Proposal: Tools, training, and professional help for large scale projects

[madjin]

Abstract

The current Builder tool provided by Decentraland enables inexperienced individuals to make small personal builds on private LAND using pre-built asset packs.

A more advanced tooling and team development platform is needed for creating the custom MVPs of the large, complex, and scripted builds envisioned by districts and larger estates.

It's hard to imagine building a city as large as Manhattan without power-tools. Districts represent over a third of the LAND in Genesis city and are excellent sandboxes for prototyping new business models and use-cases for crypto and VR gaming.

This proposal delineates a foundation of tools, processes, collaborative methodologies, and builder training that will boost land-owners' skills and facilitate both Decentraland and Genesis City development with a stronger feedback loop.

Problem

• Districts have been experiencing great difficulty in propping up their ideas.

• Some of the bigger districts are composed of hundreds of people. Collaboration is a critical feature needed for these larger scale projects.

• The current Builder is a great tool for simple scene creation, but is underpowered for complex, large scale team projects.

• No VR support makes it very difficult to develop experiences that will eventually be viewed in VR - scale and perspective will be severely affected.

Solution

For part of this program, a skilled VR squad works closely with district leaders on the two areas we need to focus on as mentioned in the Scene MVP Guidelines:

1. The basic user experience and functionality of your project.
2. The creation of a basic “pipeline”, or team workflow and content management system for building your experience and iteratively improving it.

While building MVPs with districts and estates, feedback is provided back into DCL and the tool builders in the form of detailed documentation, bug reports, and pull requests. Also, there will be virtual workshops designed to teach best practices for creating MVPs to exponentially grow the the builder community and evolve their skill-sets.

Tools

JanusVR offers a high quality building environment for Decentraland scenes, estates, and districts. Today there is a working pipeline, team workflow, and content management system for making MVP experiences and iteratively improving them using JML then converting to Decentraland's SDK, as seen in the following examples:

Janus can reduce friction between collaborators with a client that has built in chat, voip, and editor functionality for building and exporting DCL SDK compatible scenes in real-time. Multiple people can meetup at a website location, build using drag and drop, then export results to DCL.

From what we've seen in the Builder contest, people's creativity will rapidly flourish when given new tools.

Here is a 1 hour timelapse from an early 2018 demo of three people working together to construct a prototype of the Decentraland Conference Center.

full video: https://vimeo.com/320886148

This feature is particularly useful to drag and drop assets from any type of source to prototype their ideas, including but not limited to obj, dae, fbx, gltf, jpg, png, gif, mp3... files, from Decentraland curated content, Google Poly, and custom team libraries.

Janus projects can be saved or imported as HTML/JSON and have a high degree of interoperability with other tools and environments: https://github.com/madjin/janus-tools.

---

Training

The second part of this proposal is about establishing a best practices knowledge-base to be enjoyed by generations of new builders in the future. There are multiple components to how training and education will be delivered:

1. Detailed write-up inspired by DCL documentation for Scene MVP Guidelines
2. Producing short tutorials and guides to Github
3. Livestreams and other forms of video production

The Decentraland University is already gearing up to have a VR Academy within its campus which can be an in-world venue for this educational initiative.

Many of the other districts have a vision for what they want to build. We can put the Scene MVP Guidelines to practice and coordinate and collaborate with with their team and an expert VR developer in real-time.

Within a few weeks, the volunteers that have been building under the wing of an expert will become advanced themselves and can teach others. Students can watch and talk to experienced builders:

High Speed Scene Building

Based on the example above and other tests done with collaborative DCL building it takes an average of 5 minutes to make satisfactory 16x16 parcels. With a couple of squads of 3 people using this software (2 district contributors per 1 VR professional) it would take approximately this long to populate a district with content:

• Festival Land: 473 (19 man hours)
• DCL Conference Center: 799 (33 man hours)
• DCL University: 1550 (64 man hours)
• District X: 2001 (83 man hours)
• Vegas City: 6776 (282 man hours)

To get an accurate estimate of how long it would take to develop a MVP for a major district or estate several other factors would have be taken into account such as model creation and timezones. It takes roughly an hour for a skilled 3D modeler to make a low poly object from scratch.

There is also a full loop pipeline between Blender and Janus, import and export, with IPFS support:

Resources

• https://github.com/decentraland/districts
• https://dclplazas.com/decentraland-districts
• https://vimeo.com/320886148 - collaborative parcel building
• https://vimeo.com/310668134 - collaborative district building
• https://blog.district0x.io/decentraland-districts-40b9ada0431b
• https://decentraland.org/blog/announcements/district-update
• https://medium.com/decentraland/announcing-mana-bounties-for-3d-designers-on-ethlance-73d2bb559bad

Summary

This proposal is about forming an open source, community-centric, and educational approach to accelerate the development of Decentraland and a best practices knowledgebase for future builders.

By enabling the rapid creation and deployment of customized and dynamically scriptable content at large scale, Decentraland will be well populated and richly interactive at go-live.

The JanusVR team has demonstrated immediately useful tools for this methodology, with prototypes that enable a faster and more collaborative team workflow.

As an added benefit of development using VR-capable tools, the scene designs will be proven for both 2D and future VR user experiences.

We seek a modest investment to quickly round out these capabilities and to create the educational process for new builders.

Many prominent districts and large estate owners are interested in test piloting this program and can provide LAND and volunteers to beta test outlined solutions. These districts are marked orange on this map:

[Alonzo-Coeus]

We are planning on using this for some aetheria scenes so, it would be useful if it became a popular community pipeline

[b4u5]

Powerful collaborative tools like this would be of great help to all my projects in DCL.

[cfravel]

I agree that this more advanced tooling, pipeline, professional assistance and training would aid both district development as well as the building of larger or more complex private estates. It would be a good ROI for support from Decentraland and its backers

[storm5had0w]

In Jin we trust

[PDeee]

To allow me to accelerate developing my land content with a sort of ease of user interface of the DCL builder and underlying power of the SDK, I would love to see this as another supported land development option and am prepared to spend some time learning to use it. I can only imagine that the larger the project the more the collaborative tools would come into their own.

[JamesMcCrae]

Well proposed. Such a project would greatly accelerate development.

[RobL4045]

Great ideas Jin! This would be extremely helpful for all Decentraland districts. District X would definitely participate.

[toosad]

This is a great project! 100%

[voytaa]

As a contributor in 3 districts, I think the problem as described here is real so addressing it, by the JanusVR or otherwise could benefit the community greatly

[FireFoxxG]

Very well thought out game-plan here. I've helped a couple top districts as a contractor and I think they really need the tools, training, and technical support this proposal offers. They would get a vastly improved rendering engine and a proper in world scene editor.

[ToxSam]

This tool is a MUST for the metaverse development. Right now building a scene is a slow process that has to be done by just 1 person at a time. This creates a bottleneck on the pipeline, slowing down the development of any virtual scene.
I've used the collaboration tool with Jin to build a couple of scenes, and o boy, it is fun and useful at the same time. It just feel right, using this you can spread the workload among more builders to speed up the production of any scene!

[madjin]

Thank you for all the support and encouragement, very appreciated. It's been at least a month since I've posted the proposal so I'd like to write an update.

Since the post, I've spoken directly to DCL about two weeks ago regarding this proposal . I'd like to highlight a few points here:

• Content is king, and these tools aim to greatly accelerate the content creation pipeline as well as make it fun!
• One of the byproducts for training and collaboration is the content that is made can be shared through videos and documentation. This helps drive adoption.
• These tools make it greatly easier and faster for content from other development environments to be brought into DCL.

Remember that this proposal is not only about technology, but about the offer to provide training and building services.

Every person on the JanusVR team has at least 5 years of professional experience in the VR industry, building the decentralized metaverse. Folks like me and @FireFoxxG whom are experienced using such tools have made content for LAND owners and districts in DCL and are currently 10x faster than industry standard for building high quality, optimized, custom content. Building this way with others is how some of us learned to become VR devs. Our experience being passed onto new builders would be an invaluable knowledge transfer to the DCL community. Did I mention it's fun also?

"Give a man a fish, and you feed him for a day; show him how to catch fish, and you feed him for a lifetime."

---

Use-case: 1 Designing MVPs

The Decentraland docs say:

---

Builders need to think about two areas of focus for designing MVPs:

1. The basic user experience and functionality of your project.
2. The creation of a basic “pipeline”, or team workflow and content management system for building your experience and iteratively improving it.

---

The methodology of this proposal supports this in two ways:

1. The basic user experience can be built in Janus collaboratively between team members who have the vision of the user experience and functionality and team members who have the skills and inventory to build the scene. This can be done iteratively, experimentally, until it is refined.

2. The tooling used here provides a number of options about how the team might form a "pipeline". Like Blender -> JanusVR -> SDK -> Land

As to adding user interactivity, which means scripting, It's more ideal to add interactivity into the scene after first having the basic static MVP in place.

Use Case 2: Nexus Marketplace build

During the auction around 4.5M USD worth of perceived value in MANA went towards districts and they've had very little support since then. As a result, projects have become stagnant with most districts failing to achieve proper MVPs of their visions even after having over a year to build.

To demonstrate the ability to more rapidly build out land content, we can take the case of Nexus marketplace district (279 parcels) as a case study.

Little had been done in the building of this district over the last year. Recently, a squad of 3 professional builders using Janus tools built the large scene shown below in about an hour.

Extrapolating that to the district, and polishing the work, a fair estimate to build the entire district is about one week with custom static content and art.

Faster speeds = more dynamic city.

Covering the entire area with static content/art to develop a sense of space.

Watch the full video here: https://vimeo.com/342432345

[b4u5]

If DCL funded the development of a tool like this along with building services for districts and larger scale private projects to use, it would exponentially speed up the development of and add considerable value to the Metaverse. Anyone who holds MANA and/or LAND should be very excited about this and should want to contribute to make this happen, as it will directly increase the value of their assets.

[Converg3nce]

Awesome proposal Jin! I'd love to see this get some funding from DCL because this is the type of thing that could really help speed up our district's development progress. Thank you for putting as much effort into Decentraland as you have.

[smsithlord]

I'm all for creating cool content for DCL while pioneering a smooth pipeline for other creative DCL users to utilize! Seeing people's joy as they learn how to realize their creative potential in an app that they love is a wonderful thing. Teaching a man how to fish type stuff.