Frictional Games the developers of the horror Penumbra series recently revealed the title of their next project – Amnesia : The Dark Descent which I wrote about earlier.
For this special event I interviewed the developers about their next title.

1. Hello Jens and thank you for the interview,
There is a question I ask every commercial game company that already released their previous games to GNU/Linux,  I wonder how where the overall sales of the Penumbra Trilogy per platform ? from the “Thank You” post I feel that there is indeed a market for GNU/Linux, but can you be more specific (at least percentage wise for each game and the trilogy pack)  ?

It’s a bit difficult to give any exact numbers, mainly because for Windows we have a publisher and they in turn use many partners, so even if we get sales figures they are not always detailed down to a unit level for all territories. To make matters worse the first publisher we had for the original Penumbra: Overture release where pretty much a scam and never sent any sales figures at all, so we have only a slight idea of how well that game sold.

I did my best to collect as many figures as I could for online sales of the game for Linux, Mac & Windows. To give some background, the only place to purchase Linux is through our online store, the Mac version is available in our store + 3 other online stores and the Windows version is available in our store (recently added) + all the major stores (Steam, D2D, Gamer’s Gate, etc). I could not get all the figures for the Windows version, it also further weakens the figures that the Windows version is available in retail and that it has had backup by publishers. So it has had a lot of marketing that the other platforms have not had.

Here are the percentage for each platform, based on the total sales of all the games (individual and as the Collection):

Windows: 80%
Linux: 12%
Mac: 8%

A large part of the Linux percentage is thanks to the “Thank You” post you mentioned, that the article it mentions and its readers are what we thank in it for purchasing a lot of copies during the USD 5 weekend sale.

I wish this was all very positive, but in reality our games do not do that well. The only reason the company exists is because we live on what the treasure chest has to offer! So while the Thank You post is obviously a big thanks (it is!), it’s more that it bought us some more time, at a time when we really needed it badly.

We sold 3000 copies for Linux during that weekend, which is a GREAT number of units to sell during a weekend, but for a price of USD 5, you get about USD 13 000 to put in the chest. We are five people in the company and we also have a great guy making the Linux port, so with that in mind you don’t get that many days extra to live on. Here in Sweden the tax rate is above 30% and the employer fee above 30% too, so to put say USD 1 000 in your pocket to buy food for it will cost the company around USD 1 800. That does not include the additional 12% of VAT you have to pay for the actual food! Haha.

There I went on a long trip… It was not to nag about the poor state of being an independent game developer, it is great to be part of a small company that can quite freely work on the type of games we want. But it does have some parts that are a bit of a strain on the creativity.

2. What can you tell us about the upcoming game Amnesia and how would you compare it to the Penumbra series gameplay/engine wise ? (please be detailed)
I’ll leave this question for Thomas, he spends a great deal of time with the design of the game and is properly the man to ask for this question.

Thomas: One could say that Amnesia is a sort of progression from Penumbra. Many of the core gameplay elements have been taken from Penumbra and then refined in Amnesia. This is stuff like the physical interaction system, the way which we design our levels, and so on. Anyone who has played Penumbra will feel right at home when starting Amnesia. What really makes Amnesia different from Penumbra gameplay wise (apart from story and settings) is were our focus have been this time. These are the main areas:

1) To make the player the protagonist. What this means is that the player should to the extent possible be in the shoes of the character he/she is playing as. We will not have any sort of cut scenes, there will be very little text feedback (like descriptions) and the player should be in control from start to finish with no time jumps or anything like that. We feel that this will increase the immersion and emotions experienced during gameplay.

2) Much more effort has been put on the graphics, in terms of details, quality and diversity. As we no longer have text comments that can enhance the environments we want the player to be able to get the feelings by themselves. When entering a room the mood should be apparent and work towards enhancing plot and atmosphere. So instead of the protagonist telling the player what they should think or feel towards something, it is up to the player to decide and hopefully forcing him/her to become more immersed in the game world.

3) To make a more streamlined experience where the player does not get stuck all the time on a puzzle or a gameplay element. We want to have difficulty in the game, but we are designing it in such a way that it should not make break the immersion. Whenever the player restarts from a save or gets stuck at some puzzle for too long, he/she will drift away from the game world and the quality of the experienced will be lessened. This means we are working very hard to get rid of these things and we do not want to throw in a puzzle just to lengthen gameplay time (something that we did in Penumbra).

Engine wise it is quite different and there many parts that have been rebuilt from the ground up. The Penumbra games uses a portal system to do occlusion culling (check what objects not to draw) that placed a heavy burden on the artists. In Amnesia we used a fully dynamic culling system based on a technique called Coherent Hierarchical Culling, which does not give any extra work for the artists and give them more time to make nice graphics instead of worrying about performance issues. Another example is the shadows that were stencil (shadow volumes) based in Penumbra but is done with shadow maps in Amnesia. There are all kinds of new effects added as well, like Screen Space Ambient Occlusion, proper decals, etc. This means that Amnesia will not support as many graphic cards as Penumbra did, but we felt that we had to improve and were forced to draw the line somewhere. Still, Amnesia will be a lot less demanding than most commercial 3D games released these days. If you are looking for minimum requirements, then I am afraid it is too early for that as proper testing on that front has not started yet.

3. Usually I don’t like puzzle games, that’s because I am not so good at solving them – but I really enjoyed playing the Penumbra games because of it’s dark and scary atmosphere, how are the puzzled in Amnesia compare to Penumbra ? more challenging or easier ? can we expect more of the same or the new engine will let you make more complex and interesting puzzles  ?
For Amnesia the puzzles can probably be summed up as being a bit easier, it has been a design goal for the game to be a bit more streamlined in the experience. With Penumbra it was quite clear that as you encountered a puzzle, you had to spend some time trying to solve it, in Amnesia we have tried to smooth it out a bit more. Making puzzles more integrated into the rest of the game, so that we don’t get these “puzzle breaks” in the gameplay.

Amnesia is more about the exploration and horror than Penumbra, we have definitely concentrated on a game where the atmosphere is dense and trying to create an intriguing story. Much of the game is about finding out who you are and piece everything together to get the answers you seek.

So no, the engine is not really going to give any complex puzzles, but it will however give a better interaction with the puzzles. The whole physics interaction, and interaction in general, has been tweaked a lot so it is for example much easier to open doors, turn valves and so on.

4. What tools/libraries/programs have you used to develop Amnesia ? (3D, 2D, programming, platforms, etc…)
Oh, hmm, tons of stuff. Speaking for myself I use a Mac (old PPC model) to do all the work with sounds and video editing, I also use it to write down all my notes about bugs, the todo etc. The programs are quite old, I used to run a sound & music company, so it is Cubase SX 2.0, Spark XL (defunct) and an almost none working version of Peak. I use Windows for the development, which is mainly using our own tools for Level & Model editing in combination with writing the script code for the game, which I do using Notepad++. I do some simple modelling, often placeholders, and some basic image editing, for this I use Blender and GIMP.

Don’t worry, we use Linux too! Our server is Linux and I have a cute Eee with Linux that I usually use to login to the server and also to do some work with when on the run. It’s really great that server wise it does not really matter what platform you are accessing it from, the Mac obviously works out of the box and for Windows you have PuTTY.

Sounds like a ton of computers for a poor independent developer? Well, the Mac was traded for another older Mac + some extra cash in 2006, my PC is home built with some parts dating back to 2001 (damn that SoundBlaster card is really great or not much has happened with sound hardware/software compared to the graphics industry…).

Amnesia it self is built using cross-platform libraries only: OpenGL, OpenAL and Newton Game Dynamics are the major parts and then a slew of various libraries for mixed tasks. You can peak at Penumbra to get a specific list, it has changed very little for Amnesia!

5. At the Penumbra series the player was encouraged to run/hide from monsters , while in the first Penumbra it was possible to kill the dogs, in the second game killing the monsters was impossible (without a patch/mod) , how Amnesia will be compare to those games ? will the player be able/encouraged to kill monsters ? will Amnesia be more action oriented (fighting/climbing/jumping etc…) ?
And if so what weapons or/and skills can we expect ?

If you asked this question a year ago, I would have answered that Amnesia would be more action oriented. We spent about a year creating prototypes and have tested a lot of different approaches, what we originally wanted was to avoid some of the very tedious parts of developing a game like Penumbra. With Penumbra almost every instance of the game is uniquely created, each puzzle has to be done from the ground up, each event and so fort. It’s not like a First Person Shooter where you can put a lot of gameplay hours into having the player shoot enemies and once you have set it all up it can be re-used multiple times.

One idea we had was to add a bit more combat, to go back to the Overture idea that the player would always have a last resort of defence. This time we did the combat a bit easier and worked a lot on how the AI should behave, so that it gave the player a better chance to sneak or do a bit of battling to get out of a tricky situation and then hide again. But during testing we had the same problem as always, half thought it was way too easy and the other half thought it was very difficult. Adding to this we felt that it was not as fun as we had hope, that at best it would be like a poor-mans version of Condemned, as a small developer we really have to try and be unique, we can’t afford having a lot of “well it’s like game X only worse” comments.

So if comparing to the Penumbra games, Amnesia will be most similar to Black Plague. But with the differences that enemy encounters will always be very short, more of an obstacle to overcome, than levels filled with patrolling enemies.

6. Amnesia will use the new and improved engine HPL2, what’s new in this engine ? what new features will it have ?  are you willing to license this engine to other companies ?
Also at this note, is there a chance you will release your older engine HPL under some FOSS license like GPL as id software does with their older engines ?

To make sure you got the best answer to this question I asked Thomas to jump in again, he is after all the programmer.

Thomas: I have already gone through some of the new features in a previous answer, so I will let that be.

We do not plan to release any commercial version of the engine at this point since it would take too much time in support and probably not be worth it. Also, what we want to do is to make games, not sell SDKs / tools.

As for the releasing the old engine, we have talked about it, but it is not all that good and it would take some work to get it into some state that is usable for people (meaning documenting and such). I am also unsure how many people that would do anything useful with it now that engines like Unity can be gotten for free and has a lot better tools and support than the first HPL engine. This combined with the time it would take to make a proper release, makes it quite unlikely that it will happen. But as always it depends on how big the interest is and so on.

7. The community made some mods for the Penumbra games, how mod friendly will be Amnesia ? will you supply modding tools for the community and if so what can you tell me about them ?
Amnesia will be a lot more mod friendly. Developing tools take a lot of time, with Penumbra there was not time to do any advanced tools, creating the engine was time consuming enough. For Penumbra Thomas was pretty much the sole programmer, but for the whole Amnesia project we have had one additional programmer on-board that has concentrated fully on developing tools.

What we have now is a Level Editor, Model Editor, Particle Editor and a Material Editor. The main tools are by far the Level and Model editor. In the Model editor you import a 3D model, say a door and its frame, and then you use the editor to add physics bodies, joints and set properties for behaviour. For example you specify that the door is made out of wood, what it weighs, that it is connected by a joint to the door frame, that is has a gameplay specific value of being a door (which gives it a lot of predetermined behaviour, that is also easy to change) and that the joint will make certain sounds when the bodies connected to it move around. This allows you to create a final door entity, that you then can use over and over in the Level editor. Each time you use it you can give it specific properties, such as if the door should be locked or if it should be opened a bit when the player first enters the level.

The Level Editor is the most fun to use, you build the levels using sets of static graphics, you have pieces for walls, ceilings, railings and other type of main components to create rooms. Then you add in entities that are doors, chairs, items, gameplay parts and so on, all comes with a specified behaviour so in it’s simplest form you already have some gameplay in the level. Then you can continue on and add lightning, decals, particles, fog and other type of effects to spice it all up.

There are a lot of things that you can do in the Level Editor that we previously had to script in Penumbra. This saves a lot of time and also allows the user to do a lot more without having the need for any script knowledge. But of course, when you want to add in some really fun stuff you have to start writing scripts, but it’s at a quite high level so what limits the mod creation will be more about the effort than the requirements on skills.

To summarize, we have created a lot of tools to make the development of the game much more effective for us. But we have had the community in mind while doing it, so the tools will be part of the game and the game itself has a basic support for making use of user created content, more specifically a special option to load user maps.

As it looks now the tools will be available for all three platforms, something that we are very happy about!

8. What can you tell me about the replayability of Amnesia ? will there be several solutions to each quest/puzzle ? what about different paths/ways to end the game ?
We are strong believers of making games that definitely does not necessary need a lot of replayability. Since we make a story driven game that is much about exploring and experiencing the game we feel that it should be one experience that is as good as it can possibly be. We don’t want the player to get stuck too often in the game, so the story and environment often push the player towards a specific goal, but this does not mean you have to take that specific path.

Much like in Penumbra you can explore the world in more than one specific path, there are places you can go and solve parts of a large puzzle in different orders. There are of course events that only occur based on certain criteria, it could be as simple as revisiting a level three times for a spooky event to occur (as a precaution to make sure that if someone goes back and forth a lot there will be some events going on making it a bit more interesting).

There will be story bits and puzzles that are not necessary to complete the game, so depending on how you play there will always be the option to replay to make sure you got all the story bits and puzzles. For the ending we do not want to talk about it really, but there will be some differences depending on what you have done during the game, but it’s not something we would put down as a main feature or talk too much about as it would ruin the final game.

Our type of “replayability” is perhaps conveniently compared to that of films, if you really enjoy it you see it one-two times again to really suck it up all up and notice the details you missed the first time.

9. When could we expect Amnesia to be released ? will the Linux version be released at the same time as the Windows one ? what about preorders ?
Sometime during the summer 2010 is the current goal, it might change, but we usually manage to stay on track. We have had a very confusing time with this project, with publisher involvement planned than then has gone down the drain, close to the closing of the whole company and so on and on, so we have had periods when we have had to concentrate a lot on other areas than the actual development of the game. But since this summer it has been full steam a head and keeping up with the schedule.

The Linux version should be available on or around the same time as the Windows release. We can inform already that Edward Rudd, the Linux porter, have the game up and running on Linux as well as the Level Editor. So things are looking good at the moment!

We will definitely make sure there is an option to preorder before the release, not until next year though. We can absolutely make use of all the support we can get!

10. Is there anything I’ve missed and you want to add to the interview ?  
I think it is all good! Looking forward to be in touch again next year, perhaps with some new gritty details or even playable previews!

Thanks for asking us to do an interview.

And thank you Frictional Games for answering in so much detail.

Frictional Games
Amnesia : The Dark Descent
Penumbra Series

  1. blink-179 says:

    Amnesia should be like the first Penumbra. Then it’ll have a success…

  2. grzywacz says:

    I’m waiting for this game in anticipation. I think I’m going to give pre-orders a try, given the quality of experience I’ve had with Penumbra!

  3. Social comments and analytics for this post…

    This post was mentioned on Twitter by frictionalgames: Interview about amnesia and the company posted at

  4. #1 with a bullet says:

    The problem with looking at OS numbers and sales is that if you bought the Windoze copy to play through Wine it will still count as a +1 for M$ which is a problem. This pretty much goes for all games, of course.

  5. Patrick says:

    @ #1 with a bullet:
    That’s why I hate WINE. A lot of Linux users don’t buy native Linux games, because they can play the (often much cheaper) Windows version through WINE. Those are only sales for the Windows platform. WINE really has to die, so people buy the Linux versions, instead of the Windows versions and without WINE people will start asking for native Linux versions. Nowadays the developers and publishers just say: “No, there won’t be a native Linux release. Just buy the Windows version and use WINE.” Without WINE they have to do a native Linux release and what’s the result? No more Linux users buying software for Windows, so sales for Windows will decrease and because more native Linux software will be released sales for Linux will increase.

  6. Vadim P. says:

    That’s a great interview you scored there.

    I only buy Linux games (and play). Don’t see a place to pre-order this one tho.

  7. qubodup says:

    This is definitely the most interesting interview I have red on this blog so far! I’ll finish reading it later but thumbs up already 🙂

  8. qubodup says:

    PS: slashdotted (I think.. I’m always confused by their submission form) Sorry for the typo, I should go sleep…

  9. Maxim Bardin says:

    A good interview depends on :
    1. The questions
    2. The answers
    3. How much are you interested in the game

    In my opinion one of the best interviews I’ve made was with Konrad From Bitgap – The Xenocell Developers – his answers were long and detailed – but not many people knew about the upcoming game so there was not enough interest in it.

    The Amnesia interview was also very very good – and as the developers are known in the GNU/Linux gaming community many people where interested in it.

  10. Maxim Bardin says:


  11. qubodup says:

    🙂 I’m very interested in this game, because I love the atmosphere of Penumbra. Also the details about Linux use and Linux sales are very interesting (and able to influence porting decisions of developers).

  12. T_U says:

    I really do appreciate the way Frictional Games makes use of humble resources (manpower + hardware). I’m glad I bought the Linux version twice ; though a bit annoyed they barely earnt enough to survive financially.

    I got definitely much more hooked than by any big budget game and hadn’t been so scared by a videogame since… 1997 (Fallout 1).