Thinking about infrastructure

Open: Building Myst/Uru-style worlds and games on the OpenSim Hypergrid

Moderator: OpenSim Hypergrid Managers

User avatar
Dot
Member
Posts: 222
Joined: Sat Dec 13, 2008 7:42 am

Thinking about infrastructure

Post by Dot » Sun Jan 10, 2010 7:16 am

Originally, I was discussing with JW about setting up a project for Devokan alone. But from recent developments it seems that a better plan might be to pool efforts in 'cooperative independence' (to use Nalates' helpful phrase) and see where we get to.

This new discussion was initiated by pm to a group of individuals I felt might be interested. However, it was thought that a better place for discussion would be here to enable all to take part.

The aim as originally stated was to eventually agree on a basic structure for communicating plans and ideas with each other (perhaps run as separate subprojects, so that we could each retain our independence), together with working out a practical way of linking to the different grids based on the OpenSim Hypergrid system -- which would probably include exploring the server idea mentioned by JW. Then we could set up the resources here for the umbrella project, and here and/or in other places for subprojects depending on each group's needs.

However, it seems that a broader discussion would be in order, as will be seen from the posts below...

User avatar
JWPlatt
Member
Posts: 1100
Joined: Sun Dec 07, 2008 7:32 pm
Location: Everywhere, all at once

Re: Thinking about infrastructure

Post by JWPlatt » Sun Jan 10, 2010 7:33 am

The post of mine Dot referred to is this one in the Systems Concepts Project:

Server Provisioning & Billing Services On OpenURU.Net

I will also be referring to that post in a little bit. Additionally, this discussion may well enter into the space of what the Systems Concepts Project does, so let keep an eye on that.

I appreciate Dot reposting from the PMs so others can watch, learn and participate. This could remain a single thread or it could grow into a project with many threads, or many subprojects as Dot mentions. Just so I don't get too absorbed in creating a plethora of projects and so the site doesn't get a lengthy index page, an umbrella project would be nice if it could be structured that way. That keeps the index page as short as possible. Within, you could break out any major subprojects (things that need more than a few threads in the umbrella forum).

So... what's the umbrella, I guess is a question. In my mind, it's the resource of bringing up an infrastructure of servers and a way to provide and bill for them when they are not donated. But it's not just servers - it's a complete environment of being able to seamlessly order, receive and use servers and software. It could be Uru; it could be an SL shard or whatever they call it, it could be the emerging things that Dot and all of you are interested in pursuing. You could call it the "Virtual Worlds Project" or some such name. You could call it The Uru Cloud. Heh. Or, oh, yeah - the title of Dot's PM thread: "OpenSim Hypergrid Project"! I am beginning to look into how horrendous it might be to have a front end to Amazon Web Services (AWS) for use by the community. I have an account and am expecting a call from an AWS representative to do some discovery about what's possible.

In the Systems Concepts Project, I made it clear that Uru is the concern on this domain. If it weren't for that, this kind of project could happen anywhere; be done by anyone. And so it probably is better done elsewhere except for the fact that we're talking about Cyan IP, Uru IP specifically, and Uru generally - even broadly. The premise of cooperative independence makes it possible for me to be very enthusiastic to support even unrelated sim work because it means building an infrastructure that the Uru community can use. That's my interest. I intend to promote services where talent can make profit for their creative work. Or so services can make profit to keep it a living system. That's the .net or .com side of things perhaps in the future. The .org side of things now and in the past is and has been for people who eschew that kind of distasteful commercial thinking. That has been the prevailing attitude of this community so that's very important to support at the root of all things. It is open source, after all. This is all to give you an understanding of my motives and interests. Providing web services costs money and so money must be made to keep something like OSGrid available, but without a free Uru I would have no motive and am not interested in building this as a public service company if Uru is not at the heart of it.

That said, I should probably repeat my disclaimer that this all depends on Cyan open source licensing of Uru. That's a big caveat, but I have faith and I don't think we have a moment to spare in planning for it because when Uru hits we want to be on the ground running.
Last edited by Mac_Fife on Sun Jan 10, 2010 3:30 pm, edited 1 time in total.
Reason: Fixed URL
Perfect speed is being there.

User avatar
Nalates
Member
Posts: 437
Joined: Mon Dec 22, 2008 7:50 pm

Re: Thinking about infrastructure

Post by Nalates » Sun Jan 10, 2010 10:03 am

JWP, thanks. Your thinking works for me.

On commercial... I agree. I have a connection and the GoC Tech Data site piggy backs on a commercial site's extra space that I free lance for and have a history with. But some where things have to be paid for. I am of the mind that if no one but me wants to see something or participate then I can do it at home. Money spent for others that never come to play is a waste. The bills tend to focus us to those things people enjoy and are willing to support.

I understand others like free. (place long rant on the evils of free social services here). In our community a number of people are willing to share. That works for me. The division of .com and .org is a good idea and one I like. A presentation of choice so blatant should be more acceptable than any thing I've seen in the community yet and I hope it works well.

On sections... If you have ever looked at the SL forum... that index page is a nightmare. There is so much to consider. For all that is there I think they did well. But, I would like to avoid three screens of sections.
:(
Perhaps a top section for the SL and OS related grids. I would toss Blue Mars in the group. When SL/OS grids add meshes these VW's will have much more in common.

I will have to find someone in GW that could tell me if there is anything Uru fans are doing there that might parallel the SL/OS/BM type building. I did just see a Team Fortress building post on MOUL. So, there are other possibilities. In all these cases people are doing something that is related to there current game and trying to connect it to Uru and their friends.

In SL Devokan I decided to experiment with mixing my Devokan RP with my Land of NoR (LoN - abbreviations of abbreviations now :roll: ) RP. It is a huge contrast. Devokan is PG and non-violent. LoN is adult and all about violence. Is it possible to mix the two? I think mostly. Dot and I are pre-planning the direction of the RP. My thinking is if this works, even if no more than an example of concept, many players could RP the time out of the cavern with their other characters on other VW's. I think for many that could be fun. I've enjoyed figuring out how to mix the character's different environments.

The possibilities suggest to me that a top section might be Non-MOUL Uru Related VW's (give me time I might think up some bloviated named). Add sub-sections for other STYLES of VW’s as needed. Within those sections add subsections for specific regions (an SL type age) like Devokan, servers, hosts, setting up home servers…

We can go public and decide once we have a top section.

Some of the people that will start to find us will not know Uru. So, defining Uru as a PG kid friendly Player-with-Player-cooperative environment and that this is a Myst-Uru oriented effort probably needs to be made obvious.

Again thanks,
Nalates
GoW, GoMa and GoA apprentice - Guildmaster GoC - SL = Nalates Urriah

User avatar
Dot
Member
Posts: 222
Joined: Sat Dec 13, 2008 7:42 am

Re: Thinking about infrastructure

Post by Dot » Tue Jan 12, 2010 7:16 pm

Tai'lahr suggested that we might copy and paste relevant information from the helpful discussion in the Eder d'Uru section. Here's my attempt to do just that.

The thread grew out of discussions within the SL Uru community of what was possible within grids based on the open-source OpenSim (OS) software, as a more affordable route to build Myst/Uru-influenced or other style areas for the Uru community to enjoy.

Various grids and their costs were mentioned, such as:
  • Builders of Arda grid
  • Inworldz
  • OSgrid -- different hosting options are available for this, including BYO DIY
  • standalone OS region set up on a home computer
Blue Mars was mentioned as another option, and various other virtual world (VW) possibilities.

Several of us chimed in with our own experiences of two of the OpenSim-based grids: Arda grid and OSgrid.

Then came the suggestion for a project, from which the pms behind this thread arose:
One thing I'm hoping we will be able to do is to explore the methods of teleporting between the various open-sim Myst-Uru areas on different grids, maybe setting up a special hub region using the Hypergrid. Then explorers could link to the hub, and from there use a special portal to the Myst-Uru-friendly builds on the different grids. I'm currently discussing this with JW as a potential project here on OpenUru.

To my mind there would be several aspects:
  • the technical aspects of setting up and running the server software and portal software
  • creating a meaningful inworld model for explorers to use to select destinations and to link (akin to the Nexus in Uru)
  • gathering, organising and keeping up to date the information about the different builds on the various grids, publishing it in-world and on the web
  • publicising the facility among the wider Uru community
At the moment, I'm not sure that we are all talking about the same thing, which makes it hard to outline a clear structure for a possible project.

Perhaps a way forward would be for each of us to write a simple bullet list like the above, of what we personally see the main aspects of the umbrella project to be. Then we might be able draw out commonalities for the main project, and spot what would be subprojects.

User avatar
JWPlatt
Member
Posts: 1100
Joined: Sun Dec 07, 2008 7:32 pm
Location: Everywhere, all at once

Re: Thinking about infrastructure

Post by JWPlatt » Tue Jan 12, 2010 11:12 pm

I had a quick preliminary discussion with an Amazon rep this evening. The requirements I stated were:

1. No cost until used (on demand)
2. Front end on this domain for provisioning (without developing an AWS API app ourselves)
3. Automated billing to each client (not OpenURU.net) through the front end here
4. Server instance (EC2) and storage (S3) control for each client through the front end here

Requirement #2 is currently a question mark. We'll probably either need a service at some cost, or develop our own front end (or find an open source AWS front end solution). The rep will be putting me in contact with a potential "white label" service soon.

One question I could not answer is how many people do we expect to actually sign up and pay for server instances? And many instances? That would be helpful. I guessed certainly less than one hundred, if many at all (not big). Support your numbers.

One goal here is to not burden OpenURU.net with administration beyond installation and maintenance of the front end (no service billing and management!). OpenURU.net might set a margin on pricing to pay for any white label system fees and any for-hire techs.


(P.S. If it weren't for the fact that my post here is more or less a small detail of your larger discussion about grids and Virtual Worlds, it should be in the System Concepts Project. Like I said, let's keep our eye on that and see how this should be structured. System Concepts is a solid project that should be considered more an umbrella for other projects like Open Uru - which itself is one subproject of something like this - than something to be part of any one project.)
Perfect speed is being there.

User avatar
Nalates
Member
Posts: 437
Joined: Mon Dec 22, 2008 7:50 pm

Re: Thinking about infrastructure

Post by Nalates » Wed Jan 13, 2010 3:51 am

@Dot, This gives me a headache. I’ve looked it at so many ways and added stuff, subtracted stuff, renamed… This seems the simplest.

Orientation
* Myst-Uru close to Cyan story
* * Client/Player (Plasma)
* * Server Operators (Plasma & Alcugs)

* Myst Style - distant story related or style only related
* * Client/Player – may not be Plasma/Alcugs related. SL/OS/BM
* * Server Operators - SL/OS/BM

* Modeling – This is going to become more related to SL, OS, BM and MOOU as meshes come to SL & OS grids. So rather than repeat it in section, have a single section.

Non-Myst Related section? – I doubt we want to support this… those interested could work inside existing style sections.

As to the number of people interested in running servers and/or purchasing hosting, do we have any idea?

I can’t see incurring a monthly cost that requires some minimum number of subscribers to cover costs. If it is on demand that would be nice.
Nalates
GoW, GoMa and GoA apprentice - Guildmaster GoC - SL = Nalates Urriah

User avatar
Dot
Member
Posts: 222
Joined: Sat Dec 13, 2008 7:42 am

Re: Thinking about infrastructure

Post by Dot » Wed Jan 13, 2010 5:50 pm

Just to check my understanding -- there are several inter-related strands:
  • The first is to do with the technical aspects, part of the existing System Concepts project (JWPlatt, possibly Nalates).
  • The second is to do with structuring content (where my interest primarily lies; possibly Nalates).
  • The third is to do with running costs and how to finance them (GoH; JWPlatt, Nalates).
Yes? Or am I being dense and should retire gracefully from the discussion?

User avatar
Nalates
Member
Posts: 437
Joined: Mon Dec 22, 2008 7:50 pm

Re: Thinking about infrastructure

Post by Nalates » Wed Jan 13, 2010 7:09 pm

Dot, I think you have it. I see the logical divisions you list. I also think story and in-world building have strong influences on the infrastructure. In RL we build the water lines and stuff to support what people want to do. No one builds a house/building to support a water line.

I'm trying to think of how someone trying to get a sim/region/age, whatever, up and running or someone new wanting to get to one of the alternate ages would think of it.

I anticipate Uru fans will want to see places that age writers/builders (may be there is a distinction there - writers out-world tools, builders in-world tools?) have made. Because of my SL connection I may be biased, but I think as they see what others are doing they will want to try their hand at small things. From there larger things and eventually entire ages.

So, things start with getting there. Then how to get around. Next how to build little things, then bigger things then ages. Once they can build ages then how to share them (servers and hosting). Once they are at that stage it is then about how to get word out and players in and how to share the age so all can play. I think that process needs to be supported in the forum structure and information arrangement.

Some of the challenges for MMOG makers is how to allow players to shape the story (Devokan journals, public in-world bulletin boards, forum threads). Another challenge is in how to allow players to affect the world. In Devokan I was thinking of tracking down the squees and catching them for a trip to the OS Devokan. Having a place where I could set the travel cage would be neato! Once one has met a griefer they know the problems with that. I think we can work those things out and create new ways to create a game that is more dynamic than MOUL was. I am anticipating Crux Isle opening and seeing how they have done with their puzzles.

I have lots of ideas for things we can do. I am hesitant to put massive effort into something only a few will be interested in. I'm also reserved in what I suggest for others until I have gotten try some things my self, which is why I'm building a guild hall in my sim (as well as the fun and learning). However, if we do all this as we play with OS Devokan and can help OpenUru.org organize and lay foundation for a future 'shard' and 'region' hosting system... \o/ ...and getting to do it with tolerant and thoughtful people... I think it should be fun and attract others.

Lots of the information will fit in more than one place. I tried to make the broad areas that could hold most of the areas of related interests and seem workable for me. How that gets shaped to make sense for others is where yours and others input is important.
Nalates
GoW, GoMa and GoA apprentice - Guildmaster GoC - SL = Nalates Urriah

User avatar
Dot
Member
Posts: 222
Joined: Sat Dec 13, 2008 7:42 am

Re: Thinking about infrastructure

Post by Dot » Thu Jan 14, 2010 2:45 pm

Nalates wrote:Dot, I think you have it. I see the logical divisions you list. I also think story and in-world building have strong influences on the infrastructure.
Thanks for the reassurance.

Re infrastructure following from story and in-world building: One way forward (using the OS-based grids as a trial run) might be to build on what is already happening, with a mix of home-based part-time sims and full-time regions on different grids and simhosts, and to develop a server/region that allows relatively straightforward travel between grids. That would allow each building group their independence, while starting to trial ways of supporting cross-grid travel. In UU terms I guess this would be akin to facilitating cross-shard linking.

JW, how does this tie in with what you are wanting to do? And should we take the discussion of content on OS-based grids elsewhere, given that it is not purely Uru-based?
I'm trying to think of how someone trying to get a sim/region/age, whatever, up and running or someone new wanting to get to one of the alternate ages would think of it.

I anticipate Uru fans will want to see places that age writers/builders ... have made. Because of my SL connection I may be biased, but I think as they see what others are doing they will want to try their hand at small things. From there larger things and eventually entire ages.
Umm, it didn't work like that with Paislee and myself, where we STARTED in SL with building areas with pre-built well-designed objects, working up to whole sims, and then using a mix of our own and others' creations -- but this might be because we started our 'Uru exile' in There, where all things are prebuilt and land is relatively cheap. The next step for us, with hosted regions on the OSgrid, is a combination of factors from both SL and There: build it (mostly) yourself, plus cheap land.
So, things start with getting there. Then how to get around. Next how to build little things, then bigger things then ages. Once they can build ages then how to share them (servers and hosting). Once they are at that stage it is then about how to get word out and players in and how to share the age so all can play. I think that process needs to be supported in the forum structure and information arrangement.
Suggestions for prioritising and organising information (mostly Wiki based):
(a) How to get there, sort out avie, move around, tp within grid, tp between grids (this might need to be set up per grid)
(b) How to set up region -- home-based; where to look for hosted options, and costs

Tips on building would be further down the line because there is already a lot of info out there -- and skills learned in SL are transferable to the other grids.

I see the publicising of what is out there to the wider community as a separate task/project. It could be both in-world, maintained at the linking region, and off-world, akin to the Age list maintained for Drizzle at http://www.the-deep-island.de/uru-ages/uam.updates.htm, for example.
Some of the challenges for MMOG makers is how to allow players to shape the story (Devokan journals, public in-world bulletin boards, forum threads). Another challenge is in how to allow players to affect the world. ... I think we can work those things out and create new ways to create a game that is more dynamic than MOUL was. ...
I would suggest that allowing fans to shape their world is essential -- whether through designing/building or through events or storytelling. That is going to be a challenge for pure Uru. Some of us have been spoilt by our time in exile -- we've seen (and shown) what more might be possible. (Nalates, let's talk about squee tracking in GoMa. ;) )
I have lots of ideas for things we can do. I am hesitant to put massive effort into something only a few will be interested in. I'm also reserved in what I suggest for others until I have gotten try some things my self... However, if we ... can help OpenUru.org organize and lay foundation for a future 'shard' and 'region' hosting system... \o/ ...and getting to do it with tolerant and thoughtful people... I think it should be fun and attract others.
Agreed :-)

User avatar
JWPlatt
Member
Posts: 1100
Joined: Sun Dec 07, 2008 7:32 pm
Location: Everywhere, all at once

Re: Thinking about infrastructure

Post by JWPlatt » Thu Jan 14, 2010 3:07 pm

Dot wrote:JW, how does this tie in with what you are wanting to do? And should we take the discussion of content on OS-based grids elsewhere, given that it is not purely Uru-based?
That's a good question to allow me to explain how, um... open... OpenURU.org should be to loosely coupled ideas. You would be correct to think that the Writer's Club Project does not necessarily apply to Uru. There is no requirement to write Uru-specific work. It's just there for the community to be creative. But it does provide a space for Uru-specific work too that could be used in UCC sometime or is even simply for creative people to maintain a presence in the community. The more projects like that, the better the chances of achieving something more than each individual project. So if your project helps to build an infrastructure and serves as an example development platform for cross-grid travel, that has implications and benefits to Open Uru as well. Even if the code is not directly usable, the architecture and algorithms might be useful to implement cross-shard travel. There's not much better than an infrastructure for support and good examples for inspiration. In other words - cross pollination, or keeping up with the Joneses. ;)
Perfect speed is being there.

Post Reply

Return to “Myst/Uru-Style Hypergrid & Open Simulators”

Who is online

Users browsing this forum: No registered users and 1 guest