Google stadia game changer for opensim?

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

Google stadia game changer for opensim?

Dr Ramesh Ramloll
Hello,

At the last OSCC, there was a lot of positivity around a new modern viewer
using Unreal Engine.  Now with Google Stadia, it appears  that it will be
possible for opensim builders to access the same audience as other
platforms on any device. I know  'streaming' is pretty old now, but for
Google to go in this way suggests that they might have made it affordable.
Do we know what's happening with the Unreal Engine viewer effort?

Best regards,
Ramesh

--
'Consider how the lilies grow. They do not labor or spin.'
*Rameshsharma Ramloll* PhD, CEO CTO DeepSemaphore LLC, Landisville, PA;
Affiliate *Research Associate Professor*, Idaho State University,
Pocatello, ID 83209 Tel: 208-240-0040
LinkedIn <http://www.linkedin.com/in/rameshramloll>
_______________________________________________
Opensim-users mailing list
[hidden email]
http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: Google stadia game changer for opensim?

Haravikk


> On 20 Mar 2019, at 12:39, Dr Ramesh Ramloll <[hidden email]> wrote:
>
> Hello,
>
> At the last OSCC, there was a lot of positivity around a new modern viewer
> using Unreal Engine.  Now with Google Stadia, it appears  that it will be
> possible for opensim builders to access the same audience as other
> platforms on any device. I know  'streaming' is pretty old now, but for
> Google to go in this way suggests that they might have made it affordable.
> Do we know what's happening with the Unreal Engine viewer effort?
>
> Best regards,
> Ramesh

I keep looking at streaming offerings but every time I do I'm not convinced; while there are some games where it's "fine" the quality just never lives up to claims and for twitchy games it's a non-starter.

While opensim is arguably not a twitchy game where response times are key, you're still talking about adding another layer of lag between the user and the simulator which feels like a recipe of extra-strength headaches to me.

There's also something to be said for privacy; a true virtual world is so much more personal than games like Call of Duty, Fortnite etc., you're not just dressing up the same basic player(s) that everyone else has, you're creating an avatar that *is* you, or an extension of you, in a virtual world. Introducing another party into that pipeline (other than grid, sim and you) feels weird to me.

So yeah, I dunno; if streaming can be used to help people with less powerful hardware to get better visuals then sure, streaming could be of benefit to them, but for anyone with a half-decent machine streaming is to be avoided IMO. I'm also not sure how useful mobile support via streaming would be for a platform like OpenSim, in terms of controls there's a lot to try and boil down into a mobile device; if you want to support in-viewer build tools then that complexity is even greater.

I don't think streaming platforms are a "game changer" for virtual worlds, they're an option for increasing accessibility given how demanding viewers tend to be on hardware, but the streaming aspect doesn't fundamentally change the difficulties of presenting a complex, malleable virtual world to users, streaming is really just a change in delivery method. I also don't think there's anything specific to Google's offering that makes it fundamentally different to all the past services that have tried; technology to implement it has improved, but you're still adding input and output latency plus video compression no matter what you do.

It's worth remembering as well that aside from the initial rush of assets downloading in a new sim, the actual back and forth messaging is relatively light; video streaming on the other hand still requires a fair bit of bandwidth just to look good, as well as modern processors for full h.265 support in hardware (or a dedicated GPU in which case you could probably handle a local viewer), so streaming isn't necessarily cost-free in terms of performance and hardware requirements either.

Does anyone know if there's much demand for virtual worlds on the go? I don't know that it's something I'd fully want to do on my smartphone, though the ability to chat with people in-world would be useful; does OpenSim currently support a way to do this without having a zombie avatar sign into a region?


Sorry for the text-wall and if I come across as too negative; it's an interesting technical space, I'm just a bit skeptical how much it really affects something like OpenSim.
_______________________________________________
Opensim-users mailing list
[hidden email]
http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: Google stadia game changer for opensim?

Dr Ramesh Ramloll
 'in terms of controls there's a lot to try and boil down into a mobile
device; if you want to support in-viewer build tools then that complexity
is even greater.  '
I agree no one wants develop tools on their phone, if we are talking about
such platforms, the assumption is that we are talking about the people who
are purely users with a interest in returns for their time investment. I
think for mobile devices, or even large screen applications, we really have
to make an effort to simplify the UI.  We spent quite a lot effort to
automate camera control (eliminate it all together) and convert all
interaction possibilities as single point interactions. And with this
approach, Stadia becomes relevant for  us and am sure for others too.

On Wed, Mar 20, 2019 at 9:29 AM Haravikk <[hidden email]> wrote:

>
>
> > On 20 Mar 2019, at 12:39, Dr Ramesh Ramloll <[hidden email]> wrote:
> >
> > Hello,
> >
> > At the last OSCC, there was a lot of positivity around a new modern
> viewer
> > using Unreal Engine.  Now with Google Stadia, it appears  that it will be
> > possible for opensim builders to access the same audience as other
> > platforms on any device. I know  'streaming' is pretty old now, but for
> > Google to go in this way suggests that they might have made it
> affordable.
> > Do we know what's happening with the Unreal Engine viewer effort?
> >
> > Best regards,
> > Ramesh
>
> I keep looking at streaming offerings but every time I do I'm not
> convinced; while there are some games where it's "fine" the quality just
> never lives up to claims and for twitchy games it's a non-starter.
>
> While opensim is arguably not a twitchy game where response times are key,
> you're still talking about adding another layer of lag between the user and
> the simulator which feels like a recipe of extra-strength headaches to me.
>
> There's also something to be said for privacy; a true virtual world is so
> much more personal than games like Call of Duty, Fortnite etc., you're not
> just dressing up the same basic player(s) that everyone else has, you're
> creating an avatar that *is* you, or an extension of you, in a virtual
> world. Introducing another party into that pipeline (other than grid, sim
> and you) feels weird to me.
>
> So yeah, I dunno; if streaming can be used to help people with less
> powerful hardware to get better visuals then sure, streaming could be of
> benefit to them, but for anyone with a half-decent machine streaming is to
> be avoided IMO. I'm also not sure how useful mobile support via streaming
> would be for a platform like OpenSim, in terms of controls there's a lot to
> try and boil down into a mobile device; if you want to support in-viewer
> build tools then that complexity is even greater.
>
> I don't think streaming platforms are a "game changer" for virtual worlds,
> they're an option for increasing accessibility given how demanding viewers
> tend to be on hardware, but the streaming aspect doesn't fundamentally
> change the difficulties of presenting a complex, malleable virtual world to
> users, streaming is really just a change in delivery method. I also don't
> think there's anything specific to Google's offering that makes it
> fundamentally different to all the past services that have tried;
> technology to implement it has improved, but you're still adding input and
> output latency plus video compression no matter what you do.
>
> It's worth remembering as well that aside from the initial rush of assets
> downloading in a new sim, the actual back and forth messaging is relatively
> light; video streaming on the other hand still requires a fair bit of
> bandwidth just to look good, as well as modern processors for full h.265
> support in hardware (or a dedicated GPU in which case you could probably
> handle a local viewer), so streaming isn't necessarily cost-free in terms
> of performance and hardware requirements either.
>
> Does anyone know if there's much demand for virtual worlds on the go? I
> don't know that it's something I'd fully want to do on my smartphone,
> though the ability to chat with people in-world would be useful; does
> OpenSim currently support a way to do this without having a zombie avatar
> sign into a region?
>
>
> Sorry for the text-wall and if I come across as too negative; it's an
> interesting technical space, I'm just a bit skeptical how much it really
> affects something like OpenSim.
> _______________________________________________
> Opensim-users mailing list
> [hidden email]
> http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-users
>


--
'Consider how the lilies grow. They do not labor or spin.'
*Rameshsharma Ramloll* PhD, CEO CTO DeepSemaphore LLC, Landisville, PA;
Affiliate *Research Associate Professor*, Idaho State University,
Pocatello, ID 83209 Tel: 208-240-0040
LinkedIn <http://www.linkedin.com/in/rameshramloll>
_______________________________________________
Opensim-users mailing list
[hidden email]
http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: Google stadia game changer for opensim?

Ethan Gardener
In reply to this post by Haravikk
On Wed, Mar 20, 2019, at 1:29 PM, Haravikk wrote:
>
> There's also something to be said for privacy; a true virtual world is
> so much more personal than games like Call of Duty, Fortnite etc.,
> you're not just dressing up the same basic player(s) that everyone else
> has, you're creating an avatar that *is* you, or an extension of you,
> in a virtual world. Introducing another party into that pipeline (other
> than grid, sim and you) feels weird to me.

I don't see the problem here.  Even taking the simplest possible case: roleplaying or personal chat on IRC, the server and your local computer (which may have malware) are already handling lots of very personal expression.  
_______________________________________________
Opensim-users mailing list
[hidden email]
http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: Google stadia game changer for opensim?

Ethan Gardener
On Wed, Mar 20, 2019, at 3:59 PM, Ethan Gardener wrote:

> On Wed, Mar 20, 2019, at 1:29 PM, Haravikk wrote:
> >
> > There's also something to be said for privacy; a true virtual world is
> > so much more personal than games like Call of Duty, Fortnite etc.,
> > you're not just dressing up the same basic player(s) that everyone else
> > has, you're creating an avatar that *is* you, or an extension of you,
> > in a virtual world. Introducing another party into that pipeline (other
> > than grid, sim and you) feels weird to me.
>
> I don't see the problem here.  Even taking the simplest possible case:
> roleplaying or personal chat on IRC, the server and your local computer
> (which may have malware) are already handling lots of very personal
> expression.

[I pressed the wrong key combination on my Twiddler & sent it before it was done, sorry.]

Apart from that one point, I've given up paying attention whenever any corporation or blogger announces a "game changer" because there are so many announcements and so very very few which make any real difference.  I'd pay a little more attention if I was in the biz, just in case it's got a chance, but I've seen so much hype that the very words "game changer" no longer brighten my day! (chuckling)  A true game changer can only be recognized in hind sight.  Perhaps Google have got some sort of price drop in the offing, but I don't see how they could bring it to everyone in the same way that the price of remote storage was suddenly hugely cheaper when AWS and GCS appeared.  

Taking a guess, I think this announcement is a political move against certain US states with internet access worse than the third world and effectively protected by law.  If lots of companies jump on the streaming bandwagon it will widen the gulf between states with reasonable and unreasonable internet laws, increasing democratic pressure for change.  Until that change happens, a streaming service will be accessible to less people, not more.  

Aren't viewers built with SDL for display?  There's an SDL library for Android.  OpenTTD is a C++ program built with SDL, and has an Android port which plays just the same.  It has some extra UI options to facilitate use with a touchscreen and no keyboard.  I don't know how much other work they had to do to port it.
_______________________________________________
Opensim-users mailing list
[hidden email]
http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: Google stadia game changer for opensim?

Dr Ramesh Ramloll
' Aren't viewers built with SDL for display?  There's an SDL library for
Android.  OpenTTD is a C++ program built with SDL, and has an Android port
which plays just the same.  It has some extra UI options to facilitate use
with a touchscreen and no keyboard.  I don't know how much other work they
had to do to port it.'

This would have been great except that this is not a simple UI problem that
can be ticked off through a selection of UI options. Half of the
optimization problem involves in world coding that does not touch the UI
aspect.

On Wed, Mar 20, 2019 at 12:28 PM Ethan Gardener <[hidden email]> wrote:

> On Wed, Mar 20, 2019, at 3:59 PM, Ethan Gardener wrote:
> > On Wed, Mar 20, 2019, at 1:29 PM, Haravikk wrote:
> > >
> > > There's also something to be said for privacy; a true virtual world is
> > > so much more personal than games like Call of Duty, Fortnite etc.,
> > > you're not just dressing up the same basic player(s) that everyone
> else
> > > has, you're creating an avatar that *is* you, or an extension of you,
> > > in a virtual world. Introducing another party into that pipeline
> (other
> > > than grid, sim and you) feels weird to me.
> >
> > I don't see the problem here.  Even taking the simplest possible case:
> > roleplaying or personal chat on IRC, the server and your local computer
> > (which may have malware) are already handling lots of very personal
> > expression.
>
> [I pressed the wrong key combination on my Twiddler & sent it before it
> was done, sorry.]
>
> Apart from that one point, I've given up paying attention whenever any
> corporation or blogger announces a "game changer" because there are so many
> announcements and so very very few which make any real difference.  I'd pay
> a little more attention if I was in the biz, just in case it's got a
> chance, but I've seen so much hype that the very words "game changer" no
> longer brighten my day! (chuckling)  A true game changer can only be
> recognized in hind sight.  Perhaps Google have got some sort of price drop
> in the offing, but I don't see how they could bring it to everyone in the
> same way that the price of remote storage was suddenly hugely cheaper when
> AWS and GCS appeared.
>
> Taking a guess, I think this announcement is a political move against
> certain US states with internet access worse than the third world and
> effectively protected by law.  If lots of companies jump on the streaming
> bandwagon it will widen the gulf between states with reasonable and
> unreasonable internet laws, increasing democratic pressure for change.
> Until that change happens, a streaming service will be accessible to less
> people, not more.
>
> Aren't viewers built with SDL for display?  There's an SDL library for
> Android.  OpenTTD is a C++ program built with SDL, and has an Android port
> which plays just the same.  It has some extra UI options to facilitate use
> with a touchscreen and no keyboard.  I don't know how much other work they
> had to do to port it.
> _______________________________________________
> Opensim-users mailing list
> [hidden email]
> http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-users
>


--
'Consider how the lilies grow. They do not labor or spin.'
*Rameshsharma Ramloll* PhD, CEO CTO DeepSemaphore LLC, Landisville, PA;
Affiliate *Research Associate Professor*, Idaho State University,
Pocatello, ID 83209 Tel: 208-240-0040
LinkedIn <http://www.linkedin.com/in/rameshramloll>
_______________________________________________
Opensim-users mailing list
[hidden email]
http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-users
Reply | Threaded
Open this post in threaded view
|

Re: Google stadia game changer for opensim?

Ethan Gardener
On Wed, Mar 20, 2019, at 6:56 PM, Dr Ramesh Ramloll wrote:

> ' Aren't viewers built with SDL for display?  There's an SDL library for
> Android.  OpenTTD is a C++ program built with SDL, and has an Android port
> which plays just the same.  It has some extra UI options to facilitate use
> with a touchscreen and no keyboard.  I don't know how much other work they
> had to do to port it.'
>
> This would have been great except that this is not a simple UI problem that
> can be ticked off through a selection of UI options. Half of the
> optimization problem involves in world coding that does not touch the UI
> aspect.

That was the moment I forgot about the difference between 2D and 3D. ^.^;;;  OpenTTD's graphics are rooted in 1992.  It's had some additions, but still doesn't even have full scaling, only 2x, 4x, etc.

I'm not sure it's possible to optimize a SL/OS viewer enough for tablets.  Viewers are already capable of extremely heavy LOD-culling as an optimization, but it ruins builds.  For a few months, I tried to explore with a rather unsuitable computer: decent CPU but Geforce 9500 graphics card.  With viewer preferences set to try to cope with all the high-lag mesh around, I went to a big, nice, low-lag mesh build, and it looked *terrible!*  I could see that there were buildings, but I couldn't see the facades even at a moderate distance, they were *destroyed* by LOD-culling!  I suppose the viewer algorithm of when to apply that optimization might be improved.

I'm concerned about caching too.  It's needed, but how well is it going to work with the slow write speed of Flash memory?

A cheap work-around for both of these problems would be to eliminate mesh objects, (no mesh uploads, no connection to the OpenSim hypergrid,) maybe restrict texture sizes too, and stop sending meshed prim data from the grid to the viewer.  Yay, it's 2006 again! ^.^;  It could be somewhat better than old SL, but still... "I'd play it, thousands wouldn't," perhaps? :)  Actually, I wouldn't play it, I get enough neck pain without using tablets and phones.  "Thousands" would play it who wouldn't play SL/OS because they don't have a gaming computer -- compare Minecraft & Roblox.  Less cheap would be building into the viewer a mesh editor which can't produce high-lag meshes, but that would be quite a challenge just to design.  A built-in texture editor would perhaps be even more desirable.  But I digressed, I believe the original point was business rather than pleasure?
_______________________________________________
Opensim-users mailing list
[hidden email]
http://opensimulator.org/cgi-bin/mailman/listinfo/opensim-users