next »

[Sanojian]

I decided to re-visit my Assemblee entry now that I have a little bit more knowledge and experience.

Here is an early prototype

It is a little MMORPG made entirely out of javascript.  I am using Craftyjs and Raphael for graphics.  No Canvas.  I have not decided on what I will use for Audio yet.  Currently using Assemblee graphics from dbb, oryx, oddball, and geeze.

On the server side is Google App Engine and the Channel API for networking.

create character


fighting


looting


dungeons


multi-player

Stuff that is broken:

• server only remembers your character name and class
• you can't see other players fighting, just walking and talking
• you health has no effect
• your mana has no effect
• your items have no effect
• not all classes implemented (not enough graphics)
• lots more stuff I am forgetting

I am looking for help to complete this beast if anyone is interested.  Thread here  Progress will be painfully slow (kids take priority) but hopefully steady.

[SirNiko]

I remember this from Assemblee!

I'm curious where you plan to go with this. Given recent F2P mmos (like Realm of the Mad God and Spiral Knights) I'm not sure what niche this game will fill.

[emacs]

Just a FYI - the link to your Assemblee entry is broken.

[kamac]

It look nice  '

If you'll somehow finish it, then I applaud to your skills.

[Sanojian]

Fixed the link, thanks.

I am planning to make it play like a low-rent Diablo MMO.  Much faster than my last version. That is not saying much since the last one played like swimming through tar.  But it will be slower-paced than a game like Realm of the Mad God.  I have never played Spiral Knights so can't comment on that.

Some plans:

• NPC scripting and quests
• 3 player teams
• PVP zones with objectives to fight over

I feel like my chances are pretty good for making a playable game out of this. Hey, I made the first version in one month and had no idea what I was doing.  Now at least I have a little idea.

[Sanojian]

What is working:

• Account and Character creation (requires Google account)
• Login
• Simple-minded MOBs
• Fighting and looting
• Zones (2 populated with MOBs)
• One wizard spell (magic missle)
• Inventory drag and drop
• See other players moving
• Chat with other players

What is broken:

• MOBs and loot are client-side only (vanish after logout)
• Your health and mana bars affect nothing
• Your items affect nothing
• Things get wonky in multiplayer when moving between zones (ghost textboxes, etc)
• Graphical glitches in fighting/casting animations
• MOBs can sometimes crawl through walls

Coming soon:

• Move MOBs and loot to server-side
• Make loot permanent in inventory
• Friendly NPCs
• Player death
• MOB respawn
• Player levelling
• More wizard spells
• Better wall boundaries (current ones just squares)
• Lots more zones and MOBs

In the plans:

• More player classes (priest, thief, ranger, paladin)
• Simple NPC quests
• Buy and sell loot
• 3 player teams
• PvP zones
• PvP objectives (capture the castle, etc)
• Better MOB AI and casting MOBs
• Touch controls for phone/tablet support

[Sanojian]

Here is the latest version

What is new for 2012-03-11:

• Moved MOBs to server side
• Distributed MOB AI between clients
• Implemented zoning for players
• 4 zones to explore (intro, inn, goblic cave, kingdom)
• Limit world event broadcast to local zone
• Improved player movement sync

What is newly broken:

• Bad sync on MOBs movement

[Windybeard]

This sounds awesome, best of luck with this

[Sanojian]

Oh wow.  Mozilla Browser MMO

This is kinda exactly what I was going for.  And its open-source?   

Should I even continue?  I made really good progress lately too. I had three players last night working and finally squelched a maddening sync-bug.

My game does have a couple of advantages:

• they only support websocket browsers, I support everything
• they seem to have discreet x,y locs (tiles).  I have analog x,y locs
• my game can be run for free (Google App Engine)
• my graphics, sound, and production values are better (this is totally a lie  )

To continue or not to continue...   

[Sanojian]

Okay, so more than a year later after losing hope I am back again.  I scrapped everything and am starting over.  I built this cool (I think) lighting and shadow engine for 2D graphics in HTML5.



The tech will be include Canvas and nodejs on the server.  I am using Liberated Pixel Cup graphics until I find a replacement.

I will overcome this MMO demon!  Once more into the breach, dear friends!

[kamac]

Very nifty shadows. Any details on implementation?

[Sanojian]

Thanks! 

To start I am using out of the box CraftyJS.  I grab the canvas context and then use the globalCompositeOperation to draw the gradient lights and shadows.  I compute the shadows by projecting lines from the light source to the edges of the collision box.  I am just using z-index to decide what to light.

For the darkness I overlay another canvas on top of the first and then make holes in it for every lightsource.

[kamac]

Nifty. (Although it might become slow when there are many objects to cast shadows from on screen, I suppose?

Also, how do you handle (or want to handle) a situation where player doesn't yet see the lightsource by it's origin, but should be already able to see a shadow (that is casted by an offscreen object) and some light (casted by the offscreen light)?

Also, if you have any pre-alpha/alpha/test version up, I'd be happy to check it out.
And another question, any details on server architecture? (Is it JavaScript and run in the browser? Node.js perhaps? Or maybe C++ native app?)

[Sanojian]

So far I haven't seen any slowdown with 10 or so light sources.  I was a bit surprised actually.  I am sure the limits are just around the corner and I will hit them soon.

I made a quick example video here

Also, how do you handle (or want to handle) a situation where player doesn't yet see the lightsource by it's origin, but should be already able to see a shadow (that is casted by an offscreen object) and some light (casted by the offscreen light)?

I haven't hit that testcase yet (I just started this last week).  I will either shrink the "viewable" area of the screen to be smaller than the rendered part or I will just fudge it and the light will pop on when the object comes into the rendered area.

Also, if you have any pre-alpha/alpha/test version up, I'd be happy to check it out.
And another question, any details on server architecture? (Is it JavaScript and run in the browser? Node.js perhaps? Or maybe C++ native app?)

Soon.  The servers are node.js.  I don't have any node hosting yet so its just local.  There is not much to the game yet.  The server just handles character position and speech so far.  The architecture is being designed by another programmer so I will let him speak up here if he feels like it.

[Sanojian]

I started a dev blog for this project where I talk about overcoming some of the technical hurdles creating a small MMO.  I am new to blogging but I tried to make it humorous and informative.

http://tinymmo.blogspot.se/

[Sanojian]

I want this game to have much bigger zones than last time and that requires map tile management.  Here is my blog post discussing my method of managing big maps with large amounts of tiles.

http://tinymmo.blogspot.se/2013/06/tile-maps-of-unusual-size.html