Lowering the barrier to entry

Hi guys,

I was flicking through some of the posts in the Developer Tools section, and came across Tecknowolf's post about Multiverse's high skill requirements.

Now, he has a fair point - I'm still struggling to learn all the required knowledge, but that's mainly because I've been busy with my own stuff. But even then, there is a lot to learn, and sometimes the tools don't really make it easy.
Of course it goes without saying that the Wiki is a great help, and will continue to help, so long as we keep it updated with the latest advances with the Open Source Project (which I will refer to from now as OFP, because I'm a lazy typist!)

My aim is to attract as many new developers to the project as possible, in order to create a burgeoning community, where people can help each other out with everything, and even get some worthy projects in the pipelines. The problem is, if it's so difficult to get into developing with Multiverse, we won't get very far as a community!


So, what do you reckon we could do to lower the barrier to entry? (so to speak)

Maybe to have some pre coded plug-ins would ne nice. Not sure because it is hard sine this is a metaverse platform and even though I find it difficult to learn everything I have to learn, there is so many components to think about and to involve.

I also say to make a simpler server. Instead of what we have now a simple download that downloads all necessary components and a start button that will start everything correctly. That would be good.

I have given it some thought and this is what I say:
Make an easier to use server with a download bundle that downloads all neccessary components.
Have everything that is neccessary for a morpg already built into the platofm.... currency system crafting animals monsters etc.
Have an easy way to add new characters that have animations.
Have an easy way to add music to certain regions.
Make the master server the users server so all accounts that are create are created on that server.
Allow the creator to create an account on his server during the installation period.
I believe from what I have seen the map edditor is pretty simple
NPC easy creation with dialog
When water is on map allow after x dept character swims instead of walks
Easily add food and craft recipies with a type of manager.

That`s what I have up to now

Keeping things simple is grate. What drew me to Multiverse was the networking setup. Getting the server up and running was a pain however it showed you what peaces could be pulled away and run on different servers. The reason I left was that it seemed that the program was dropped and was no longer being developed. After looking at many different game platforms I came to admire what Multiverse had going for it. The client, log in server, user data all ready for you.

Something you have to keep in mind most of us do not have the cash to sink into a huge supper computer how ever by splitting the game up on many cheep small computers it becomes possible to expand with out a lot of cash. So far Multiverse is far above every thing I have seen out there. I would hate to see it lose this just to make it simple.

I ended up going with Torque 3D because of the support and the fact that it was being developed but it had no networking built in and right now this has my project at a stand still.

What is needed is not to make it more simple but a clear set of instructions that work every time. If you over simplify the program then you can lose a lot of the versatility of the program and I would hate to see that. Sure it might be hard to add a not monster in to the mix but by the time your done you know ever aspect of that critter and can truly make it different that the other monsters in the game. This is where a clear set if instructions help and people to lean on if you run in to issues.

Keep in mind I have not taken a look at Muliverse in over 2 years. The only other issue I ever had with Muliverse was the master log in server that I had no control over. I am one of those that want everything under my control and not have my project down because someone else server is down and there is nothing I can do about it. I could understand why they did it, to keep control over the program so people did not steal it, but now that it is open I hope this will change.

What I am hopping is that Muliverse will keep all the good things, add grate docs, source code, be able to have your own master control server and get a following behind it. Then maybe if I can bring all my assets over from Torque 3D, ok I know that is way to much to hope for but I can dream.

Whilst I have no previous experience with multiverse, I non the less agree w/ what Wolfen said!

Should there be an example set up to work from, one that work's "out of the box"(at least as close as practically possible), yes!
But especially the server needs to remain as versatile and configurable for different set-up's/needs as possible!

Off course, I'm pretty sure I'm just stating the obvious here, but redundancy is never truly redundant, if that makes sense, so...

You can now control your own master server...most likely we will also provide a master server for teams that do not want to start their own.
The master server is setup the same way as the regular server but with a little different mysql schema

We will need to work on making the server split up easier...currently you can split up some parts but you will probably still need one powerful central server to run the world manager (unless you have a small world.) At some point we will work on making a distributed world manager but it is not currently in the code.

Why not allow both options... Keep one with he source code as is and keep it with a moderate level entry and have another where you make it simpler. So a simple version and an advanced version so people can stat with the simple and when they are finally able to use the advanced then they can do so. It can make the best of both worlds.

We could have the best of both worlds with ease, and still only have one project instead of 2. If you keep it where the program can be broke up to many servers does not mean that it has to be, you could all ways install all the parts on one computer. The biggest thing is having a clear set of instructions that is EASY to follow and explain not only how but why when you set it up. The more detailed the better.

My pref is rundowns using ubuntu server, because of the cost and having one starting point for everyone to start from. Not everyone has a copy of windows XP laying around or wants to install a copy of Cygwin on their main computer esp if they are not sure what they are doing, but everyone can down load a copy of ubuntu 10 and if you do not have the hardware to run it on, you can even grab a copy of Virtual Box and run ubuntu in it (so it will be a lot slower we are talking testing and dev work to get started not production). The easier it is to get your feet wet the better.

On a side note why do it on bare bone hardware or a VB box? Easy “KISS”. Keep It Simple Stupid. Otherwise you have to deal with what else the user has installed on their computer, and hunting down what broke what or let’s say you do get it to work and you dev a project that you want to go live. Hmm do I install my desktop computer in production or try to rip out the project and figure out all the software and settings that may or may not be needed. Also easier to reproduce and test, you can go from the install of the operating system to logging in to the server step by step with nothing getting in the way or forgetting that you needed some piece of software install first. Everyone can do it the same way and expect the same results.

The issue with have 2ver is twofold. 1 – Now you have 2 different ver of the program that has to be developed and supported, and soon they start growing apart where things will work on one but not the other. 2 – A lot of times when a program has a simple and a hard ver there is still a step learning curve when you switch to get more functionality not to say anything of the issues of porting your project from one to the other, a lot of time it is easier to start over. (Can you tell I have gotten burned before?) LoL

It does not matter how simple you make the program if the instructions are not there. However even the most complicated system in the world can be a joy to set up if you have a guided path.

Personally I have 37 1u servers sitting around doing nothing and would be a good start on a huge MMO if I can divide it up among them.

This weekend if I have a chance I am going to take Multiverse for a new spin on my VMware server and see how it goes. If all goes well I will migrate it over to a few of the 1u. I will let you know how it goes.

With proper code architecture, running with a master server in a different process than the world server or running both in the same process is pretty easy. As long as the OO is sound and the architecture is properly written, it should be extremely easy to do. Now, I've been watching MV and reading the code for a while (read: years) and so I have a good feel for what the original MV guys were shooting for.

What would help the current MV devs would be lists of 'user requirements' from our community members. (much like what gabzo posted above) We can bounce those requirements off the current code base and see what changes would be needed to support that specific requirement. From there, we can develop a realistic road map and charge forward!

Quote :

My pref is rundowns using ubuntu server, because of the cost and having one starting point for everyone to start from. Not everyone has a copy of windows XP laying around or wants to install a copy of Cygwin on their main computer esp if they are not sure what they are doing, but everyone can down load a copy of ubuntu 10 and if you do not have the hardware to run it on, you can even grab a copy of Virtual Box and run ubuntu in it (so it will be a lot slower we are talking testing and dev work to get started not production). The easier it is to get your feet wet the better.

So................why not build a linux boot cd, that has everything installed and configured on it?

Food for thought...

Mulitverse hopes to create a revenue stream. One way of doing that is, having a hosting server online, whereby users can link to it, via an paid for account, in tiers. Say 10 users for 1 hour, or 1 user for 10 hours, increments. I could then purchase say 10 user hours for testing, then use another paid for tier of 10 users for 1 hour, for exhibitions/showcasing my work. If it works out well, I could then, with the potential revenue I would generate, host my own server.

If Multiverse is to showcase itself, at some point there will have to be a 'home world' virtual space for users to login. In the initial stages, perhaps the funding to run this server could come from donations. I'm sure, if the pricing structure was right, it would be a cost effective way of getting started and generate new users/developers to the platform.

Yes that is the plan. We have the demo world currently running and will be developing a "Showcase" world for others to enter. We are working with our hosting and VPS partner, Vanquish VPS, to setup plans to accommodate Multiverse users.

Long story short, we actually have many many incredible ideas on what to do. We have not yet shared them, and perhaps that is something we need to lay out for the community. I'll work with Tristan on this.

I realize that some of the posts on this thread are very old but I am just now getting around to starting up my project and wanted to add some things that I have found and thought about in the last six months and I figured we needed a fresh bump for this thread.

I just set up Webmin on my Linux server and while I was rereading this thread Gabzo's initial post struck a chord. With all the development that has and is going into Multiverse, there should be a "Server" web-based front-end that would allow an admin to:

• Upgrade the server
  
• Upgrade the database
  
• Start/Stop the server
  
• Activate/restart core server modules (such as mob engine, item, economy, combat, etc...)
  
• View server statistics (i.e. up time, error statuses, etc...)
  
• Clone modules - This would allow an owner to copy a core module and modify the code for that module so that the module would be custom for that world. Users would not be allow to touch the core modules so upgrades to the core modules will always be available. Custom modules are the owner's responsibility. The Server Manager would allow the owner to register custom modules to use in place of the core modules.
  

I am sure we could think of dozens of things to add to this list that need to be managed but I think the focus should be on making the initial set up easy and manageable. Once there is a easy management system for the server I think we can focus on making the necessary modules and specifics.

Just some thoughts.

On another note, after reading the DB thread.

I do not believe that access to the core database should be open, now, I know there is a strong argument that open source is open source but the underlying goal is to make an open source MMO that our teams can use to create our own custom worlds.

What that means is that as part of the core distribution, the database setup would be set and unchangeable. If teams want to customize modules that require additional database modifications then we should allow teams to extend the core database connectivity module to create their own database connections for their custom module. This would allow the teams to use the core data and their extended data. This maintains the integrity of the core database but still allows for modifications.

Eliminating the need for someone to be a database architect would also help lower the barrier to entry.

The more we lower the technology aspect of the project the more likely we are gain teams that may not be as technically savvy to join the community. This may be good as I am encountering a lot of good game designers that are not technical enough to put together a basic Multiverse MMO.

I agree 110%
Do you have any development experience in this area?

Not directly, or at least not yet, or maybe some.

Confusing? Sorry about that but I have to think about it. My initial thought is no, not for what I described; however, I have experience with MySQL, PHP, Javascript, etc... I could probably poke around, clumsily at first, and see what I can come up with.

Without completely oversimplifying it, the GUI would be a simple administration program that calls the appropriate scripts with the appropriate permissions. The initial install package would still need to be complete and make the initial attempt to install the servers, database, and a listener so the user can log in to the manager's web GUI.

A web based java or PHP control panel would be best. I was originally shooting for a Windows based control panel since that is where most of my experience is. However, there are obvious limitations to a Windows based app. Having it web based would allow it to be installed on the remote linux server that runs the MV servers. This is a great idea and one that should be moved forward. The hard part is of course, actually making the tool and without willing developers to contribute it won't go very far. That is why I asked if you have any experience in that area. I'm a little lacking when it comes to that technology. Of course I could learn it.. but things could move quicker if someone is already an "expert" in building that type of tool.

Thank you for the ideas they are very valuable and good. Perhaps you could "mock up" your ideal user interface. Something that you feel would be provide the best work flow?

I need to read up on PHP and the like anyway

Merry Christmas and Happy New Year!

Shane C Fischer
CobaltBlues

Merry Christmas and Happy New Year to you too.

I agree it is difficult to do it from the ground up but maybe we do not have to.
Why reinvent a perfectly good wheel.

Here's a link to the Module Development site for Webmin (http://doxfer.com/Webmin/ModuleDevelopment). Maybe we can just create a custom module directly for Webmin, since based on what I just read this is exactly what we want to do. There are modules to manage MySQL, Apache, FTP, DNS, etc. We can create one to manage Multiverse. Then the user will have access to both MySQL and Multiverse from the admin same tool.

I am going to play around with this over the next several days/weeks.

I will also mock up some thoughts, regardless of if we can leverage Webmin directly or not.

Ah ok that's a good idea. I'll take a look at how this tool works. I've not seen it before.

Shane

just fyi..

Installing Webmin over YUM for CentOS
http://chrisschuld.com/2009/11/installing-webmin-with-yum-centosrhel/

I installed Webmin on our Foundation server and Wow. It sure makes administration of the system a ton easier. I wish I found this a year ago. Thank you! We should definitely investigate creating a module for this admin tool.

Shane

Should we start a new thread to explore this option so we can post our findings?

My brother showed this to me earlier this year and I did not pay much attention to it until I had the server admin wipe my server and I installed it myself. As I did, I was reading about it and realized how useful it is.

I have installed the example module from the development page and looked at the code. It seems we could create the module easily and build it from there, incorporating the steps necessary to install and monitor the server one piece at a time.

- Rich

New thread here

https://multiverse.forumotion.co.uk/t361-multiverse-webmin-module-development#1946