FIFE - a next generation Fallout-like engine

[mvBarracuda]

To bring up the most important fact at the beginning: the FIFE team will finally release the 2007.2 milestone to the public at Wednesday, the 10th of October, 2007. Over five months have passed since the 2007.1 milestone had been released at the end of April and a lot has changed since then. Now that the most important changes are in place and work we feel that it's time for another release to give the community a sign into which direction the project is heading.

 

Let's summarize the most important changes for the 2007.2 release and have a (small) look into the future of FIFE as well.

 

After having utilized Lua5 for the scripting side of FIFE for over 18 months we agreed that it would be worth taking a look at SWIG again. SWIG was promising as it offered a way to use our engine with a number of scripting languages; furthermore we could get rid of the semi-manual lunar.h bindings approach this way as well. We ran into some hurdles while trying to integrate SWIG into FIFE but in the end it worked out quite well. Now we had the ability to wrap up the engine into python instead of the old way of embedding the Lua library into FIFE.

 

The so-called metamodel branch was an undertaking to seriously redesign major parts of the engine. We decided to go for an MVC (model - view - controller) architecture pattern for FIFE and refactored the engine modules accordingly. Another aim was to remove unneeded bi-directional module dependencies to have a clean module hierarchy:

http://wiki.fifengine.de/index.php?title=Engine_Core

 

The Zero-Projekt team agreed to work with the FIFE developers on an example game for future releases together. This way we can give game creators a starting point and show what can be done with the engine. Furthermore working with the Zero team directly together brings up a number of advantages for FIFE:

* We won't need to worry about producing content for the example game.

* We're in contact with game creators so we get feedback what works well and what aspects should be improved.

* We can test and adjust the FIFE API while we work on the example game. This wouldn't be possible without building an actual game on top of the engine.

 

A first snapshot of the example game will come bundled with the 2007.2 release. Don't expect too much: it is a work in progress project and especially the first versions will naturally lack a lot of polish!

 

After the release of the 2007.2 milestone we plan to finally start working on the editor tools for FIFE again. We tried to do so several times before but everytime something went wrong: editor programmers disappeared and later we did find several engine issues that prevented building an useful editor tool on top of the engine.

 

Now that the engine is in pretty good shape after the metamodel refactoring, we feel that working on the editor tools will bring up better results this time. Jwt, who also wrote the new python-based maploaders for the 2007.2 release, does currently investigate how we could create python-based mapsavers. This is meant to be the first step for the editor tools.

 

Anxs & prock are currently redesigning the FIFE audio module in a separate branch. If everything works out as planned we might be able to ship the new improved audio module already with the 2007.3 release. If you want to read some detailed information about the audio module redesign, head over to the wiki:

http://wiki.fifengine.de/index.php?title=A...n_Documentation

 

Now that we're heading to the finish line we can need every helping hand to test the upcoming release. Furthermore help with packaging would be appreciated as well. Therefore we invite every interested community member to join our project IRC channel to lend us a hand. We hope to see some of you there. If you're too busy, it would be great if you could be around at the release day to celebrate the new milestone with us at least )

http://wiki.fifengine.de/index.php?title=IRC

 

And finally after all the text, some new eye-candy as well. This is a screenshot of the product we're working on together with the Zero-Projekt team. This content won't ship with the upcoming 2007.2 release and is work in progress.

[mvBarracuda]

We've made it! The 2007.2 release is on the server and we're all happy that it shipped in time.

 

Download

* Source package (Linux, Macintosh): http://downloads.sourceforge.net/fife/FIFE...7.2_src.tar.bz2

* Win32 binaries: http://downloads.sourceforge.net/fife/FIFE_2007.2_win32.exe

 

Screenshot

 

Important changes since the 2007.1 release

* Removed the rather difficult to maintain lunar.h bindings in favour of SWIG.

* Replaced the Lua scripting support with Python. Now the main loop runs in the scripting language. For the 2007.1 release the main loop still ran in the engine side and the Lua library was embedded into the engine. By utilizing SWIG game creators can add Lua support back to the engine again if they have the need for it.

* Metamodel refactoring. This resulted in a better engine structure and a decreased number of dependencies between the engine modules. Furthermore bi-directional dependencies were removed completely.

* Improved engine modularity. Our aim is that you can cherry-pick specific engine modules of FIFE and just use the ones you really need for your game.

* Pathfinding integration. Exchangable backends so you can already replace the currently rather simple linear pather with a more sophisticated system.

* New 3d geometry system. Camera supports now tilt, rotation and zoom. Geometries are affected by camera adjustments, but in addition to that, layers can contain separate scale and rotation.

* New event channel module. Scripts have full access to mouse, keyboard & widget input.

* New MVC architecture pattern. The whole engine has been refactored to follow a model - view - controller pattern. This leads to a cleaner engine module hierarchy compared to the engine design that was used before.

* Improved action-based animation system.

* A techdemo to give game creators a starting point. This is a common undertaking of the Zero-Projekt team and FIFE.

 

Read the full release news update:

http://mirror1.cvsdude.com/trac/fife/engin...007/10/10/16.23

[Rosbjerg]

Congrats man, it looks great! Graphics are actually quiet decent..

[mvBarracuda]

The FIFE project is trying to fill various positions on the team with fresh blood. Feel free to take a look at the the detailed help wanted thread we posted at gamedev.net recently:

http://www.gamedev.net/community/forums/to...topic_id=471055

[mvBarracuda]

No fancy new screenshots this time but we updated our compile SDK for Windows systems. You'll need the new SDK to build the latest SVN revisions of FIFE on Win32. Grab the new SDK here:

FIFE Win32 compile SDK sfx 7zip version

FIFE Win32 compile SDK zip version

[mvBarracuda]

It's update time!

 

It was pretty quiet in the last weeks. One reason were the number of university tasks but another factor was my personal lazyness and lethargy caused by the holiday period. Now FIFE is back with yet another news update )

 

Here is a nifty new screenshot of the currently work in progress editor tool:

 

Furthermore our latest blog post covers the latest issues we encountered but also progress reports from the audio module, editor development and pather front.

 

You can check out the whole news update at the FIFE developer blog:

Time for heroes - yet another FIFE blog update

 

Last but not least: we're of course still trying to recruit new developers for the project ) Don't be shy and visit us at our developer IRC channel.

[mvBarracuda]

It's update time again ) three weeks have passed since our last official blog update and we're glad that we can report to have made several steps forward in this time.

 

To just summarize the most important points:

- We decided to revive the island_demo concept from the 2007.1 release and ship a island_demo like game with future FIFE releases

- View code changes to improve performance and resolve z-order issues

- Floating text renderer support for map instances

- VFS improvements for editor tool

- Vastly improved pychan extension

- FIFE documentation server with fresh doxygen, epydoc documentation generated on a daily basis; live IRC logs for developers

- Tweaked FIFE forums design

- Fresh blood on the team: two new programmers, one additional project manager

- New precompiled Win32 SVN snapshot available

 

Feel free to give the whole detailed blog update a read here:

http://mirror1.cvsdude.com/trac/fife/engin...008/01/28/17.22

 

In case you're just interested in fancy screenshots we got something for you as well. Here is a list of all new screenshots that are shown in the update:

Floating text support:

 

Continuum graphics pack: 45

[mvBarracuda]

Is there no way to edit the last post? I tried to find the EDIT button for about five minutes but failed :grin:/

 

The download link for the new SVN snapshot has changed, the new URL is:

http://downloads.sourceforge.net/fife/FIFE_r2057_win32.exe

 

In case I missed the EDIT link, could a moderator please edit the post above with the new link and delete this post? Thanks )

[Dagon]

Great work , can't wait to see what I can do when editor will be ready!

Your work is awesome!

[mvBarracuda]

We've done it again! We proudly present the FIFE 2008.0 release )

 

Download:

* Source package (FreeBSD, Linux, Macintosh)

* Win32 binaries

 

Screenshot of the new editor tool:

 

Youtube promotion trailer:

 

Theme of this release:

For the 2008.0 release we decided to pick up the island_demo concept again. Right: again. We already used this concept for the 2007.1 release but we put it on ice as we worked together with the Zero-Projekt developers on a different example game for the 2007.2 release. The cooperation didn't work out as planned for either side and so both teams stick to their own projects again now.

 

After the cooperation has ended we considered all possible options and finally decided to flesh out the old island_demo concept and base our future example game upon it. Although we were able to finish the initial planning work for the new island_demo game some days ago, there is almost none of the new planned content in place yet. On the other side the engine itself and the editor tool that ships with it now made _huge_ steps forward since the last release about 4 months ago. Therefore we decided to release the current status as stable release although it offers not many new aspects on the content side and the content that is in place is very likely to change with the 2008.1 release.

 

The milestone itself seems to be a very important step into the right direction. You can build your own maps now with the help of the editor tool. Furthermore a basic application structure is in place that should give you an idea how can create your own FIFE-based game. We'll surely adjust the structure of the island_demo client over the next months but this release is stable enough to start working on your game now.

 

This release is surely not the beginning of the end of the road for us, but it's the end of beginning for the FIFE development team. Enjoy it as we do )

 

Important changes since the 2007.2 release:

* New audio module with support for some more advanced audio features.

* Optimized view code to improve the performance and resolve map instance z-order issues.

* Switched back to island_demo concept; this will be the example game that ships with future FIFE releases as well.

* Ability to save maps

* Routing pathfinding (previous release only had linear pather)

* Pluggable rendering system with several new plugins included.

* New editor tool, which can be used to layout maps

* XML-format improvements: stack position, z-position, camera definitions, animation x/y offsets

* Vastly improved pychan extension; XML based GUI definitions

* Quadtree adjustments

* Ability to define own settings formats and interpret them from scripts

* Ability to attach cameras to instances (so that they are tracked while they move)

* Separate x/y scaling for cellgrids, allows more flexible geometry definitions (e.g. rectangle)

* Basic mouse map instance picking.

* Improved engine-script exception reporting

* Mouse cursors can be changed from scripts

 

Few non-code related changes:

* Rough task list for the whole island demo has now been created

* Vastly improved documentation with nightly updates for engine core & scripts

* Lots of new content gathered and created for island demo e.g. agent definitions, clouds graphics, music...

 

Read the full release announcement at the FIFE developer blog.

[Lare Kikkeli]

So, is anyone actually developing a game on this engine?

[@\NightandtheShape/@]

Not a bad result, i suppose if anyone seriously wanted to do a games based on solid 10 year old technology then this engine is perhaps the best for that purpose in the world. The only real question now is... WHY?!

[mvBarracuda]

Not a bad result, i suppose if anyone seriously wanted to do a games based on solid 10 year old technology then this engine is perhaps the best for that purpose in the world. The only real question now is... WHY?!

I wouldn't call it "solid 10 year old technology". Especially for independent developers modern 2d engines can be a good option for their games.

 

Concerning your "why?" question: I think that for the majority of the people who work on FIFE the motivation is twofold:

On one side there is the motivation to work on something that could help to bring back the games that you would like to play yourself. Quite many modern RPGs seem to suffer from the problem that they're targeted towards a more and more increasing audience. Usually developers try to attract younger gamers who often spend more of their pocket money on games than adults do. This is also related to the shift of development focus from PC to console games. There are fewer quality RPGs for PC and the ones that are in production are often cross platform (read: developed for modern consoles and PC) projects.

 

Oblivion was a fine example of what I would call the downfall of mature RPGs for PCs. A lot of aspects where simplified compared to Morrowind and the game appeared as a console to PC port in some aspect (e.g. the poor, very large user interface so the XBox users can read the fonts while sitting 3m away from the TV). My personal hope is that FIFE gets adopted by indie developers who would like to create RPGs in the spirit of Fallout, Planescape Torment or Arcanum. While these games didn't feature top notch graphics, I enjoyed them far more than any of the mass market RPGs that were released in the last several years (Vampire: Bloodlines by the new defuct Troika might be the only exception to this rule).

 

On the other side there is some personal motivation involved for every developer who works on FIFE. By working on such a project you learn a lot about how to create games, work in a team, communicate with other individuals, discuss concepts with them. Basically working on FIFE helped to build up my soft skills in a lot of areas and learn how to work in a team with people from all over the world. So spending all the time on the project might help me to get a job once I finished university )

[mvBarracuda]

The FIFE team released a new version of their compile SDK for win32 systems today.

 

The new SDK addresses an issue with MSVC2005 debug builds that was just recently discovered. Furthermore libSDL and SWIG have been upgraded to the latest versions.

 

Read the full announcement at the FIFE development blog:

New compile SDK for win32 release announcement

 

Download URL:

* FIFE 2008.0 win32 compile SDK (selfextracting 7zip version)

* FIFE 2008.0 win32 compile SDK (zip version)

[@\NightandtheShape/@]

Not a bad result, i suppose if anyone seriously wanted to do a games based on solid 10 year old technology then this engine is perhaps the best for that purpose in the world. The only real question now is... WHY?!

I wouldn't call it "solid 10 year old technology". Especially for independent developers modern 2d engines can be a good option for their games.

 

Concerning your "why?" question: I think that for the majority of the people who work on FIFE the motivation is twofold:

On one side there is the motivation to work on something that could help to bring back the games that you would like to play yourself. Quite many modern RPGs seem to suffer from the problem that they're targeted towards a more and more increasing audience. Usually developers try to attract younger gamers who often spend more of their pocket money on games than adults do. This is also related to the shift of development focus from PC to console games. There are fewer quality RPGs for PC and the ones that are in production are often cross platform (read: developed for modern consoles and PC) projects.

 

Oblivion was a fine example of what I would call the downfall of mature RPGs for PCs. A lot of aspects where simplified compared to Morrowind and the game appeared as a console to PC port in some aspect (e.g. the poor, very large user interface so the XBox users can read the fonts while sitting 3m away from the TV). My personal hope is that FIFE gets adopted by indie developers who would like to create RPGs in the spirit of Fallout, Planescape Torment or Arcanum. While these games didn't feature top notch graphics, I enjoyed them far more than any of the mass market RPGs that were released in the last several years (Vampire: Bloodlines by the new defuct Troika might be the only exception to this rule).

 

On the other side there is some personal motivation involved for every developer who works on FIFE. By working on such a project you learn a lot about how to create games, work in a team, communicate with other individuals, discuss concepts with them. Basically working on FIFE helped to build up my soft skills in a lot of areas and learn how to work in a team with people from all over the world. So spending all the time on the project might help me to get a job once I finished university )

 

It was more of a sly tongue in cheeck referance to the fact that it's been about 10 years since a new adn acclaimed 2D isometric engine was developed for commercial purposes. That is what was meant my point being that, despite it's quality it's still an excercise in fuitility to develop a 2D engine of this vain.

 

When I stated "Why?" it was in referance to why would someone choose such a 2D engine, not why to develop such an engine... The point being that an engine is not responsible for making game like Fallout good, I suppose you get those die hard 2D fanatics whom are firmly offended by any idea of progress and bitch and complain at every turn because of change.

 

Basically put what I am trying to say is something like, lets take say Planescape : Torment, lets apply the same generic game mechanics, point click etc... The only thing which I'm going to hypothetically change is the 2D to 3D, is it really a different game? I don't think so, infact it wouldn't need to even look that different, if anything it could potentially add to it.

 

It would have made more sense to me personally to take the generic mechanics of something like Fallout, but make it 3D, that I would have considered to be progress, instead its not, it's a step back... That said I do imagine that the project will have a varied amount of success in indie circles and may even find a nice indie project or two. There is always a luddite lurking around somewhere.

[mvBarracuda]

It was more of a sly tongue in cheeck referance to the fact that it's been about 10 years since a new adn acclaimed 2D isometric engine was developed for commercial purposes. That is what was meant my point being that, despite it's quality it's still an excercise in fuitility to develop a 2D engine of this vain.

Please elaborate ) As expressed: I seriously believe that FIFE will be a great option for hobbyist and independent developers. This engine is not meant for modern AAA titles; I do completely agree that this engine will never be featured in a mass market game.

 

Let me ask a simple a question: if you would be an independent or hobbyist developer why would you create your own engine from scratch and invest a lot of time into it if there would be a free (as in free beer and in free speech) solution that suits your needs? What major drawback do you see that stops the mentioned target audience from adopting FIFE?

 

When I stated "Why?" it was in referance to why would someone choose such a 2D engine, not why to develop such an engine...

Because writing a good engine takes a lot of time, prolly around two years if you start from scratch. So why would somebody who just wants to create a game spend two years trying to write something that is already available for free and can be customized to suit his / her needs?

 

The point being that an engine is not responsible for making game like Fallout good, I suppose you get those die hard 2D fanatics whom are firmly offended by any idea of progress and bitch and complain at every turn because of change.

It's not about being a 2d or 3d fanatic. My point is that it's easier for hobbyist and indepentent developers to use a 2d engine in bunch of cases than to go for full-featured 3d.

 

There are also some problems associated to the new possibilites that 3d engines introduced. In case you add a first version view to your game you will adjust the gameplay to work from the first person perspective as well. E.g. Party-based RPGs work far better in isometric perspective than in first person view.

 

The graphical restrictions of a 2d engine seem to help the developer to focus on the IMO most important element of game design: to come up with solid and enjoyable gameplay. Recent RPGs like Oblivion went more and more into the direction of first person shooter gameplay: shiny graphics but lacking a lot of soul. I want complex dialogues like in Planescape Torment, I want real choices with real outcomes like in Fallout, I want a party of adventurers who are going along with me like in Arcanum. I couldn't care less about shiny bloom or "immersion" of a real 3d world I can see with my own eyes in first person view.

 

Basically put what I am trying to say is something like, lets take say Planescape : Torment, lets apply the same generic game mechanics, point click etc... The only thing which I'm going to hypothetically change is the 2D to 3D, is it really a different game? I don't think so, infact it wouldn't need to even look that different, if anything it could potentially add to it.

I wouldn't mind a 3d remake of Planescape Torment. But Planescape would NOT work in first person view; Planescape featured strong party-based gameplay so it's essential IMO that a remake would just feature isometric perspective. As soon as you introduce first person view you start to make gameplay compromises. E.g. turn-based combat as featured in Fallout doesn't work really well in first person view, so let's make it a real-time button mash fest.

 

It would have made more sense to me personally to take the generic mechanics of something like Fallout, but make it 3D, that I would have considered to be progress, instead its not, it's a step back... That said I do imagine that the project will have a varied amount of success in indie circles and may even find a nice indie project or two. There is always a luddite lurking around somewhere.

Time will tell ) I don't mind a graphical step backwards if it helps to focus on bringing the gameplay some steps forward. If we need to go back to the graphics standards of 1999 to see some new games with excellent gameplay as the ones from the golden age of RPGs I'm fine with that.

Edited February 11, 2008 by mvBarracuda

[@\NightandtheShape/@]

It was more of a sly tongue in cheeck referance to the fact that it's been about 10 years since a new adn acclaimed 2D isometric engine was developed for commercial purposes. That is what was meant my point being that, despite it's quality it's still an excercise in fuitility to develop a 2D engine of this vain.

 

Please elaborate ) As expressed: I seriously believe that FIFE will be a great option for hobbyist and independent developers. This engine is not meant for modern AAA titles; I do completely agree that this engine will never be featured in a mass market game.

 

 

If anything I agree, for indie devs, hobbyist etc... But I don't see the point.

 

Let me ask a simple a question: if you would be an independent or hobbyist developer why would you create your own engine from scratch and invest a lot of time into it if there would be a free (as in free beer and in free speech) solution that suits your needs? What major drawback do you see that stops the mentioned target audience from adopting FIFE?

 

If I was an indie developer, I would suppose I already fall under that "Specification", and I would do a better job, in that dream scenario I'd develop an isometric 3D engine.

 

It's not about being a 2d or 3d fanatic. My point is that it's easier for hobbyist and indepentent developers to use a 2d engine in bunch of cases than to go for full-featured 3d.

 

Easier, yeah I suppose there is a significantly easier learning curve in 2D, but there is also a significant increase in asset development time, and a significant increase in asset sizes.

 

There are also some problems associated to the new possibilites that 3d engines introduced. In case you add a first version view to your game you will adjust the gameplay to work from the first person perspective as well. E.g. Party-based RPGs work far better in isometric perspective than in first person view.

 

So because party based RPG's work better from a 3rd person perspective I should us a 2D engine, insane absolutely INSANE logic there my friend.

 

The graphical restrictions of a 2d engine seem to help the developer to focus on the IMO most important element of game design: to come up with solid and enjoyable gameplay. Recent RPGs like Oblivion went more and more into the direction of first person shooter gameplay: shiny graphics but lacking a lot of soul. I want complex dialogues like in Planescape Torment, I want real choices with real outcomes like in Fallout, I want a party of adventurers who are going along with me like in Arcanum. I couldn't care less about shiny bloom or "immersion" of a real 3d world I can see with my own eyes in first person view.

 

I wouldn't say you can really compare Oblivion or the eldar scrolls series in general to the previous works of Trokia and Black Isle.

 

Complex character interaction doesn't require a 2D engine, please see MotB.

 

Time will tell ) I don't mind a graphical step backwards if it helps to focus on bringing the gameplay some steps forward. If we need to go back to the graphics standards of 1999 to see some new games with excellent gameplay as the ones from the golden age of RPGs I'm fine with that.

 

2D doesn't improve gameplay... The concept that it could is bizzare to say the least you can use a 3D engine and have the exact same mechanics. So your arguement doesn't make sense, you keep bring oblivion up, but you fail to mention games such as NWN's series and Kotor series, which certainly have elements of gameplay that you desire. I would add that gameplay has been dumbed down in some games of late, but the truth of that is more about mass appeal than 3D gfx.

[Musopticon?]

Can you lay off him, Nightguy? It's not like either of you are making a point here.

Edited February 12, 2008 by Musopticon?

[Gorth]

Can you lay off him, Nightguy? It's not like either of you are making a point here.

I tend to agree with my nordic friend here... I know the thread title says "feedback appreciated", but I suspect that was feedback regarding the Engine itself, not the comparative merits of 2d and 3d engines

[@\NightandtheShape/@]

You guys always spoil all my fun!

[Musopticon?]

Well, this isn't your playground.

[@\NightandtheShape/@]

I was actually trying to have a constructive conversation, as I couldn't understand the general motivation behind developing an engine of this manner, why not develop something entirely different... All said, I have tried to maintain that the product itself is of a fairly high quality.

 

I was honestly trying to understand why so much effort had been put into a product of this type, but really it isn't all that different to say "Devils Whiskey"...

 

*Goes back to playground* WEEEE!!!!

[Gorth]

I was actually trying to have a constructive conversation, as I couldn't understand the general motivation behind developing an engine of this manner, why not develop something entirely different... All said, I have tried to maintain that the product itself is of a fairly high quality.

 

I was honestly trying to understand why so much effort had been put into a product of this type, but really it isn't all that different to say "Devils Whiskey"...

 

*Goes back to playground* WEEEE!!!!

Hey, easy now

 

(don't forget your ball!)

 

Fair enough question, just not sure that it came across the right way though? I think he answered your original question with something that could be written in short as "Personal Preferences"

[mvBarracuda]

Hopefully we can bring this thread a bit back on track with this on-topic update )

 

It's time for yet another FIFE update )

 

This time we got something special for you: a new series entitled "FIFE community spotlight". The idea behind these new community spotlight articles is to have a closer look at the projects which are utilizing FIFE for their games. The first community spotlight covers the computer roleplaying game project "Zero" that is currently in development.

 

You can read about it at the FIFE blog:

FIFE community spotlight #1: Zero - a postnuclear cRPG

 

Here is some eye candy from Zero to convince you to actually read the whole community spotlight article:

 

 

 

 

 

 

 

 

Last but not least and slightly related to the community spotlight series: we did recently declutter our wiki starting page. Check out the new and hopefully improved version here:

FIFE wiki

 

All kind of feedback concerning the community spotlight and the new wiki starting page is appreciated by the team )

[Gorth]

Man, Fallout 1&2 never looked that good. Who vonlunteers to do a remake?