[go: up one dir, main page]

Skip to content

Latest commit

 

History

History

2019-05-16

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 

M3 2019-05-16

Schedule

6pm PST Thursday May 16, 2019

Agenda:


Topic of discussion:


Weren't able to get M3 name, M3-org is the organization name settled on for now.

12:20 - Assigning different roles

15:40 - Chris Platorres: Past Metaverse Attempts

22:00 - QA: What is the base reality for the metaverse?

Chris Platorres: Past Metaverse Attempts

Coming from more cynical end, science fiction term but there is a more practical way to implement it. Don't want to get lost in the sci-fi vision of it.

The metaverse is "wikipedia definition"

Many have preconceptions of the metaverse from Snow Crash and Ready Player One

Cannot directly translate from science fiction to reality

What makes for such great eye-candy on the big screen doesn't always translate into the real world very well.

When it comes to the metaverse, we should take it down a notch and find out wheres the real value.

There's a framework for creating achievable goals: S.M.A.R.T. to help us keep grounded and on track.

Few premisis: projects reluctant to adopt new standards.

Past attempts: create something amazing "Hey adopt our world/standard/protocol"

Unless you achieve critical mass and people incentivized to do so, its a very hard sell.

-XKCD-

Web standards has been arguably the most successful attempt at having a standard platform. Immersive web is building ontop of it. Once people thought CD-ROMs were important than internet, wasn't true. Hyperlinks became more important, what is the hyperlink of the metaverse?

Instead of just 1 video game or client, there should be a hyperlink between these worlds in 3D.

If you look at the web there is no base reality. If you see sci-fi there's a base layer of the metaverse (The Street in Snow Crash, Oasis in RPO).

Hyperlinked multiverse of experiences.

Looking at past attempts, we should generally be oriented on is being practical solving specific problems, focusing on compatibility, starting simple and allowing for emergence. Building blocks.


Base Reality for the Metaverse

Web has no base reality, but it seems that in recent home platforms trying to own that top layer.

Do you think there's room for some company to build interconnecting layer jump into? Maybe CV / HiFi / Supermedium?

How do you see base layer emerging?

Should exist but not forced to have only one: should be customizable home pages like a start page / new tab page. Offline startpage! Browsers can browse local files too.

P2P systems work in the way of local first. PWA also have property of being able working offline.


2D in 3D

Need to bring 2D browser in a 3D environment

Why abandon the 2D web? Can't pull up a browser in WebVR, missed opportunity.

What am I doing in browser anymore (self-crippling), might as well get a native app.

Plugins push browsers forward but plugins are getting increasingly restricted.

Browser standards push/pull, some parties want new features but bad actors cause browser vendors to restricting extensions.

Going native route you are not policed by app store model.

Exokit wants to drive use-cases, first to experiment with things with maybe other browser vendors are afraid to experiment with.

Loading native apps (like minecraft.exe) and displaying it in a web context.

It's on groups like us to prove out the case, prove that there's demand, and that there is a way to secure it for everybody.

Idea of an overlay (Pluto that overlays social over other games).

Something in between hubs and exokit reality tabs - hopping across divergent spaces but staying connected via some substrate.


31:00

What could be consistent between one layer to the next? (avatars?)

What aspects when jumping between incrongruent spaces stay same?

Maybe the metaverse is just a state of mind? The Matrix has you.

Given these walled gardens, hard to inject yourself above them in kosher way.

How do we reach app-store approved layer above them?


31:30

What is the shared context or standard people agree on between apps?

33:00

What carries over between different experiences?

Avaer: There isn't very much you can carry over, but there is a common ground reality we can program against (pixels and input - like hands/head). These are only the kind of constants we can rely on in XR context.

As to Chris's point, in many ways the Metaverse may have to reflect social structures of reality

when you go to certain place that place has rules / laws / aesthetics / cultural norms that one adheres to. Metaverse layers of reality will be equally as divergent as reality is. It will be on person visiting that reality to decide if they want to be there, agree to the norms, spend time with. Not sure if its a realistic goal to unify everything under a single common framework (like Chris talk on interconnected multiverses).

No grand platform to cover all cases, more likely connected multiverses.

Invite people to personal space


36:00

Agree not to put too much infrastructure in protocol

Whatever decisions had with html/http early on carries with us today

If you don't include certain functinos into protocols, other structures fill in gap and can form monopolies on the network

Would of been great if web had payments included in it

protocol people can implement

absense of app store, a giant then dominates

marketplace of app stores

if you don't provide it, aws will provide it or someone else will

in vacuum potentially not great outcomes will come in and lock things

whens last time cool OS for mobile phones came out and dominated?

in China with vacuum of Google more froth for things built on top of AOSP

If you want to create environment for less restrictive, need to think marketplace of app stores

Leverage existing technology instead of reimplementing?


39:55 Avaer: Permissionless Innovation for VR

Exokit does a lot of things browsers can't do

Can load 2D web browsers in front of you and layers of reality

Whole idea is around permissionless innovation

Free to experiment, build, invent things without somebody in one way or another saying "no you can't do that"

Web is best place to do those kinds of experiments - can take any 2D idea you have to throw it on a screen and share with rest of the world.

VR is just 2D screen x2

Quickly ran into corner cases where permissionless innovation breaks down in web platform such as:

CORS extensions loading 2D pages

First thing you want to do is load 2D page usually

Reason you can't do that in browser is because there's so much legacy and "security-issues"

Vendors are scared to ship features that allows people to show 2 in 3D context

Perhaps ways to sniff out sensitive information via eye/hand tracking

XR is people experimenting with ways of interacting with computers, similar to mobile days

Appstore model - things are approved with some amount of safety assurance with programs you're installing.

Going through same thing with web, figure out what security issues are, fix them, train users what it means to be in certain reality vs another, slowly as generation shifts can solve many problems.

Need to be experimenting right now (How to blend 2 layers of reality).

Exokit experimenting with loading web content and exe, what does that mean for security?

If over here this is my bank and only I can see it and everybody elses reality is separate from that, what does that mean and look like?

Hope to test out some of these ideas with Exokit and maybe eventually one or two of these things can be a standard by showing people this can work, can be secure, and will be awesome.


46:00 QA: Thoughts on IPFS/Dat for permissionless Innovation?

In order to share projects on 2D web currently requires going through central authorities also need a web server, domain, etc.

New protocols have early internet problems with new standards, education for end users and developers.

Google not on board fully with Dat.

Huge amount of intertia with existing protocols and way things are done now.

Problems we're solving need to show people how much better it is, needs to be better by 10x by some use-case for some major amount of adoption to occur.

Tide might be shifting with people getting upset with data breaches, wisening up with how their data is being handled by faceless corps.

49:00 QA: Z-buffers and security with Exokit

Way that Exokit currently works is that since fundamental building block is frame buffer and everything is composited in terms of Z there is issue of security: what does it mean for application to own a certain set of pixels?

In terms of space a lot of that is solved because you can partition the world into octrees and cubes, over here this space can only be rendered by this application and the user-agent / top level OS can make decision. Can play into inputs - can make it such that this cube over here as long as you are not inside that cube cannot read inputs.

Most of all, need to decide what all this means in context of a 3D bank. (BE YOUR OWN BANK?)

On all of us to experiment with what problems are there are - technical details AND user education. How do you teach someone that over here is secure space. In 2D space we solved this with padlock icons. What does that look like in a 3D space? Whom do I trust? Side channel or outline on ground or watch that cannot be controlled by sites?

Fundamentally trust has to start somewhere - headset hardware trust pixels are the ones you want to display - layer of reality doing what you want it to do - need to give users awareness where they are. Technological and educational issue, teach people new spatial computing way of navigating XR space securely.

Still unsolved in 2D web, browser vendors are moving slowly because of these issues. Cryptocurrency opportunity to build from ground up?

App can only see its own layer, if you're inside application space it has control over those pixels. A lot of research needs to be done here.


54:10 Chris Novello:

Computer artist that teaches at art school

Teach by navigating the metaverse, how to spatialize links and web assets? Memory palace - how to spatialize knowledge?

2D web not going away for awhile, main places of thinking, need to bring it into 3D How to browse the 2D web in 3D?

Cards?

Watch with cards, pull out certain aspects?

Trying to advocate for a browser in a browser. Webview? Janus?

Bundled with immersive page jump?

Teach code to enact a new type of renaissance, destroy the "user" - flip into a "creator"

We're in a computation environment, why are we not computing? Why are we not making computation?

Incubating culture around reimagining what programming is.

Class last semester students made dog park - program by training dog, can combine tricks.. turtle-esque.

Approaching from many angles, need a way to create software from VR.

Start thinking of sane way for collaborative software development in VR.

Don't want to end up with just creating 3D spaces.

How can we safely run scripts/sandbox code in shared spatial environment?

Discussion on how we can think about memory palace teacher practice when not having a 2D browser in 3D yet.

Hubs is great, can browse the web and improv perform. Cannot save things in time though, output is a video usually.. (ghost recording?)

1:03:00 Starting an ephemeral university / fablab / makerspace - burn it down

Things we think we want we don't want

In future there is no such thing as a user - everyone will have ability to create things as easily as consuming media.

We're not just going to be decorating spaces, we'll be thinking and computing collaboratively.

Metaverse is a burn it down process

Computing will not slow down and keep changing.

Ycombinator is an institution that responds with how fast technology is moving.

State of mind / process / way of being

Browsing the internet is education, eventually we'll be doing it spatially

Want it archived, not like what unity is doing.

Oldweb.today

Want way experiences can be archived, unity emulation in future is finnicky.

Why does web team want unity to win??

Its still damn hard to make web pages, most people will never learn (Myspace was an exception until Facebook killed it).

Important not to be locked into particular flavor of openness - have some amount of flexibility.

People freaking out about Unity IPO because it makes it sound more like a company - it's always been a company.

Eliminate boundary between creator/consumer.. instagram / squarespace great but it locks people into their intertia.

1:10:00 First off: How do we browse the web in 3D?

Really like what hubs is doing

Are we to pursue 2D browser in 3D? How do we organize, what buttons to push? (Janus is a good example)

Are we spinning up a whole new browser?

1:11:00 - Exokits approach

Exokit loads 2D page by spinning up fresh copy of chrome in a sandbox

Web is in a very weird place right now where it started off as open thing but now that its centralizing in Google's hands they're turning it into their version of Unity.

On one hand with efforts like AMP and portal tag its somewhat unheard of before: Google weild lot of power that they're making version of unity with existing web standards and costs externalized.

Exokit handles it by throwing away a lot of legacy, abuses a lot of existing software principles to do 3D parts of the web and move forward in that direction without security issues of 2D pages and things browser vendors are arguing.

Isolating different sites in different contexts and rethinking browser model and what security and information interchanging means.

1:13:20 - QA: If Exokit has this approach how can that be pointed to be upstreamed / palletable

Can someone WASM compile servo as polyfill? Is there someway can use exokit to use CEF? Use web browser as it exists in 3D space again.

How to bring rectangle out into 3D world? What would person really want to do? (Co-working)

Like suggestion of VMs in VR (Qubes in XR).

If minecraft can run inside its own VM and solve compositing problem that'd be awesome!

N-dimensional hypervisor

BigscreenVR really embraced existing 2D aspect to make that as a layer of VR

Limited version of DOM inside XR space making its way through spec

Exokit is not trying to invent things, predict where things will be in future and implement now in hope it will be standard


1:20:15 - Explaining the World Hopping Event

Implement S.M.A.R.T. framework to traveling between unrelated virtual worlds, find out the landscape of them and how they interconnect.


1:25:57 Shortfalls of previous attempts

When joining metaverse apps look at things they're doing right and where they fell short from doing cool things and how we can fix that.

Literally VRchat and minecraft next to each-other, how to walk between them?

Wouldn't developers have to target half the available specs if two things were running at same time?


1:28:25 James: Border between Minecraft / VRChat

Back in the day did a pilgrammage between disparate applications

Best way to look like you're moving between worlds is from compositing tricks but if you want true interoperability the solution would have to go deeper

In theory Minecraft / VRchat have some kind of partnership there can be more opportunity for data portability

N ways FActorial exposure

With Janus excited about creating boundary free experiences

Use WebXR to glue to enable


1:30:36 Smalltalk

Real security nightmare - no real boundary between programs.

Look at ActivityPub


1:35:50 Stadia

Anything that was on previous platforms is gone now, no way to archive. Google would be arbiter of content moving forwards.

When netflix isn't carrying movie the users are left out, no ownership of movie like used to with physical media.

Shift to software streaming is taking away ownership.


1:40:10 Chris Novello - Connecting computer programs

Example with using tetris to control ableton live

1:42:14 Alan Kay

1:42:46 Learning about Smalltalk

Two apps expose everything they can do

Talk from one program to the other

Basic notion of object oriented programming (objects talking to each other through messages)