Asset Manager “Basics” #73366

Open
opened 2020-01-24 17:16:15 +01:00 by Dalai Felinto · 136 comments

The description was moved to the #asset_browser project description (click the tag to see it), the Milestone 1 description to #asset_browser_milestone_1.


Milestone ?: Online Repository

Here we want to be able to append assets from an online repository like the Blender Cloud, BlenderKit and similar.

There would be a Python API to register custom asset repositories as part of add-ons. The main operations would be to provide a list of assets and metadata, provide .blend with an asset to append/link into the scene, and functions to add/remove/update an asset.

Milestone ?: Usability

Once the asset manager is working for the basics and the UI can do its job, we can start to make things nice. This includes better handling of thumbnails, re-configuring the ID browser to allow for asset browsing, a nice UI widget for managing tags and tag-based filtering.

A key part of making the UX nice, is to make it effortless to drag assets into the scene. When dragging in objects & meshes, they should be able to snap to surfaces. When dragging in materials, the underlying target objects should highlight on rollover. You should be able to add multiple assets at a time.

Milestone ?: Collection Variations

Not strictly part of the asset manager since this would be a core feature of Blender also usable without the asset manager. We would like to support variations for all datablock types, but the first step could be collections (it's effectively like automatically toggling visibility on subcollections to show only one).

Once this is in the place, the asset manager will need to be modified to understand variations.

Milestone ?: Project Repositories

Up to this point the asset manager will handle only the use case of appending files from a local or online repository. The next step would be making it works also for projects.

  • The root project directory would contain a .blender_project file indicating that it is a project (similar to e.g. .git)
  • When opening a .blend file within the project folder, Blender will automatically know which project it is part of
  • Folder structure, file names and file saving are left to the user. The asset browser will not add or remove .blend files in the project. That's all done using File > Save and file browsers as usual.
  • The assert browser will have a "Project" repository that can be chosen, which will show all assets within the project
  • The user will have to mark relevant datablocks as assets in each .blend file in the project so it shows up in the asset browser.
  • Assets can be linked between files within the same project. This is different from other repositories, where only append is possible. Linking between projects or to other repositories should not be possible, a project must be self-contained and not link to any .blend file outside of it.
  • The project folder would be indexed exactly like other asset folders.

The idea here is that we don't want to make many assumptions about how users might structure their project. It should be possible to simply add or remove files, reorganize project directories, and the asset browser should keep working. It will merely provide a view on the existing project structure, and functions to create links to other assets in the project. It will not introduce an alternative linking system as an earlier iteration of the asset manager design did.

This system does not preclude a studio from making an add-on that takes more control over file saving, version control, partial repository checkouts and so on. Such an add-on would likely need to integrate in various places in the Blender user interface, one of which can be the asset browser. For example it might include a custom asset repository to browse assets that are not on the artist's computer, and right before the user links an asset into the scene, download the corresponding .blend file.

But for most projects that are not e.g. feature films, the built-in project repository and automatic indexing system should already be quite powerful.


Further links:

The description was moved to the #asset_browser project description (click the tag to see it), the Milestone 1 description to #asset_browser_milestone_1. ---- **Milestone ?: Online Repository** Here we want to be able to append assets from an online repository like the Blender Cloud, BlenderKit and similar. There would be a Python API to register custom asset repositories as part of add-ons. The main operations would be to provide a list of assets and metadata, provide .blend with an asset to append/link into the scene, and functions to add/remove/update an asset. **Milestone ?: Usability** Once the asset manager is working for the basics and the UI can do its job, we can start to make things nice. This includes better handling of thumbnails, re-configuring the ID browser to allow for asset browsing, a nice UI widget for managing tags and tag-based filtering. A key part of making the UX nice, is to make it effortless to drag assets into the scene. When dragging in objects & meshes, they should be able to snap to surfaces. When dragging in materials, the underlying target objects should highlight on rollover. You should be able to add multiple assets at a time. **Milestone ?: Collection Variations** Not strictly part of the asset manager since this would be a core feature of Blender also usable without the asset manager. We would like to support variations for all datablock types, but the first step could be collections (it's effectively like automatically toggling visibility on subcollections to show only one). Once this is in the place, the asset manager will need to be modified to understand variations. **Milestone ?: Project Repositories** Up to this point the asset manager will handle only the use case of appending files from a local or online repository. The next step would be making it works also for projects. * The root project directory would contain a `.blender_project` file indicating that it is a project (similar to e.g. `.git`) * When opening a .blend file within the project folder, Blender will automatically know which project it is part of * Folder structure, file names and file saving are left to the user. The asset browser will not add or remove .blend files in the project. That's all done using File > Save and file browsers as usual. * The assert browser will have a "Project" repository that can be chosen, which will show all assets within the project * The user will have to mark relevant datablocks as assets in each .blend file in the project so it shows up in the asset browser. * Assets can be linked between files within the same project. This is different from other repositories, where only append is possible. Linking between projects or to other repositories should not be possible, a project must be self-contained and not link to any .blend file outside of it. * The project folder would be indexed exactly like other asset folders. The idea here is that we don't want to make many assumptions about how users might structure their project. It should be possible to simply add or remove files, reorganize project directories, and the asset browser should keep working. It will merely provide a view on the existing project structure, and functions to create links to other assets in the project. It will not introduce an alternative linking system as an earlier iteration of the asset manager design did. This system does not preclude a studio from making an add-on that takes more control over file saving, version control, partial repository checkouts and so on. Such an add-on would likely need to integrate in various places in the Blender user interface, one of which can be the asset browser. For example it might include a custom asset repository to browse assets that are not on the artist's computer, and right before the user links an asset into the scene, download the corresponding .blend file. But for most projects that are not e.g. feature films, the built-in project repository and automatic indexing system should already be quite powerful. --- **Further links**: * [Asset Manager]] design blogpost*-- March 2020** [[ https:*code.blender.org/2021/06/asset-browser-project-update/ | Asset Browser Project Update ](https:*code.blender.org/2020/03/asset-manager/) *-- June 2021* * [Asset Browser Workshop Outcomes ](https:*code.blender.org/2021/06/asset-browser-workshop-outcomes/)*-- June 2021// * #54642 (Asset Project: User Interface)
Author
Owner

Added subscribers: @WilliamReynish, @JulianEisel, @Ton, @dfelinto

Added subscribers: @WilliamReynish, @JulianEisel, @Ton, @dfelinto

Added subscriber: @tibicen

Added subscriber: @tibicen

Amazing! :)
I would strongly suggest few thought:

  • have as a rule that library is divided over elements in each file. I've done Simple Asset Manager in os directories, cause the basic ".blend file as library" was not working in the "version control times" and updating my library online was hard with one big file as a library.
  • Possiblility of os directory structure, with own categories etc.
  • Have posibility to use library as a multitool importer. (it would recognize active import options/addons), and could handle other formats (images, obj, fbx, etc). As I tried to do that in SAM: "one button for all"
  • FileBrowser menu should go minimalistic with almost all the ui hidden. The less buttons, the less clutter, and more elements to drag to scene. (preset setting somwhere in hidden sidebar).

Don't know if it right place for such comments, but its somwhere in between ui and implementation.


Amazing! :) I would strongly suggest few thought: - have as a rule that library is divided over elements in each file. I've done Simple Asset Manager in os directories, cause the basic ".blend file as library" was not working in the "version control times" and updating my library online was hard with one big file as a library. - Possiblility of os directory structure, with own categories etc. - Have posibility to use library as a multitool importer. (it would recognize active import options/addons), and could handle other formats (images, obj, fbx, etc). As I tried to do that in SAM: "one button for all" - FileBrowser menu should go minimalistic with almost all the ui hidden. The less buttons, the less clutter, and more elements to drag to scene. (preset setting somwhere in hidden sidebar). Don't know if it right place for such comments, but its somwhere in between ui and implementation. ``` ```

Added subscriber: @Scaredyfish

Added subscriber: @Scaredyfish

Added subscriber: @AditiaA.Pratama

Added subscriber: @AditiaA.Pratama

Added subscriber: @einsteinchen

Added subscriber: @einsteinchen

Added subscriber: @Pipeliner

Added subscriber: @Pipeliner

Added subscriber: @Josephbburg

Added subscriber: @Josephbburg
Member

Added subscriber: @JacquesLucke

Added subscriber: @JacquesLucke

Added subscriber: @Cedch

Added subscriber: @Cedch

Added subscriber: @dee909

Added subscriber: @dee909

Changed status from 'Needs Triage' to: 'Confirmed'

Changed status from 'Needs Triage' to: 'Confirmed'

Added subscriber: @AndrewPeel

Added subscriber: @AndrewPeel

Added subscriber: @Constantina32

Added subscriber: @Constantina32

Added subscriber: @robblau

Added subscriber: @robblau

Added subscriber: @slumber

Added subscriber: @slumber

Added subscriber: @DuarteRamos

Added subscriber: @DuarteRamos

If feasible I'd request that whenever possible relative paths be used in library definitions, both in "indexes", user defined folders or paths to assets.

This helps make assets more resistant to change, so that paths don't break or require re-indexing if for example a drive letters change, or when accessing from network or different computers, assets in a removable drive, etc.

Another possibility is to store both (relative and absolute in parallel) like Inkscape currently does for external images, if one breaks the other may be used as backup.

If feasible I'd request that whenever possible **relative paths** be used in library definitions, both in "indexes", user defined folders or paths to assets. This helps make assets more resistant to change, so that paths don't break or require re-indexing if for example a drive letters change, or when accessing from network or different computers, assets in a removable drive, etc. Another possibility is to store both (relative and absolute in parallel) like Inkscape currently does for external images, if one breaks the other may be used as backup.

Added subscriber: @lemenicier_julien

Added subscriber: @lemenicier_julien

Added subscriber: @MaciejMorgas

Added subscriber: @MaciejMorgas
Member

Added subscriber: @MikeErwin

Added subscriber: @MikeErwin
Member

In #73366#879066, @DuarteRamos wrote:
If feasible I'd request that whenever possible relative paths be used in library definitions, both in "indexes", user defined folders or paths to assets.

+1 to relative paths for local files, with full path as backup.

> In #73366#879066, @DuarteRamos wrote: > If feasible I'd request that whenever possible **relative paths** be used in library definitions, both in "indexes", user defined folders or paths to assets. +1 to relative paths for local files, with full path as backup.

Added subscriber: @MichaelHermann

Added subscriber: @MichaelHermann

Added subscriber: @ChristopherArcher

Added subscriber: @ChristopherArcher

Added subscriber: @xan2622

Added subscriber: @xan2622

Added subscriber: @ckohl_art

Added subscriber: @ckohl_art

I realize the design is not finalized but can we get some clarification on:

Individual .blend file per asset in local folders (support loading multiple in a .blend too)

You mean to say that you would store every asset in .blend format instead of its native format? (.png, .tga, .wav, .mp3, etc.)

Or does this subsequent point mean that for, say, a folder full of brush alphas there would be a structure like bark.blend, bark.tga, scales.blend, scales.tif, star.blend, star.png, and so on?

External files can use pack, append, then unpack (into e.g. textures/ folder next to .blend)

I realize the design is not finalized but can we get some clarification on: > Individual .blend file per asset in local folders (support loading multiple in a .blend too) You mean to say that you would store every asset in .blend format instead of its native format? (.png, .tga, .wav, .mp3, etc.) Or does this subsequent point mean that for, say, a folder full of brush alphas there would be a structure like bark.blend, bark.tga, scales.blend, scales.tif, star.blend, star.png, and so on? > External files can use pack, append, then unpack (into e.g. textures/ folder next to .blend)
Member

Added subscriber: @jendrzych

Added subscriber: @jendrzych

Added subscriber: @brecht

Added subscriber: @brecht

@ckohl_art, I've updated the description to be more specific, hopefully that clears it up.

@ckohl_art, I've updated the description to be more specific, hopefully that clears it up.

Andrew Peel has published an interesting video on Youtube: Asset Libraries in Blender 2 83

Andrew Peel has published an interesting video on Youtube: [Asset Libraries in Blender 2 83](https://youtu.be/ZphqykbmHeQ)

Added subscriber: @lbtk

Added subscriber: @lbtk

Added subscriber: @shanberg

Added subscriber: @shanberg

Added subscriber: @kvick

Added subscriber: @kvick

Added subscriber: @item412

Added subscriber: @item412

Added subscriber: @johny.zlo

Added subscriber: @johny.zlo

Added subscriber: @karmaral

Added subscriber: @karmaral

Added subscriber: @RomboutVersluijs

Added subscriber: @RomboutVersluijs

Why is that Technical Doc link closed for the reason of "Invalid"?

Why is that Technical Doc link closed for the reason of "Invalid"?

Added subscriber: @kpavicic

Added subscriber: @kpavicic

In #73366#885607, @xan2622 wrote:
Andrew Peel has published an interesting video on Youtube: Asset Libraries in Blender 2 83

Some parts indeed, but it seems he's more interested in building blocks. Not so much about the asset manager itself. Though his functionality he showed of the manager itself look quite solid. All those other functions in his video already exist in other addons which already ship with Blender. It feels like re-inventing the wheel if im honest

> In #73366#885607, @xan2622 wrote: > Andrew Peel has published an interesting video on Youtube: [Asset Libraries in Blender 2 83](https://youtu.be/ZphqykbmHeQ) Some parts indeed, but it seems he's more interested in building blocks. Not so much about the asset manager itself. Though his functionality he showed of the manager itself look quite solid. All those other functions in his video already exist in other addons which already ship with Blender. It feels like re-inventing the wheel if im honest

Added subscriber: @tonton

Added subscriber: @tonton

Added subscriber: @SerjMaiorov

Added subscriber: @SerjMaiorov

Added subscriber: @derekbarker

Added subscriber: @derekbarker

In #73366#885607, @xan2622 wrote:
Andrew Peel has published an interesting video on Youtube: Asset Libraries in Blender 2 83

That is amazing, I hope the devs take many if not all suggestions from this video

> In #73366#885607, @xan2622 wrote: > Andrew Peel has published an interesting video on Youtube: [Asset Libraries in Blender 2 83](https://youtu.be/ZphqykbmHeQ) That is amazing, I hope the devs take many if not all suggestions from this video

Like i said, check Blender now and there are addons which have more advanced implementation of what he is showing. THere are already 3 addons with building elements like this ; Archimesh, Archipack, JARCH viz. Check those out

Like i said, check Blender now and there are addons which have more advanced implementation of what he is showing. THere are already 3 addons with building elements like this ; Archimesh, Archipack, JARCH viz. Check those out

Added subscriber: @Fureloka-4

Added subscriber: @Fureloka-4

Added subscriber: @LeoSch

Added subscriber: @LeoSch

Added subscriber: @Draxley

Added subscriber: @Draxley

Added subscriber: @ZoltanFodor

Added subscriber: @ZoltanFodor

Added subscriber: @AonoZan

Added subscriber: @AonoZan

Added subscriber: @aviggiano

Added subscriber: @aviggiano

Added subscriber: @OliverVillar

Added subscriber: @OliverVillar

Added subscriber: @Nerezza

Added subscriber: @Nerezza

Added subscriber: @Josh.Guenther

Added subscriber: @Josh.Guenther

Added subscriber: @nfguler

Added subscriber: @nfguler

Added subscriber: @richeyrose

Added subscriber: @richeyrose

I've been thinking about assets lately, and I was wondering if they shouldn't actually be implemented entirely as it's own separate type of datablock.

An asset will actually be a datablock appendable/linkable/browsable as a folder like all others, which among other things would contain the following data/fields:

  • Name - Name of the datablock
  • Description - A short note or usage instructions
  • Author - Name of the authors
  • Tags - For categorization and search/filtering purposes
  • Datablock - The actual data it will be containing, (a pointer, it is technically called?) to another datablock, like a collection to instance (containing a character), a material, a single object, or a shader node, or eventually an object with a geometry node setup (when everything nodes comes)
  • Thumbnail - Either an automatically generated thumbnail, or if overriden an optional custom image (or if possible sequence) of the datablock for preview purposes
  • Parent Asset - Maybe Variations could just be regular assets that would have a pointer(?) field pointing to another asset datablock they were a variation of, that way we avoid conflicts and can have an arbitrary number of them without convoluted setup processes

The file browser would then have this dedicated mode to browse Asset datablocks.

I've been thinking about assets lately, and I was wondering if they shouldn't actually be implemented entirely as it's own separate type of datablock. An asset will actually be a datablock appendable/linkable/browsable as a folder like all others, which among other things would contain the following data/fields: - **Name** - Name of the datablock - **Description** - A short note or usage instructions - **Author** - Name of the authors - **Tags** - For categorization and search/filtering purposes - **Datablock** - The actual data it will be containing, (a pointer, it is technically called?) to another datablock, like a collection to instance (containing a character), a material, a single object, or a shader node, or eventually an object with a geometry node setup (when everything nodes comes) - **Thumbnail** - Either an automatically generated thumbnail, or if overriden an optional custom image (or if possible sequence) of the datablock for preview purposes - **Parent Asset** - Maybe *Variations* could just be regular assets that would have a pointer(?) field pointing to another asset datablock they were a variation of, that way we avoid conflicts and can have an arbitrary number of them without convoluted setup processes The file browser would then have this dedicated mode to browse Asset datablocks.

Added subscriber: @DirSurya

Added subscriber: @DirSurya

Added subscriber: @GaryRitchie

Added subscriber: @GaryRitchie

Added subscriber: @KtRN

Added subscriber: @KtRN

Added subscriber: @RainerTrummer

Added subscriber: @RainerTrummer

Added subscriber: @RobertFletcher

Added subscriber: @RobertFletcher

At the moment this proposal seems to focus purely on native blender data (makes sense), It would be fantastic if this system was implemented using the Model View Controller pattern so that devs could add their own repositories via custom Python Models. and custom loading behaviour via Python Controllers.

So the View, simply renders items from the model, and perhaps allows a editor widget which facilitates custom search behaviours. When the view triggers a load event the controller takes over, interprets the item and does some work on it. (usually loading data)

This would allow studios to provide access to their existing asset management systems, or even integration with online market places via python addons.

It would also be great if the view was provided a starting context based on how it is opened. for example

  • image
  • model

so it could filter its data accordingly.

Something to consider :)

At the moment this proposal seems to focus purely on native blender data (makes sense), It would be fantastic if this system was implemented using the Model View Controller pattern so that devs could add their own repositories via custom Python Models. and custom loading behaviour via Python Controllers. So the View, simply renders items from the model, and perhaps allows a editor widget which facilitates custom search behaviours. When the view triggers a load event the controller takes over, interprets the item and does some work on it. (usually loading data) This would allow studios to provide access to their existing asset management systems, or even integration with online market places via python addons. It would also be great if the view was provided a starting `context` based on how it is opened. for example - image - model so it could filter its data accordingly. Something to consider :)

Added subscriber: @Beryesa

Added subscriber: @Beryesa

Added subscriber: @mark.b.tomlinson

Added subscriber: @mark.b.tomlinson

Added subscriber: @ReinhardK

Added subscriber: @ReinhardK

Added subscriber: @jpmonroyal

Added subscriber: @jpmonroyal

Added subscriber: @Benjamin-GO

Added subscriber: @Benjamin-GO

Added subscriber: @hlorus

Added subscriber: @hlorus

Added subscriber: @JulianPerez

Added subscriber: @JulianPerez

Added subscriber: @gusdleon

Added subscriber: @gusdleon

What would be really useful to me as someone currently having to knock together an asset manager as part of an addon to use until this is done:

  1. a firm decision about what meta data will be stored per asset.
  2. knowing how this will be stored (presumably on each data block)
  3. a way of pulling just this metadata in to the current .blend file without linking in the actual asset. Currently as far as I'm aware the only thing we can pull in is the asset name.

How assets themselves are stored (1 per .blend, multiple assets per .blend, file structure etc.), UI etc. is all stuff that is secondary. But the above three things are really crucial and mean that if you want to do an asset manager currently you either need to store asset metadata separately as i.e. a JSON file and keep it in sync with your assets, or link all your assets in to the file which is impractical. I really hope this actually goes somewhere this time and doesn't get booted into the long grass yet again.

What would be really useful to me as someone currently having to knock together an asset manager as part of an addon to use until this is done: 1) a firm decision about what meta data will be stored per asset. 2) knowing how this will be stored (presumably on each data block) 3) a way of pulling just this metadata in to the current .blend file without linking in the actual asset. Currently as far as I'm aware the only thing we can pull in is the asset name. How assets themselves are stored (1 per .blend, multiple assets per .blend, file structure etc.), UI etc. is all stuff that is secondary. But the above three things are really crucial and mean that if you want to do an asset manager currently you either need to store asset metadata separately as i.e. a JSON file and keep it in sync with your assets, or link all your assets in to the file which is impractical. I really hope this actually goes somewhere this time and doesn't get booted into the long grass yet again.
Member

In #73366#971240, @richeyrose wrote:
What would be really useful to me as someone currently having to knock together an asset manager as part of an addon to use until this is done:

  1. a firm decision about what meta data will be stored per asset.

The meta-data will be configurable. Current plan is to make them regular custom properties.
Different asset-engines have different needs (e.g. local asset manager vs. Blender Cloud one vs feature film studio X one). They should have full control over the meta-data.
There are probably some defaults that we can have builtin support for (preview image, description, tags, author, etc).

  1. knowing how this will be stored (presumably on each data block)

I might actually add a new "asset" data-block type that can reference the data-block, I see some technical benefits for this. But as long as you can get the list of assets, their meta-data and the corresponding data-block reliably and efficiently, the exact way this is implemented is an implementation detail in my mind.

The actual assets will be stored in an external data-base, which may just be a folder with .blend files and a caching system (e.g. JSON file). This is outlined in above's description and in this blog post .
Other asset-engines may want a different data-base design, e.g. we know some feature film studios using SQL data-bases. This should eventually be supported.

  1. a way of pulling just this metadata in to the current .blend file without linking in the actual asset. Currently as far as I'm aware the only thing we can pull in is the asset name.

Like said above, plan is to have some way to cache meta-data for the asset data-base. The Asset Manager will work with that mostly.

> In #73366#971240, @richeyrose wrote: > What would be really useful to me as someone currently having to knock together an asset manager as part of an addon to use until this is done: > > 1) a firm decision about what meta data will be stored per asset. The meta-data will be configurable. Current plan is to make them regular custom properties. Different asset-engines have different needs (e.g. local asset manager vs. Blender Cloud one vs feature film studio X one). They should have full control over the meta-data. There are probably some defaults that we can have builtin support for (preview image, description, tags, author, etc). > 2) knowing how this will be stored (presumably on each data block) I might actually add a new "asset" data-block type that can reference the data-block, I see some technical benefits for this. But as long as you can get the list of assets, their meta-data and the corresponding data-block reliably and efficiently, the exact way this is implemented is an implementation detail in my mind. The actual assets will be stored in an external data-base, which may just be a folder with .blend files and a caching system (e.g. JSON file). This is outlined in above's description and in [this blog post ](https://code.blender.org/2020/03/asset-manager/). Other asset-engines may want a different data-base design, e.g. we know some feature film studios using SQL data-bases. This should eventually be supported. > 3) a way of pulling just this metadata in to the current .blend file without linking in the actual asset. Currently as far as I'm aware the only thing we can pull in is the asset name. Like said above, plan is to have some way to cache meta-data for the asset data-base. The Asset Manager will work with that mostly.

Added subscriber: @Pirazza

Added subscriber: @Pirazza

@JulianEisel
Thankyou. That gives me some confidence that my hacky placeholder will be easily replaceable by your proper asset manager when the time comes and I won't have programmed myself into a corner!

@JulianEisel Thankyou. That gives me some confidence that my hacky placeholder will be easily replaceable by your proper asset manager when the time comes and I won't have programmed myself into a corner!

Added subscriber: @AlexeyPerminov

Added subscriber: @AlexeyPerminov

Added subscriber: @PetterLundh

Added subscriber: @PetterLundh

Added subscriber: @Jorgensen-1

Added subscriber: @Jorgensen-1

Added subscriber: @mattli911

Added subscriber: @mattli911
Member

Just a note, I'm currently experimenting with some technical design approaches that need to be evaluated still. I'm also writing a document for these that I'd like to present and discuss soon.

Just a note, I'm currently experimenting with some technical design approaches that need to be evaluated still. I'm also writing a document for these that I'd like to present and discuss soon.

Added subscriber: @smramsay

Added subscriber: @smramsay
Julian Eisel self-assigned this 2020-07-10 11:36:56 +02:00
Member

Added subscriber: @filedescriptor

Added subscriber: @filedescriptor

Removed subscriber: @johny.zlo

Removed subscriber: @johny.zlo

Added subscriber: @Phigon

Added subscriber: @Phigon

Added subscriber: @jon_b

Added subscriber: @jon_b
Contributor

Added subscriber: @RedMser

Added subscriber: @RedMser

Added subscriber: @Zeirus

Added subscriber: @Zeirus

Added subscriber: @Wesley-Rossi

Added subscriber: @Wesley-Rossi
Member

Added subscriber: @lichtwerk

Added subscriber: @lichtwerk

Added subscriber: @arcdipesh

Added subscriber: @arcdipesh

Added subscriber: @neo_phi

Added subscriber: @neo_phi

Added subscriber: @filibis

Added subscriber: @filibis

Added subscriber: @SimplSam

Added subscriber: @SimplSam

Added subscriber: @mcolinp

Added subscriber: @mcolinp

Added subscriber: @johny.zlo

Added subscriber: @johny.zlo

Added subscriber: @swanee

Added subscriber: @swanee

Added subscriber: @bent

Added subscriber: @bent

Added subscriber: @MeshVoid

Added subscriber: @MeshVoid

Added subscriber: @Peine_Perdue

Added subscriber: @Peine_Perdue

Added subscriber: @WeezY

Added subscriber: @WeezY

Added subscriber: @Olliver

Added subscriber: @Olliver

Personally I would like to see this get merged into master as soon as its any what useable. The sooner we have anything that saves us from having to append blend files manually the better!

Gotta give credits on the design work done so far <3 Looks clean and powerful.

Personally I would like to see this get merged into master as soon as its any what useable. The sooner we have anything that saves us from having to append blend files manually the better! Gotta give credits on the design work done so far <3 Looks clean and powerful.
Member

Added subscriber: @Kdaf

Added subscriber: @Kdaf

Added subscriber: @Harvester

Added subscriber: @Harvester

Added subscriber: @Misaelcornell

Added subscriber: @Misaelcornell

Added subscriber: @Noto

Added subscriber: @Noto

Added subscriber: @dashie

Added subscriber: @dashie

Added subscriber: @penguinburger1

Added subscriber: @penguinburger1

Added subscriber: @PaoloCabaleiro

Added subscriber: @PaoloCabaleiro

Added subscriber: @CodeComedyTV

Added subscriber: @CodeComedyTV

Added subscriber: @randum

Added subscriber: @randum

Will asset manager support custom node tree data blocks so such add-ons as animation nodes or sverchok could use to save some node trees?

Will asset manager support `custom node tree` data blocks so such add-ons as `animation nodes` or `sverchok` could use to save some node trees?

"Once the asset manager is working for the basics and the UI can do its job, we can start to make things nice. This includes better handling of thumbnails" ...

Could the following Python library be used to speed up thumbnails loading?

https://github.com/3dninjas/3dn-bip
https://blenderartists.org/t/blazingly-fast-preview-loads-images-of-arbitrary-size-bpy-utils-previews-drop-in-replacement/1303349

> "Once the asset manager is working for the basics and the UI can do its job, we can start to make things nice. **This includes better handling of thumbnails**" ... Could the following Python library be used to speed up thumbnails loading? https://github.com/3dninjas/3dn-bip https://blenderartists.org/t/blazingly-fast-preview-loads-images-of-arbitrary-size-bpy-utils-previews-drop-in-replacement/1303349

Added subscriber: @spiraloid-3

Added subscriber: @spiraloid-3

Will the asset manager support skinned armatures? I've got an armature with a skinned mesh parented inside it. when I select the armature and mark as asset, I only get the skeleton without the skinned mesh.
when I select both the armature and the mesh and mark as assets, I get them separately without the skinning connection "because of renaming".

when I put them both in a collection and mark the collection as an asset I get them together but as a collection instance without an override so I can't pose the armature.

am I missing something?

Will the asset manager support skinned armatures? I've got an armature with a skinned mesh parented inside it. when I select the armature and mark as asset, I only get the skeleton without the skinned mesh. when I select both the armature and the mesh and mark as assets, I get them separately without the skinning connection "because of renaming". when I put them both in a collection and mark the collection as an asset I get them together but as a collection instance without an override so I can't pose the armature. am I missing something?

In #73366#1172878, @spiraloid-3 wrote:
Will the asset manager support skinned armatures?

There's no such thing in Blender.

I've got an armature with a skinned mesh parented inside it. when I select the armature and mark as asset, I only get the skeleton without the skinned mesh.
when I select both the armature and the mesh and mark as assets, I get them separately without the skinning connection "because of renaming".

No. You get them separately, because they're two separate objects, and the assets are created as such.

when I put them both in a collection and mark the collection as an asset I get them together but as a collection instance without an override so I can't pose the armature.

am I missing something?

Yes. Lookup Linked Rigs.
Asset Manager is an Interface/Management system. It does not functionally change anything.

> In #73366#1172878, @spiraloid-3 wrote: > Will the asset manager support skinned armatures? There's no such thing in Blender. > I've got an armature with a skinned mesh parented inside it. when I select the armature and mark as asset, I only get the skeleton without the skinned mesh. > when I select both the armature and the mesh and mark as assets, I get them separately without the skinning connection "because of renaming". No. You get them separately, because they're two separate objects, and the assets are created as such. > when I put them both in a collection and mark the collection as an asset I get them together but as a collection instance without an override so I can't pose the armature. > > am I missing something? Yes. Lookup Linked Rigs. Asset Manager is an Interface/Management system. It does not functionally change anything.

Thanks for the response. I like that it's not changing anything. I am wondering how to put characters in it so that they can be added to scenes without additional setup.

(The first thing I tried was to link a rig and create a library override for it and make an asset for it)

image.png

in case anyone ever needs my fancy bendy tube for testing.
linked_bendyTube.zip)

Thanks for the response. I like that it's not changing anything. I am wondering how to put characters in it so that they can be added to scenes without additional setup. (The first thing I tried was to link a rig and create a library override for it and make an asset for it) ![image.png](https://archive.blender.org/developer/F10162193/image.png) in case anyone ever needs my fancy bendy tube for testing. [linked_bendyTube.zip](https://archive.blender.org/developer/F10162196/linked_bendyTube.zip))

Removed subscriber: @ZoltanFodor

Removed subscriber: @ZoltanFodor

Removed subscriber: @bent

Removed subscriber: @bent

Removed subscriber: @aviggiano

Removed subscriber: @aviggiano
Member

FYI, project update here: Asset Browser Project Update .

FYI, project update here: [Asset Browser Project Update ](https://code.blender.org/2021/06/asset-browser-project-update/).

Any chance on seeing some basic support for geometry nodes modifiers & node groups any time soon?

Any chance on seeing some basic support for geometry nodes modifiers & node groups any time soon?

Added subscriber: @Deanna

Added subscriber: @Deanna

I am super excited about this new Asset Browser. I've started using it in the 3.0 Alpha and am really happy with what is already implemented. So way to go! 🎉 🎉 🎉

  • Is there a way to search "tags" assigned to different assets? Or what's the plan with that?
  • If there is anything I can do to help with testing or UI/UX design wise (I'm an experienced artist) to help let me know.
I am super excited about this new Asset Browser. I've started using it in the 3.0 Alpha and am really happy with what is already implemented. So way to go! 🎉 🎉 🎉 - Is there a way to search "tags" assigned to different assets? Or what's the plan with that? - If there is anything I can do to help with testing or UI/UX design wise (I'm an experienced artist) to help let me know.

Added subscriber: @RobWu

Added subscriber: @RobWu

Hi all,

Regarding the Project Repositories, please reconsider having a default base set of folders for a project.
It will make cleaner project setups, and it's not that hard adding extra folders to a project if needed.

Or, like with e.g. Maya or Houdini, have some kind of project manager available to change folder names, or add extra folders., before the project creation.
With a base set of folders, you can also 'direct' the app towards these folders for certain tasks, e.g. opening the 'textures' folder when you need an image for a material. Or by default rendering all images to a folder called e.g. 'renders'.
Same for caches, go into 'simulation' etc. etc.

It's not forcing users to use a certain workflow, but it makes a cleaner one when certain things are offered. It's one of my annoyances with Blender, after working with apps like Softimage, Maya, Houdini and others for years.
And I always found it odd that unpacking a blend file will create some of these folders, but by default there's no such system in place.

So again, please consider something like a default folder structure for projects.

cheers!

Hi all, Regarding the Project Repositories, please reconsider having a default base set of folders for a project. It will make cleaner project setups, and it's not that hard adding extra folders to a project if needed. Or, like with e.g. Maya or Houdini, have some kind of project manager available to change folder names, or add extra folders., before the project creation. With a base set of folders, you can also 'direct' the app towards these folders for certain tasks, e.g. opening the 'textures' folder when you need an image for a material. Or by default rendering all images to a folder called e.g. 'renders'. Same for caches, go into 'simulation' etc. etc. It's not forcing users to use a certain workflow, but it makes a cleaner one when certain things are offered. It's one of my annoyances with Blender, after working with apps like Softimage, Maya, Houdini and others for years. And I always found it odd that unpacking a blend file will create some of these folders, but by default there's no such system in place. So again, please consider something like a default folder structure for projects. cheers!

Along with folders, I would also like to see an asset tagging system that's independent from folders. So you can tag each asset with an unlimited amount of tags and then search entirely by these tags, for larger databases with a large variety in assets I believe management would be easier this way.

Along with folders, I would also like to see an asset tagging system that's independent from folders. So you can tag each asset with an unlimited amount of tags and then search entirely by these tags, for larger databases with a large variety in assets I believe management would be easier this way.

Added subscriber: @Eliot-Mack

Added subscriber: @Eliot-Mack
Author
Owner

Just to be clear, there are no folders per-se. But instead the catalog system that creates a virtual folder system to sort the assets in a hierarchical way. On top of that a tag system can be implemented for filtering too.

Just to be clear, there are no folders per-se. But instead the catalog system that creates a virtual folder system to sort the assets in a hierarchical way. On top of that a tag system can be implemented for filtering too.
Julian Eisel removed their assignment 2021-10-19 19:09:51 +02:00
Member

This is a group project now, no reason to keep me as lone assignee.

This is a group project now, no reason to keep me as lone assignee.
Member

Unless there are major show stoppers requiring a big revert, the Asset Browser will make its debut in the 3.0 release. There is still a lot of work to do and we'll soon do further planning for that. But at this point I think we can reduce the priority of this basic umbrella task to Normal. Better to have individual parts of the design high priority (e.g. collection support).

Unless there are major show stoppers requiring a big revert, the Asset Browser will make its debut in the 3.0 release. There is still a lot of work to do and we'll soon do further planning for that. But at this point I think we can reduce the priority of this basic umbrella task to *Normal*. Better to have individual parts of the design high priority (e.g. collection support).

Removed subscriber: @tibicen

Removed subscriber: @tibicen

Removed subscriber: @item412

Removed subscriber: @item412

Added subscriber: @satishgoda1

Added subscriber: @satishgoda1

Added subscriber: @Junyi-Zhang

Added subscriber: @Junyi-Zhang
Thomas Dinges added this to the 3.5 milestone 2023-02-07 18:55:56 +01:00
Brecht Van Lommel removed this from the 3.5 milestone 2023-02-09 19:33:02 +01:00
Philipp Oeser removed the
Interest
Pipeline, Assets & IO
label 2023-02-10 08:54:19 +01:00
Sign in to join this conversation.
No Label
Interest
Alembic
Interest
Animation & Rigging
Interest
Asset Browser
Interest
Asset Browser Project Overview
Interest
Audio
Interest
Automated Testing
Interest
Blender Asset Bundle
Interest
BlendFile
Interest
Collada
Interest
Compatibility
Interest
Compositing
Interest
Core
Interest
Cycles
Interest
Dependency Graph
Interest
Development Management
Interest
EEVEE
Interest
EEVEE & Viewport
Interest
Freestyle
Interest
Geometry Nodes
Interest
Grease Pencil
Interest
ID Management
Interest
Images & Movies
Interest
Import Export
Interest
Line Art
Interest
Masking
Interest
Metal
Interest
Modeling
Interest
Modifiers
Interest
Motion Tracking
Interest
Nodes & Physics
Interest
OpenGL
Interest
Overlay
Interest
Overrides
Interest
Performance
Interest
Physics
Interest
Pipeline, Assets & IO
Interest
Platforms, Builds & Tests
Interest
Python API
Interest
Render & Cycles
Interest
Render Pipeline
Interest
Sculpt, Paint & Texture
Interest
Text Editor
Interest
Translations
Interest
Triaging
Interest
Undo
Interest
USD
Interest
User Interface
Interest
UV Editing
Interest
VFX & Video
Interest
Video Sequencer
Interest
Virtual Reality
Interest
Vulkan
Interest
Wayland
Interest
Workbench
Legacy
Blender 2.8 Project
Legacy
Milestone 1: Basic, Local Asset Browser
Legacy
OpenGL Error
Meta
Good First Issue
Meta
Papercut
Meta
Retrospective
Meta
Security
Module
Animation & Rigging
Module
Core
Module
Development Management
Module
EEVEE & Viewport
Module
Grease Pencil
Module
Modeling
Module
Nodes & Physics
Module
Pipeline, Assets & IO
Module
Platforms, Builds & Tests
Module
Python API
Module
Render & Cycles
Module
Sculpt, Paint & Texture
Module
Triaging
Module
User Interface
Module
VFX & Video
Platform
FreeBSD
Platform
Linux
Platform
macOS
Platform
Windows
Priority
High
Priority
Low
Priority
Normal
Priority
Unbreak Now!
Status
Archived
Status
Confirmed
Status
Duplicate
Status
Needs Info from Developers
Status
Needs Information from User
Status
Needs Triage
Status
Resolved
Type
Bug
Type
Design
Type
Known Issue
Type
Patch
Type
Report
Type
To Do
No Milestone
No project
No Assignees
100 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: blender/blender#73366
No description provided.