This weekend I answered a question on Jonathan Blow's Twitter with a blog link to onemanmmo.com which resulted in over 200 new visitors to The One Man MMO Project. Statistically, a few of those should have been interested in The Imperial Realm::Miranda and visited its website, but not a single one did.

So the blog got a little facelift today, it matches the game colors, has a little info about the game, a screenshot and a more prominent link.

Let me know if you notice anything broken.

The most commonly asked question about PA4 was "when can I play it?" I want to get more players into the game soon so I did some work this week to get the game ready for them: I disabled a few UI elements for features that aren't finished, and fixed a couple of bugs in the installer. I've also been working on automation so that I can push new builds to the test server with a single command which will be a huge timesaver for me, and get updated builds to players more quickly. To the question at hand though, above all, I want the game to make a good first impression so I'm trying to get the big bugs we've found in PA4 fixed before I invite more people to spend their time on it. I haven't decided exactly when yet, but I have decided that I'll be giving out pairs of Secret Lair Codes so that anyone who gets one can invite a friend to play with.

Pre-Alpha 4 is the first release of The Imperial Realm::Miranda with two of the four planned factions: Empire and Independent, so pick a side and get hunting. If you were following along on Twitter last week, you would have seen renders of a lot of the new units for the Empire faction, but here's a screenshot of all the new models.

New in this Release

• New Empire Faction: 70 Components and 34 Standard Unit Designs.
• Enhanced combat visuals, new weapon and explosion effects.
• Map rendering improved and UI reworked to support mouse drag.
• Shader support for Intel GPUs.
• The Capital City can now be found in the north-east.
• New Sky and clouds.
• New Sell and Repair cursors, plus map scrolling arrows and clicking animations.
• New Field of View setting on the off chance Miranda makes people motion-sick.
• Can now zoom UI with Ctrl-+ and Ctrl--
• 184 Changes & Bugfixes

All of the units' weapons damage and effectiveness statistics, as well as speed and unit health are in and ready for testing. Particle effects for moving units are looking a lot better, no more giant clouds of smoke when units start up. Pathing is generally looking a lot better with the elimination of the crazy steering bugs. Vendors have had several nice usability improvements. I'm hoping you'll feel the game is overall less glitchy.

Known Issues

• If you log out without a construction yard, there is no way to lay your base out the next time you log in. All you can do is sell everything.
• Once created, it is impossible to tell what type of infantry an infantry unit is - even for the owner.
• Most units' special abilities are not implemented.
• Some faster units still need their turning rates tweaked which results in units missing their target then looping back around to park.
• Fence layout is broken. They function as barriers, but it is impossible to make them pretty.
• Occasional GPU hangs on AMD GPUs.
• Some of the new unit models are causing lower framerates and long rubber-band-select times.

I wiped the database again (should be the last time) so testers will have to recreate their account using the Secret Lair Code received previously.

Occasionally Miranda has long frames, one of the big culprits for that is glLinkProgram which takes 50ms per call, sometimes even longer. That's three frames worth just for a single OpenGL call and because of the way Miranda manages materials, it can happen any time. The solution for this in OpenGL 4.1 and above (or if your OpenGL supports ARB_get_program_binary) is glProgramBinary and glGetProgramBinary. These two functions allow you to save and load the compiled version of shaders completely bypassing calls to glLinkProgram. Binary shaders are hardware specific, so they can't be distributed with the game and they have to be discarded if the OpenGL driver is updated, but once they are compiled, Miranda uses the compiled shaders every single time - no more slow calls to glLinkProgram.

The goal this year was to finish the game. While I didn't quite achieve that, Miranda is now very close to the game I have always imagined. A lot of Miranda's art and design content has been finalized and it is a much, much more polished experience than it was a year ago. There is a surprising perceptual change when you finally eliminate all the little glitches you've let slide for so long - the game miraculously transforms from broken to ahhhhhhh.

I opened the year with a major rendering overhaul, dragging framerates on my aging laptop from a low of 50FPS back up to an acceptable 120FPS. After that I realized I was tired of seeing just the salt flats in my screenshots so I added the Scorched Desert biome.



I started work on the preparations for the Friends and Family Pre-Alpha in June, fixing a ton of pathfinding and collision issues, getting Miranda set up on Amazon AWS, adding an automated crash reporter and bug reporting form as well as many more last minute additions. In-game Vendors were also new around this time. In July I added unit caps and work started on the New Combat Model which continued through August and September and culminated in the addition of the set of 65 components and 43 standard unit designs that make up The Independents Faction.



One of my breaks from this huge stretch of unit and game design was the addition of the oh-so-attractive deformable terrain. Also this summer there was a great article on the Miranda at MMOS.com. In October, work started in earnest on the first trailer for the game. The trailer is almost entirely in-game shots so I added a new Capital City, a new sky, many new weapon and explosion effects, a crashed spaceship, Mansion sur Mont, and an in-game film-maker and video recording facility.



My final feature of the year was based on feedback from my third Full Indie Demo night so now when the player scrolls the terrain with the mouse, the mouse cursor changes to red arrows and when he clicks on the terrain, it plays an animation to let them know that the units will start moving soon. While I was at it I added mouse cursors for repair mode and sell mode. I had hoped to complete the Empire's forces before New Years, but there is still a bit of work to do on the art for those units before they will be ready for players to test. Overall this year about a quarter of my time was spent on additions, a quarter on bug fixes and about half on general improvements.

I was much more diligent about blogging this year, I was trying for one post a week and I almost made it. My favourite blog post of the year was The Imperial Realm::Miranda Pre-Alpha - how could it be any other? (Although Full Indie Demo Night was a close second.)

Resolutions for 2016

• Miranda will launch.
• I'll have a month where I don't work evenings.

As for my 2015 resolutions: I bought a D&D starter set and figured out my next project, I did some PR (yesterday!) but didn't end up doing anything with robots.

2015 By The Numbers

254,102 Lines of C++ Code
5,473 Lines in largest file scene.cpp
1,124 C++ files
1,316 Tweets
855 Followers
634 Items in the "what did I do this week" log.
466 UI Files
273 Textures
192 Models
189 Materials
83 Checkins to Mercurial
68 Shaders
44 Blog Posts
31 Facebook Likes
17 Custom Tool programs
8 Server programs

Favourite Devlog Comments

• Demoed the game at Full Indie to about 20 people. I let someone else drive!
• Backup doesn't fit on 32GB stick anymore.
• Started work on Trailer.

Favourite Behind-The-Scenes Screenshot

I demoed Miranda at Full Indie at The Pint Public House tonight. I nearly missed the start of the meet because the parkade I usually park in was undergoing renovations and had only TWO non-reserved parking spots (which were full.) I then tried the mall parking ($16!) then the T&T grocery store parking (1 hour max - no signs to that effect anywhere) then finally I parked on the street for $6. If you were at the meet you got to see the raw video I've put together so far for the trailer - quite a few people loved the idea of Mansion sur Mont. I also had a number of people come up who have seen Miranda before and were impressed with the game's progress. I am quite encouraged that nearly everyone asked about public playtesting.

As is always the case, the first guy who tried the game hit the one big UI bug I've been putting off fixing. Luckily only one other player tonight hit it.

I was pleasantly surprised to have someone come up and correctly guess the provenance of the game's name. I shared a couple of the easter eggs in the game with him.

As always, I learned some valuable things from watching people play. The first is that I'm going to need to add the Starcraft style (left mouse selects, right mouse attacks) controls in addition to the classic Command & Conquer style (left mouse selects, left mouse attacks) that I prefer. For about half the people who took the helm, their expectation was Starcraft controls.

The other big lesson is that people click a lot more than me when moving units. Several people would click on a unit, then start clicking all over the terrain. What I realized after watching a couple of people do this is that visual and audible feedback for actions is needed when the action doesn't happen right away. With Miranda's pathfinding, units wait half a second after the player selects the destination before the unit starts to move. This gives the pathfinder time to calculate the path and send it to other players so that units start moving at the same time for all players. Other RTS games have a similar delay, but when the player clicks on the destination the game says "moving out" and plays a little click animation. The plan has always been to have that feedback, but with that not in yet, players would repeatedly tell the unit to move to different locations before it had a chance to start moving, resulting in no movement at all. I'll be adding that audible and visual feedback sooner rather than later.

Thanks to everyone who came out to tell me what they think of the game so far. Your help is invaluable.

One of the 5 new weapon effects: Cannons, missiles, lasers, lightning and flamethrower. Moving on to explosions.

I have written 200 blog posts on the development of The Imperial Realm::Miranda. It turns out developing an MMO is a complicated business!

When I started the blog back in November 2009, I wrote:

I'll state up front, I realize how ambitious a project an MMO is. Even a small MMO is huge, I get that. But I still want to try. Chances are at the end of my time on the project, there will be nothing resembling a game. I think the value here will be what is learned along the way.

While what I have learned has been amazing and valuable, the biggest benefit has been the opportunity to watch my little girl grow up in person instead of over Skype. Developing Miranda has been thousands of late nights, constant budget analysis, a small fortune, stress - and a much closer relationship with my wife, who has been amazing in her support of Miranda. Miranda is a game. We built her and I am thrilled.

I kind of wish I had a big announcement to make to celebrate this milestone, but if you've been following along you already know that I have a trailer and some big game features still to finish... Pre-Alpha Release 4 will be out by New Years. There, nailed it.

I'm finally feeling pretty good again after a week of having a cold, but progress was kind of sporadic this week. Friday was great fun though! While I have been working on other parts of the trailer I have been thinking a lot about why combat isn't as exciting as I want it to be. I had a bunch of ideas and that all started to come together Friday afternoon.

A couple of my biggest, baddest bugs fell this week. There are three different types of bugs programmers face: there are bugs where you can easily reproduce the bug and know right away what the problem is, those are pretty easy to fix. There are bugs where you know what the problem is but fixing it is a huge effort (usually because it is in a system made of dozens of massively complicated components.) Lastly there are the really bad bugs, the ones that aren't obvious and can't easily be reproduced. Those ones can take days.

I once worked on a crash bug in a Need for Speed game that took about an hour to set up, and it took 59 attempts to get the bug to reproduce one time. It turned out that the bug depended on the WiFi dropping a UDP packet. When that happened, two packets arrived in the reverse order they normally arrived in, generating a crash.