for quick search in spacebar menu.

So, yeah, this redesign solutions was made with goal to fit really extreme conditions.

:open_mouth:

What was the blocker for going with copying instead of instancing in cases like that? Was the performance penalty between the two negligible once you get to 30k+ objects? It looks like you came up with an interesting solution, basically you “queue up” several changes then doing them all at once.

Point move and point copy are a bit trickier to use without any GUI. Were all the on-screen graphics stripped for performance reasons or just time constraints dictated by the project?

I did a quick look over the code, 1 small suggestion, I would recommend this

bpy.ops.object.add(type='MESH')

instead of

bpy.ops.mesh.primitive_cube_add(enter_editmode=True)
bpy.ops.mesh.select_all
bpy.ops.mesh.delete(type='VERT')
bpy.ops.object.mode_set(mode='OBJECT')

(in np_1dp.py, lines 88-91)

The end result should be the same: origin only mesh object with no vertices.

Well, there is no reason to make copies by default, because instances are needed most of the time and it is much easier to get copies from instances than vice versa.

Yes, it is much more useful to get one queued freeze at the end, that gives you ability to make other useful things during it, for example, in another instance of Blender, than get a lot of little friezes that obligates you to wait through all of them.

Suddenly, there is no more need in graphicks at that point - code was simplifid enough to have no options, so in both fuctions it does pretty much straightforward thing that is the most useful in most cases.

Nice solution for optimizing empty, thanks)

Maybe a bgl based feedback as in Jacke luke “paint objects” could help and still remain fast ?

@stephen_leger That is what came to my mind also. I am not sure will the performance suffer though, if the objects are complex, with lot of polygons. I never took the time to measure the speed of certain processes.

@1D_Inc Ok, i understand now, interesting strategy…

Was thinking about box placeholders for performances

Ah, ok. I was thinking of complete bgl ‘ghost’ copies of the selection, but that probably wouldn’t help…