Blog for Thomas \"Man of Ice\" Lund at GarageGames.com

Super simple script based name generator + Air Ace bomber

Thomas \"Man of Ice\" Lund — 2007-01-22T22:48:30+00:00

I wrote this yesterday as I was in need of AI pilot names for our game Air Ace Online www.airaceonline.com

Its so super simple that it hurts, but works fine for my usage. So why not share it with you.

The basis is to use a list of first names and combine randomly with a list of surnames.

For Air Ace in version 1 we have British and German pilots, so I fetched a little list of those.

To get some good names I went to
en.wikipedia.org/wiki/Surnames
and
en.wikipedia.org/wiki/First_name

In the bottom they have lists for common names for various countries.

Then I fixed this little code snippet and thats it. Enjoy!

// Random Name Generator
// Made by Thomas Lund
//
// Creates a random name from name lists
// per team/country

function NameGenerator::getRandomTeamName(%team)
{
	return NameGenerator::getFirstName(%team) SPC NameGenerator::getSurName(%team);		
}

$NameDB::UK_first = "James Benjamin Joshua Ryan John Patrick Samuel William Jack Thomas Daniel Matthew Jacob Ethan Ronald Michael Richard Jonathan Joseph Oliver Harry Charlie Sean Jack Conor Adam James Daniel Luke Michael Aaron Jack Matthew Adam James Ryan Joshua Conor Ben Daniel Dylan Lewis Jack Callum James Ryan Cameron Kyle Jamie Daniel Matthew Peter";
$NameDB::GE_first = "Alexander Maximilian Leon Lukas Lucas Luca Paul Jonas Felix Tim David Jens Joachim Johan Heinrich Hugo Hubert Xaver Wilmar Wolfgang Waldmar Uwe Udo Ulrich Theo Thorsten Sven Sigfried Rubens Rudi Niklas Otto Max Manfred Leopold Guido Gunther Friedrich Fritz Frank Frans Eckard Erich Diedrich Dietrich Diether Bernard Bodo Aldo Arnold August Arne";

function NameGenerator::getFirstName(%team)
{
	if (%team $= "UK") {
		%index = getRandom(0, getWordCount($NameDB::UK_first)-1);
		return getWord($NameDB::UK_first, %index);
		
	} else if (%team $= "GE") {
		%index = getRandom(0, getWordCount($NameDB::GE_first)-1);
		return getWord($NameDB::GE_first, %index);

	} else {
		error("Get first name - unknown team " @ %team);
		return "Unknown";
	}
}

$NameDB::UK_sur = "Anderson Brown Byrne Campbell Clark Davies Doyle Edwards Evans Gallagher Green Hall Harris Hughes Jackson Johnson Jones Kelly Kennedy Lewis Lynch Macdonald Martin McCarthy Mitchell Moore Morison Morris Murphy Murray O'Brien O'Connor O'Doherty O'Neill O'Reilly O'Sullivan Patel Paterson Quinn Reid Roberts Ross Ryan Scott Smith Stewart Taylor Thomas Thomson Walker Walsh Watson White Williams Wilson Wright Young";
$NameDB::GE_sur = "Altenstein Altmann Bach Becker Berger Brand Denke Dunkel Eisner Engels Feuerbach Fischer Fuchs Geissler Grass Hahn Hankel Heilmann Hoffmann Huber Hueber Jost Kirchner Klein Krause List Lorenz Mann Meyer Moltke Mueller Obenauer Reich Reuter Sachs Schanze Schindler Schmidt Schneider Schulze Simmel Sommer Spitz Thiele Volkmann Wagner Weber";
function NameGenerator::getSurName(%team)
{
	if (%team $= "UK") {
		%index = getRandom(0, getWordCount($NameDB::UK_sur)-1);
		return getWord($NameDB::UK_sur, %index);
		
	} else if (%team $= "GE") {
		%index = getRandom(0, getWordCount($NameDB::GE_sur)-1);
		return getWord($NameDB::GE_sur, %index);

	} else {
		error("Get first name - unknown team " @ %team);
		return "Unknown";
	}
}

Edit - damn. Almost forgot rule #1. No blog without images.

So hopefully Phil didnt post this one yet.

We recently got this baby into the engine - B17. First a render shot, and then a shot from ingame (glass material still being worked upon)

Edit 2

Ahh what the hell. Here is a little movie I made yesterday too. The first shot at the bomber AI controlled:

airaceonline.arcadersplanet.com/movies/AIBomberTest1.wmv

I was going to save that one for a different blog, but in the end I dont have much time currently. So feel free to read more about the AI here:

www.airacepilots.com/viewtopic.php?t=244

Air Ace ready for alpha testing

Thomas \"Man of Ice\" Lund — 2006-09-28T21:16:44+00:00

Ahhhh - FINALLY.

Definitions of alpha test are many, but we finally have a stable and fun game that doesnt crash anymore. With a basic dogfighting mission that looks OK'ish, I'll try to convince Phil that we should open up for alpha tests. Check our forums for latest news and the important "why should we pick you as a tester" post in the announcement section.

Forums for anyone not signed up yet: airaceonline.arcadersplanet.com/

After merging with MS4 we still have some issues, but we managed to fix a lot of bugs the last few weeks - both in TSE and in Air Ace. So here we are 3 days before leaving with a game thats ready to be burned onto a CD and taken to Eugene.

But before going to the US, I wanted to take the time to post a new blog.

No blog without a screenie:

Well - we still have tons of things to fix, optimize, test, add, remove etc. Especially our menus are crap, but we got a nice concept going with some layers. I like it, and hopefully so will everyone else.

Air Ace is now fully playable with no known crash bugs. Particle effects have been added, the alpha test level is done and in general we are happy campers (OK - I am at least happy. Phil isnt happy, as he is "homeless" as you might have read)

So without much further words a link to an ingame movie!

airaceonline.arcadersplanet.com/movies/AlphaTestMovie.wmv

Obviously there are some visual bugs still. In the upgrade process we lost the material mappings on the plane and the canopy cube map. Dunno what happened.

The big thing is Adam deGrandis fantastic effects. We are sponsoring his explosion pack, and he is also helping with level design and other things. Very cool, and fantastic work on the explosions. Check out how the spawned effect is determined by what it hits. Water spray when hitting water, dirt on the ground etc. Cool!

But heck! Time to ramp up internal testing and finding all glitches.

One of the hotter things we will work on after returning are the new planes we got in the pipeline. We found another great guy - Jesse - who has a few nifty planes that we bought a license for.

Read the entire post here: airaceonline.arcadersplanet.com/viewtopic.php?t=94

And check out this nice P-38 - comming to Air Ace very very soon

(Yes - this is a slight rewrite of a forum post on our forums - bite me ;-) )

Regarding community weekend - Steelhead next Thursday at 7 pm!! Say no more. And dont drive, as we will drink all night and have great fun ;-) Eugene beware.

Adventures under a blood red sky (clouds)

Thomas \"Man of Ice\" Lund — 2006-09-05T11:23:37+00:00

I'm one of those guys that do not jump onto the big next tech craze without a need, and thus I'm usually a few steps behind. On the other hand, it has the nice effect of everyone else beating their heads against non-proven tech + most usually die before I waste precious development time on it.

One of those things that I never got into was - believe it or not - shaders. Heck - I've seen the demos around and thought they looked cool. Even downloaded and started up Render Monkey once. But no development of any kind.

Even with Air Ace being developed using TSE, the need for shaders hasnt come up. We mostly use TSE for the large terrain.

As posted earlier, I have been working on/off on some basic cloud rendering. Last week I coded a more generic "cloud area" object, that spawns several clouds inside it and moves them across the sky over time. When they hit the end of the defined area, they are respawned in the other end again.

But the cloouds themselves do still not look like real clouds. And the thought process of how to render nice clouds the way I do them started nagging in my mind to a point where I started to dream about it (after dozens desperate attempts of figuring out the inner workings of the GFX system and blending.

Thus came the need for shaders. From what I read on the net, I could see it being handled much much easier in a shader than in engine code. At least I had a chance to read various external books/websites on shader programming, instead of guessing about texture modulation arguments and whatever :-D Never been much of a low level guy, and this is a bit to low level for me fiddling with gfx card registry settings.

I thus took my cloud code, turned off blending and added hooks for a shader into the renderObject code. One day later, this is what I had:

Impressive!!! My first shader in action - hehe. All it does is pass through the vertex shader and adds a red color in the pixel shader. But hey - it works! From now on, the task is to blend the individual cloud puff textures into a coherent "cloud" mix and getting sunlight to brighten the top side and darken the bottom side of the cloud.

My road to this step didnt go without some smaller problems - which I would like to share with anyone else trying to do this. I cant post code here in a public place, so here are some more general observations.

* When one tried to hook up shaders (not materials) directly into the engine code, one has to send over all the "magic variables" yourself for the shader to get them.
I took some example shaders from RenderMonkey to color pixels red, and then found great this article: http://www.garagegames.com/index.php?sec=mg&mod=resource&page=view&qid=7595

It made a world of a difference, as I soon discovered the problem of hooking up those magic VC_WORLD_PROJ, VC_EYE_POS variables etc. As the referenced article says, hooking up the various world/eye/projection space isnt necessarily trivial.

It took me the most of a day from the "why the hell doesnt anything render after copying the shaders over" to finding the article and to having found the correct way to initialize the variables correctly.

The best way to do this is through the shaders/shdrConst.h header variables. One can set the registry positions manually, but it soon gets very muddy.

I seriously had hoped things to be initialized by TSE itself (more on that later), leaving some registry areas for user defined data - but everything is open for doing it as you please - totally ignoring the defined names. So what is in registry position C0 in one shader isnt necessarily the same as in another. And the engine code doesnt use the shaderConst header very much, leaving even more confusion to the n00b shader coder :-)

* There is a difference between materials and custom materials
When I later experimented a bit with materials, I found that the non-custom ones are damn easy to use. Kudos to whoever made them at GG. Making custom materials brings one back to the above mentioned problems of having to hook up everything yourself again. Materials make things a bit more easy to use though, so I would suggest to anyone using materials instead of just the shaders if possible. Allowing you to easy making multiple passes and defining textures to be used.

* Start developing using existing materials and existing shaders
Might be obvious, but then again. Attack things on one front at a time. Trying to both write custom shaders, custom materials, code to hook up shaders to your object etc makes your head spin more than once. Take it one step at a time, and solve the issues you run into before moving on to the next level. I kept going back and forth between my own and existing working shaders/materials/objects to see that I did things right.

* Last hint - reuse
Once you got the code down for hooking up one material/shader, save that code and reuse it! The process is the same over and over again. Again a fairly obvious one. I had to many things going on at the same time, so at one point I had different implementations of the same mechanism - and couldnt keep my head straight on which one was working right or not.

At one point I was sick and tired of the cloud code, so I started a second feature. Air Ace has a fairly large environment, but it does have a limit. So we have to have a way of restricting a player to not go outside a defined area. I remembered some old resources for TGE that drew a forcefield, and thats what I wanted - draw some "wall" when the player got close to the mission area boundary. I decided to allow the player to fly through the wall, and then have a certain amount of time to get back again.

Time to reuse my learned knowledge about shaders/materials :-)

I used my 1337 coders art Photoshop skills to make the above images. I wanted to have one material on one side of the wall, and another material with a different texture+shader on the other.

After drawing up some vertex buffers with triangles and matching up texture coordinates, it was time to look at the shaders again.

The inner side of the box should not be visible during normal play. And when you got to close, the wall should gradually appear with a "dont try to fly through here" texture. Thus the shader takes a alpha value from the engine based on plane distance to the nearest wall.

On the other side, I need to give the player a chance to return to the mission area. The wall should thus always be in view - but with a different "come back here" texture.

After a little fiddling with things, I finally got it all working. Check out the video here

airaceonline.arcadersplanet.com/movies/Mission_boundary.wmv

The main trouble I had here was the way I had set up the vertexbuffer. I used the same buffer to hold the triangles for the "inside" and the "outside" triangles, and then fed the primitive drawing route a start position and a number of prims to draw. But it didnt work. Dunno if thats a bug, or just me doing something wrong. In the end I split the vertex buffer into 2, and draw all the prims inside - one buffer for each material.

I'm still not entirely satisfied, but it starts looking OK. I would like to fix the shaders, so only a small area of the wall is visible. E.g. having a radius around the player position that is rendered - everything else invisible. That is a task for later for me - a little beyond my shader skills for now :-)

Air Ace is starting to look really good now. We are getting more and more features completed, and are working on the looks and initial impressions like better plane models, great looking menus and such. The new menu system is for a plan of its own. Look forward to seeing what Phil has cooked up! Its ultra cool to look at, and everyone who has seen it go "wow!".

Until next time!

Air Ace Update Server + Launcher

Thomas \"Man of Ice\" Lund — 2006-06-17T13:31:21+00:00

~2 years ago I had a student do a project with/for me. I was in need of a combined launcher+updater client, and we worked together on that 2-3 month. He wrote a project, got good grades and everyone was happy. The code wasnt 110% solid and finished, but it was good none the less.

Well - as things happen, the code never really got used and it went into my SVN "for later".

Since we are approaching alpha for Air Ace very fast, it was time to pull out this old stuff and brush it up. Yesterday I finally got the last pieces running again - and what better excuse but to write a blog and post some screenies :-)

The requirements for the system were (from top of my mind)

Client:
* "platform independent"
* C++
* display news
* update the game
* launch the game

The update process had to be fairly robust and secure. Encryption, CRC checks and rollback on any failure

For the client side we decided to use wxWindows and C++, so it would be easy to get it running on multiple platforms. The client talks to several webservices on the server to fetch news, check for updates, grabbing updates etc.

Server:
* "platform independent"
* J2EE
* support for unlimited games from same server
* webclient for uploading updates/writing news/configuration etc

All my backend stuff is coded in Java (Java2 Enterprise Edition), as its (in my mind) the best scalable language to use for server side. I've released several boxed non-game domain products with it and it just runs!. Having application servers like JBoss makes it _very_ easy to support clustering, sessions, database independence, transactions and webservices with minimal code.

The idea was an architecture like this:

The administrative client was easily made as a web client included in the application server, so thats what we ended up with.

Various screenshots:

Boring ones first - the running server:

Not much to see :-)

The administrative webclient (still under graphical development - but fully working):

And lastly the progression from the "student art" client to the almost done Air Ace layout. Still to be done is the transparent wxHTML widget

and finally a fully connected client up against the backend

Air Ace Clouds!

Thomas \"Man of Ice\" Lund — 2006-03-26T10:41:17+00:00

Ahhhh

Yesterday I was playing Heroes of the Pacific for inspiration, and noticed how much clouds mean in an air combat game.

So what better than to sit down and code a cloud renderer object for Air Ace?

(The observant reader would already now guess that I succeded - why else post a new blog) :-)

So without much further talk here are some screenies!!

One of the first attempts where "something" appeared on screen

Now it can scale the cloud puff billboards

And finally a nicer texture and something that actually looks like a cloud!!

But screenshots dont make this any justice.

Check out a little movie I made. 2 large clouds with approx 40 billboards in each of them.

Movie link (~7 MB)

Single cloud movie - fluffy fluffy
Movie link (~2 MB)

The clouds are very network efficient (single position for the entire cloud combined with a random seed + info on sizes of cloud+puffs).

I'm still fighting with getting texture lighting modified, so that the puffs on the underside are darker than the ones towards the sun. The code is in, but changing color has no effect for some reason.

Still room for improvement, but it looks damn cool!

"I'll be back" and here I am - Air Ace update too!

Thomas \"Man of Ice\" Lund — 2006-03-05T10:57:16+00:00

Looong time ago (or so it seems) I ran into the stress wall, and had to drop everything to relax. Well - I'm up on my feet again. No need to dwelve too much into that, but the most important thing is that I'm back and alive.

One of the things I promised myself is that I do not want to work "alone" anymore. Being indie is hard enough on your own, and having a soul mate is essential for survival.

Mind you - I've been working with others on my projects. But none that really had the same indie mentality, and none that were coders/producers kind of people. Sparring about problems, issues and simply throwing a problem to your partner saying "I cant find the bug - help me" instead of having to do everything yourself. Always.

For the last 2 IGCs I've hung out with Phil Carlisle a lot, and we had a lot of discussions about everything. Well - we finally decided that its time to partner up! Phil is definitely a cool guy, and we share a lot of the same problems with artists, contractors as well as the same views on life.

So for the last month, I've been secretly working with Phil on getting Air Ace done.

My work so far has been to build a backend system - master server, auth server, community server. This work is now "complete" (is there ever such a thing) to a point where it can support initial alpha tests of Air Ace.

Main features:
* master server functionality compatible with stock TGE/TSE
* authentication of accounts
* pilots (the name for the player avatar)
* squadrons (the name for clans)
* buddy lists
* saving stats
* security mechanism to control game servers

System is build a bit like Battlefield 2, where you have "official" servers that are allowed to save statistics, while unofficial servers can (in this system) be allowed or disallowed. Disallowed servers simpy are not allowed to start up, and clients cannot join them.

From a technical side, the backend server system is a multithreaded java server that uses Hibernate to connect to any JDBC compatible database (MySQL, PostgreSQL, MS SQL, Oracle etc.). It has full transactional database support, so any error rolls back changes made. Meaning that its almost impossible to mess up the database due to errors.

On the client/game server side it uses the existing mechanism for talking with master servers. Its based on async UDP communication and script callbacks. Its 100% compatible with stock TGE/TSE, and requires very few engine changes.

And without much further talk, here are some screenies!

(Warning - I'm not a GUI artist. So its all coders art!!! Still to find someone who can make this all nice and shiney)

First up is the pilot page.

On the left is the buddy list, where your buddys show up. You can see their online status (offline, online, playing). If anyone is playing, you can simply press a button and you connect to the same server as your buddy!!

In the middle you change your pilot name, and see your current squadron membership in form of the insignia and name (here my current squadron is "Base Rapers").

On the right I can see all the squadrons I am member of, and can change which one is my currently active one.

Next screenie is from the squadron page.

Left is a list of all squadrons. Right we have the name, tag and insignia. In the bottom there is a list of the squadron members, as well as their online status.

Squadron leaders can change the details of a squadron as well as invite/kick members from here.

Last image this time is a screenie of the server. Instead of taking a screenshot of the Linux console (a bit boring), here its running inside IDEA. In the bottom the server is showing what requests it has processed.

Not much more to tell about it now. Its very very very easy to expand with new request/response commands, and picking Java as the backend language has paid of 1000 times already. Ultra stable, threaded, simple to code and fantastic database abstration through Hibernate. All exceptions are caught and will (very soon) be emailed to me for bugtracking. Since Java doesnt have problems with buffer overflows and similar, its also very safe.

Initial support for webpages are also under construction. They are so ugly, that I didnt want to post any screenies of it. But each player will have his personal webpage for tracking statistics and managing his profile from a browser (in addition to the ingame GUIs). Administration of the backend server is also done through a webpage.

With the backend "done" for now, I'm moving into helping with the actual game. Soon we will have it ready for internal alpha testing, and its shaping up nicely!!

We also have some announcements very soon that are totally awesome. Watch Phils plans for more info!

Thats it for now!!

-----

Oh - on a totally unrelated note. My project until now was Steel Sentinels - an arcade ship game. Its been mothballed for now, but I'm using art and code bits in Air Ace. I've promised myself to only work on one game/project at a time. And Air Ace has a much more promising future. But who knows what will happen ;-)

My Steel Sentinels partner Brian Wells has also gone MIA. If he is reading this or anyone knows where he is - please buzz me. I've been trying to reach him by any means, but he has gone totally offline since early December!! Where are you!!!???!?!??!

Going down....

Thomas \"Man of Ice\" Lund — 2005-12-17T13:48:49+00:00

(Not your usual cheerful post and quite personal - so skip if you dont like those)

Well - it finally happened to me.

Having too many project and too little time to actually do them all, my brain and body said stop. I'm demotivated, no energy and just sit in front of the computer without getting anything done.

For quite some time, my consulting job has taken its toll. I'm currently a project manager for Denmarks largest switch in outsourcing partner for IT operations, and the last 2-3 month have been really really rough.

Trying to do game development next to that beast of a contract, as well as being a father, husband and generally a social being has not been compatible.

Game dev has for me always been what I wanted. So thats why I've pushed hard to do it. Its not for the money - heck. I earn more on regular contracting in 1 month than most do in a year. But I enjoy it too much to let it go. Having tons of ideas for games, packs, improvements.

But it was just too much for me. A few weeks ago I came to a point in my life where I have to decide what to do. Its all about focus, and I have not been very focused. Trying to do everything at once without pursuing a single (or two) goals.

So lesson #1:
Focus on a few goals, get them done and out of the way. Move to next project.

Once my body and mind said stop, it was time to listen to it. I've been here before in the good ole .com days, where we worked 16 hours a day / 7 days a week. I know when I have to throw the towel and step back. And I've seen what happens to people when they do not stop, but cross the fine line. A handful of my friend have tumbled from great careers to sitting in a rocking chair for 2 years unable to work more than 10 minutes at a time.

Lesson #2:
If you feel very low on energy, you snap at your loved ones, unable to sleep because you dream about work - STOP!! It is not worth it. No one will applaud you for burning out.

What works for me to get out of this situation is to stop doing what was stressing me, and do something completely different for a while. As I didnt have the luxury of dumping my contract, I had to drop game dev. :-(

Its really unfortunate, because I almost have a game ready, I have several 80-90% done packs, and Phil Carlisle and me had just started work on a few prototypes. I'm really really sorry Phil - I'll get back to you soon. I promise.

I simply had to make a clean cut from one day to the next one. Not answering emails, not taking phone calls, not reading/responding to forum posts.

So how will I move ahead from here?

For one just writing this is part of getting back. I've made up my mind, that I'm not dumping game dev (which was one of the options). But I'll most likely focus on joining up with someone more closely - being part of teams of peers is one of the things I've been missing. Too many tasks, too few ppl taking ownership and getting some of them done. Always ends up on my own shoulders. *sigh*

I'll either do real games or concentrate on code packs. During my time with GG, I've created quite a lot of half baked stuff that could use an overhaul and brought up on a commercial quality level (camera pack for example). Doing code packs could be one way to bring various smaller projects to market, but the market is damn small and from a commercial point of view not sane to do.

Another thing that I'm starting to realize is, that I need a non-computer based hobby. I spend way too much time at the computer, resulting in not being productive. I can sit an hour or so, just browsing websites, reading stupid stuff and not being able to get going. Having focus on "when at computer - work!" will help. Meaning I need to do something else - at the moment I'm trying to do some work on my car. Putting in a sports exhaust system in my Porsche, doing some aluminium parts mods in the cabin and in the spring going to race it on a nearby track. I've also always wanted to have a rc helicopter. That might be some other option.

Well. Enough from me now. I'll be back some day in the near future. And I hope ppl havent given up on me!

No plan without a pic. So here is a screenie of my battleship game as it looked when I stopped doing coding on it. It has been turned into a arcade shooter with extremely easy controls. I _will_ finish it. Just a matter of when and how.

Been offline - status

Thomas \"Man of Ice\" Lund — 2005-07-23T15:42:18+00:00

Well - first off an apology for all ppl trying to email me without getting answers. I'll catch up with you!

This has been my first major "offline" time from the GG community, and its been really hard. My fingers are itching hard to get back into game dev for real, and now its time.

So what happened?

2 things - contract work and moving

Contracting
Since start of this year I've had a full time contract work in non-game dev that is paying major $$$. Its a real evil devil.

As most of you know who have been there, working 40-50 hours/week for someone else and then doing game dev on the side is really hard to pull off. I do have a family with kids, and they are (rightfully) demanding more than 30 minutes a day together with me.

It did go well until approx April. Doing some unannounced game dev work on the side.Then I was contacted to do a mod for TGE, so a group could use it for machinima. They have done some really cool commercial machinima with it, and I hope that I soon can show some screenshots from it.

This ate up all my time, and the game project came to a grinding halt.

Having just a small team with 2 full time + 2 contractors it does take quite some work to coordinate and keeping everything running smoothly. My partner in crime also got married, and suddenly the team was left with none in charge and pushing ahead.

We are now at a standstill, and have to push hard to get running again. And the next few month the family will have to pay the price, as my non-game dev contract just got extended the rest of the year.

While it would be very easy just to take the money I've earned until now and do full time indie game dev the next year, its very hard to say no to a big sum of $$. And one has to grab the jobs while you can get them. Suddenly its over, and cash has to flow in from somewhere else. Having a larger war chest will make everything easier down the road.

Moving

Soon after I ended my machinima mod, we (my family) decided to sell our house in Denmark and move to Sweden - just 30-40 km away, but a hell of a difference for us economically (cheaper everything). This did eat the rest of any sparetime I ever had, and also has left me without any broadband until September. It simply doesnt exist yet in the little village I now live in. And running Internet over GPRS (3rd gen mobile network) is DAMN expensive. I currently pay $1/mb that I down/upload. So its text only for the next 2 month, and then I can get 10 mbit/s fiber into my house!! What a difference.

Moving to Sweden has until now been the right move. Some time in the future I wont have a full time contracting job to pay the bills, and at that point in time I'll be very happy that we did this. Because as someone famous said - right size your life when doing indie game dev. And by moving jost 30-40 km we effectively cut our monthly living expenses by 50% and still have the same living standard (if not better).

Near future

Besides settling in at our new home, I'll have to get time set aside to do start the game again. The plan earlier this year was a fully finished game for IGC. That is now impossible, but I hope that we can have the entire core game done with 1-2 missions finished.

As to commit myself a bit, I'll post a fast (old) screenshot from the game. Its a 3rd person battleship shooter game. Single player mostly with just a pinch of multiplayer (for version 1).

A proper announcement will come soon, so this is it for now!

New toy
Some of my warchest recently got "invested"...... into my new toy!

2000 911 Carrera 4 - and its the best money ever spend on myself :-D

Resource Updates - Advanced Camera and Flash GUI

Thomas \"Man of Ice\" Lund — 2005-02-02T15:55:56+00:00

Ahhhh - finally something done that seems to work

www.garagegames.com/index.php?sec=mg&mod=resource&page=view&qid=5471

The advanced camera resource has been updated and is now better than ever.

There is a new orbit camera mode with mouse movement support when run in single player game, a static camera mode, better collision support and a vertical freedom mode where the 3rd person camera moves up/down with the head of the player.

Thanks to all who contributed ang gave feedback.

It also marks the end of me developing engine technology for my battleship game (except that pesky flash gui - more later). From now on its a matter of implementing the levels and gameplay only. I bet more will surface later, but for now its all done.

The flash gui component I was working on has not been finished. GameSWF compiles, the gui component seems to work and it runs the movie, but there is something wrong with the resolution. For those who didnt see my cry for help, there is a nifty screenshot here below

Its trying to render a 11000 x 8000 pixel movie if you are wondering what is going on.

I have given up for now (my OenGL skills are just not good enough), and pushed the job out on some of the other community member to help. Hopefully soon we have a result. Brian Bzztbomb has gotten it to render in the right size, but that was with lots of hacks.

Hopefully progress soon, and I'll officially buy a lot of beers @ next IGC to the one that gets it running.

Thats it for now. Hopefully the next .plan can have content on my game projects. There should be basis for some nice screenshots and ingame videos soon

Friday Jan 28 13:17

Thomas \"Man of Ice\" Lund — 2005-01-28T13:17:08+00:00

First plan of the year. Flash GUI component, adv. camera update, having interns, website work and !!! game dev

Damn - already 1/12th of the year is gone, and I dont even feel I've been doing all that I planned.

For the moment I'm working on all the items below

== Flash Gui Component ==
I was going to work a bit on an MNG component for large animations, but then someone (Thx Nic) pointed me in the direction of GameSWF.

So I've it compiling and stripped of SDL calls. Now onto integrating it as a GUI component for TGE. Unless I run my head into a showstopper, that will be a resource soon. But compiling GameSWF is not a breeze on Windows - took me more than planned at least.

I want to use the flash files for ingame animation sequences for my battleship game (see later)

== Adv Camera ==
For a few weeks I've been planning to update the adv. camera resource. Both with contributions and with my own code. Once the flash component is done (or dead) I'll start on that. Nothing major planned, but a lot of cleanup and working on interpolation. About time :-)

== Website ==
We finally managed to get a website up running. Phew. After several month of poking at it, we went live a few days ago. There is still a lot of smaller details to get fixed, but overall I think the result is quite cool. Its 100% database driven and with an easy home cooked CMS for everyone to write content. The URL is www.questless.com - criticism is always appreciated - and yes I know there are a few gfx items that aren't done.

Content will be added soon for real, and I plan to write a TGE related article each month or so. We will try to do some marketing of ourselves and our games/products periodically - lesson learned at the sessions of IGC 2004.

== Game Dev ==
Oh - my favorite.

The top secret battleship game is shaping up. We have contracted someone to develop a robust and fully featured naval vessel code class, and that is under development at the moment. Meanwhile we are working hard to implement tutorials and the actual campaigns.

The game is a 3rd person battleship shooter game using TGE (what else), and we have high hopes of getting this done and shipped before summertime. Screenshots, movies and marketing will come your direction soon :-)

Some of the big unknowns are still the balance between single player vs multiplayer and realism vs arcade. We want to have both, but the mix is hard to determine up front. So it will come down to playtesting and tweaking as soon as possible to adjust everything for fun gameplay.

We are a bit scared of the multiplayer aspect. ThinkTanks, Zap and Lore are showing clearly that getting players to play multiplayer games isnt just a matter of putting up servers and releasing a great game. Thats why I think we will put way more singleplayer content in than multiplayer, and then try some alternative approaches to the multiplayer part.

One idea is to have something similar to Tribes 2 duel mod. Thus I'm going to have an intern code up a "duel mod" server, that can pair up gamers who wish to duel someone. The idea is to press a button saying "find someone to duel" and within 15-20 seconds it should have paired up 2 players. If there are none, then it should create a bot on a server to duel against. Auth server and master server are already done for this game, so it should be quite easy to pair up players with an additional server process.

Realism vs. arcade is much harder. One one side you want realism - wakes after the ship, reloading guns, limited firing arcs, subs, planes, aircraft carriers - phew. All the stuff you keep dreaming about. But as everyone knows realism != fun.

On the other hand going too much arcade and doing a "1942" kind of arcade game doesnt feel right either. Then again people always like to blow up stuff with big guns - and the guns on our ships are big :-)

I recently read a nice little article on the net about Halo and playtesting. They had given the aliens very low hitpoints and you could blow them to crap in no time. Great fun, but people felt that the AI was too easy. Then they changed the number of hitpoints for an alien, and suddenly all playtesters reported that they thought the AI was much better and harder to beat! And that only by adding more hitpoints to the enemy.

So I think this could apply to "realism" vs "arcade" too in our game. By having many hitpoints you can have very long and interesting tactical battles with only 2-3 enemy ships - on the other hand we would need tons of ships if you could blow them out of the water in no time. Thus this could make the game "feel" more realistic, the AI better and we do not have to put in as much content. A winner for all.

I guess playtesting will show if our initial approch is the right one. The earlier we reach that point, the better.

Thats it for now. Lot of work to be done, and I still have my game installer/updater + ActiveX installer products to wrap up too. They have been untouched for too long now waiting for feedback. Might as well try to pick someone to test them and get them up to product level in quality.