Nordenfelt - 2D Steampunk Shmup

[Udderdude]

In my opinion you should ditch the tiled background and go for a fully drawn background. You seem really good with drawing on paper, so just draw it there and then scan it, then color with Photoshop or something.

[hermitC]

I'm not sure about the sub genre yet. Can anybody give me a list of shmup types? I divide them in "bullet hell" and "others". What types are in "others"?

Just stumbled over this discussion about genres.

[hermitC]

In my opinion you should ditch the tiled background and go for a fully drawn background. You seem really good with drawing on paper, so just draw it there and then scan it, then color with Photoshop or something.

I've already thought about this approach. My concern with it is that I want to try some procedural content generation. One solution would be offline generation of levels (only in my level editor) and adapting them for gameplay. Another possibility would be PCG for creating missions and their interconnections (let's call it background story) according to player input. Both need a building set of level parts which handcrafted backgrounds can't provide.

The crucial point here is time. At the moment I'm fleshing out gameplay and fun. Then I will create a few levels and estimate the left effort for the whole game. If there is time left I will try to add some variance with PCG. If PCG does not fit into the schedule the tile engine is not necessary and more handcrafted stuff will come into play.

[Udderdude]

In my opinion you should ditch the tiled background and go for a fully drawn background. You seem really good with drawing on paper, so just draw it there and then scan it, then color with Photoshop or something.

My concern with it is that I want to try some procedural content generation.

Procedural levels don't really need full backgrounds like that. You could get away with just something generic there ..

I did something similar with XOP EX level, where it randomly throws enemies from all of the levels at you.

[hermitC]

Just a quick status update:

I'm busy implementing an editor for Nordenfelt. Hopefully the basic version works within the next few days so I can drop a video creating a small level here.

[hermitC]

6 weeks after the last post :embarassed1: I'm back to show you the Nordenfelt editor in action:

Nordenfelt Editor Version 0.1

I'm going to publish demos of the game in the near future. I will drop a line here when I've something to play for you.

[BPzeBanshee]

Looks easy enough to edit, though at this point I'm a bit unsure as to where your emphasis in subgenre of shmups is. I dont mean to seem a bit annoying by bringing this subject up again but since it has been six weeks: Have you actually decided what subgenre to use? Will I be seeing bullet-hell stuff ala Magic Broom Shooting or manic stuff ala Raiden or something inbetween like Garegga (subject to argument here but I personally say it's inbetween these tiers) in relation to this game?

[hermitC]

Looks easy enough to edit, though at this point I'm a bit unsure as to where your emphasis in subgenre of shmups is. I dont mean to seem a bit annoying by bringing this subject up again but since it has been six weeks: Have you actually decided what subgenre to use? Will I be seeing bullet-hell stuff ala Magic Broom Shooting or manic stuff ala Raiden or something inbetween like Garegga (subject to argument here but I personally say it's inbetween these tiers) in relation to this game?

A really good question.

To be honest: I'm not sure yet. I'm not a fan of those shot-cluttered bullet-hell games. IMHO they shrink the gameplay to wowing graphics and dodging. Correct me if I'm wrong. I just know a few of them.

I want to add a little bit more to the "kill everything and the boss" approach. Tactical weapon choice, searching items, alternative routes, etc. There are some ideas floating around. Therefore I don't want to restrict the play style too much.

To answer your question: Garegga is something you can expect. The feedback on the demos will decide which features work and what's crap.

[BPzeBanshee]

I want to add a little bit more to the "kill everything and the boss" approach. Tactical weapon choice, searching items, alternative routes, etc. There are some ideas floating around. Therefore I don't want to restrict the play style too much.

I notice some games have a suttle way of tactical weapon choice with the amount of efficiency a certain weapon will do to an enemy, as well as in some cases the upgraded forms used for different purposes from not-so-upgraded versions of the same weapon. I see this in Thunder Force IV with the standard doubleshot taking me two minutes to beat the Stage 1 boss and 8 minutes with the upgraded Blade weapon. Also in Garegga you get more score not from just killing off the boss but by taking out all its key points. I see alternative routes in the latest incarnation of Afterburner which provides good replay value as well as change in gameplay depending on which level you chose.
If you can do it right, such functionality can be implemented real well, especially as there are more examples of successful implementations of these additions to the standard gameplay. This is also why I suggested Tyrian before in my earlier posts since it has all these in its Story Mode to a point (weapon selection dictated how well you did in a certain stage but not so much for the bosses, also the game counted a lot on having a better shield for later stages more than anything else, which I hope this game doesn't go in that direction).

To be honest: I'm not sure yet. I'm not a fan of those shot-cluttered bullet-hell games. IMHO they shrink the gameplay to wowing graphics and dodging. Correct me if I'm wrong. I just know a few of them.

Fair enough assumption, especially since there aren't many bullet-hell games that actually try to go beyond the wowoff and insane dodging style of things. A discussion I had with a friend the other day pretty much ended with us both not really seeing any modern commerical games that succeed with the bullet-hell genre outside the old-skool lovers of the stuff which you have mentioned.
Personally for this project I'd recommend somewhere in between bullet-hell and pure manic since if I'm to consider Battle Garegga as a good influence for example, its somewhere in between true bullet-hell and the standard manic, in other words its not heavily orientated around either subgenre since it requires equal reflex motion to destroy and dodge quickly like manics but also does not have insanely fast bullets and it does have a fair few put on the screen so you still have to dodge a fair bit in similar way to bullethell games sometimes. A game project of mine failed for not really picking subgenre right as well as failure in optimisation early on the project (bad programming skills on my part) and I've seen similar stuff happen with other things so my point is basically: eventually you really will need to clearly define whether you're somewhere in between the subgenre tiers or right in to one tier in particular.

[hermitC]

My basic reference titles are Garegga and Tyrian. Garegga for its play style and steampunk touch. Tyrian's story mode always has been the main inspiration for the equipment management mechanics. Those became secondary in favor of gameplay development.

Bullet-hell is no longer an option. At the moment Nordenfelt is planned to be more manic. You are right by pointing to the lack of clarification about the subgenre. The upcoming first demo will give me first impressions how it's style is received. This brings me to the deadline for the demo:

Nordenfelt 0.1 will be available for download on July 4th

You can find a list of included features here.

[hermitC]

Concept of first boss in Nordenfelt 0.1:

[hermitC]

End boss rendered:

Concept:


Rendered:

[Udderdude]

You're kidding, right? Sorry, but after seeing the other art you've made, that just doesn't match up at all.

[BPzeBanshee]

You're kidding, right? Sorry, but after seeing the other art you've made, that just doesn't match up at all.

It's a steampunk box fortress with a cannon. Personally I think it needs more cannons.

Besides that it seems consistant graphically but not with the shape, I think this is what Udderdude meant.

[hermitC]

@Udderdude: No joke but an experiment. I want to check out how people receive such odd designs. First feedback arrived I'm always interested what exactly the problem is. Can you give me more details about your opinion, please?

@BPzeBanshee: Finally the boss has 4 cannons. They are all the same because there is no time left (until July 4th) for variety. The blocky shape is intended as I mentioned above.


Objects with simple forms are easier to create but seem to miss some expectations. What do you think about this simplicity?

[Udderdude]

@Udderdude: No joke but an experiment. I want to check out how people receive such odd designs. First feedback arrived I'm always interested what exactly the problem is. Can you give me more details about your opinion, please?

@BPzeBanshee: Finally the boss has 4 cannons. They are all the same because there is no time left (until July 4th) for variety. The blocky shape is intended as I mentioned above.


Objects with simple forms are easier to create but seem to miss some expectations. What do you think about this simplicity?

It doesn't look steampunk. It doesn't look like anything but a bunch of boxes and girders, really. I would maybe save the cannon part, but scrap the rest.

[hermitC]

@Udderdude: No joke but an experiment. I want to check out how people receive such odd designs. First feedback arrived I'm always interested what exactly the problem is. Can you give me more details about your opinion, please?

@BPzeBanshee: Finally the boss has 4 cannons. They are all the same because there is no time left (until July 4th) for variety. The blocky shape is intended as I mentioned above.


Objects with simple forms are easier to create but seem to miss some expectations. What do you think about this simplicity?

It doesn't look steampunk. It doesn't look like anything but a bunch of boxes and girders, really. I would maybe save the cannon part, but scrap the rest.

Ok, noted. Version 0.1 will use this boss, following versions may improve it or cut it at all.

[hermitC]

Kamikaze sprite, from design to render:


Concept


3D Model


Ingame Sprite


The demo will be out within the next 24 hours. I'm looking forward to your feedback.

[BPzeBanshee]

The kamikaze ship looks allright, I can imagine hoardes of them appearing from the top of the screen and aiming towards you, or even spinning and moving down while spinning towards you like Stage 2 of Xeno Fighters R. Seems more likely they'll be the former though.

[hermitC]

The kamikaze ship looks allright, I can imagine hoardes of them appearing from the top of the screen and aiming towards you, or even spinning and moving down while spinning towards you like Stage 2 of Xeno Fighters R. Seems more likely they'll be the former though.

Spinning sounds good! I will give it a try in a further demo version.

[hermitC]

Now you can get your hands dirty on Nordenfelt 0.1:

Download Nordenfelt 0.1

Tell me what you think. I'm looking forward to your feedback.

[BPzeBanshee]

I'm actually finding the boss impossible to defeat.

I also like how you've done the Kamikaze dudes, they're exactly like how I did an enemy for my first game project. More bullet types, variations in flying styles for the aircraft are a must though, they're quite boring as it is now. For that you should be able to get away with doing color edits of the existing sprites and doing a different path and speed pattern for them. It's also difficult to dodge considering there is no hitbox either (or at least it seemed to be like there was no hitbox).
For v0.1 this is good progress though, I had no issues running it on the best PC in my house (I should hope not since my family bought it in 2008). I'm interested to know though: Is this a C++ Program or is it made with something like Game Maker?

One other thing: Spotted this after I played the game:

Code: Select all

2010-Jul-05 14:15:41 | info:    Log system is running
2010-Jul-05 14:15:59 | error:   d:\air task\airtask\Registry.h(47): Throw in function class stable_ptr<class AirTask::MissionProgress> __thiscall AirTask::Registry::Get<class AirTask::MissionProgress>(void) const
                                Dynamic exception type: class boost::exception_detail::clone_impl<class AirTask::NotAvailableError>
                                std::exception::what: class AirTask::MissionProgress is not registered
                                [struct tag_function *] = __thiscall AirTask::EnemyShip::~EnemyShip(void) throw()
2010-Jul-05 14:18:15 | error:   d:\air task\airtask\Registry.h(47): Throw in function class stable_ptr<class AirTask::MissionProgress> __thiscall AirTask::Registry::Get<class AirTask::MissionProgress>(void) const
                                Dynamic exception type: class boost::exception_detail::clone_impl<class AirTask::NotAvailableError>
                                std::exception::what: class AirTask::MissionProgress is not registered
                                [struct tag_function *] = __thiscall AirTask::EnemyShip::~EnemyShip(void) throw()
2010-Jul-05 14:20:37 | error:   d:\air task\airtask\Registry.h(47): Throw in function class stable_ptr<class AirTask::MissionProgress> __thiscall AirTask::Registry::Get<class AirTask::MissionProgress>(void) const
                                Dynamic exception type: class boost::exception_detail::clone_impl<class AirTask::NotAvailableError>
                                std::exception::what: class AirTask::MissionProgress is not registered
                                [struct tag_function *] = __thiscall AirTask::EnemyShip::~EnemyShip(void) throw()

I think I saw the game lose a frame after going into the grass section but besides that the game ran perfectly and had the log not existed I would never have noticed these errors. Keep up the good progress.

[hermitC]

I'm actually finding the boss impossible to defeat.

I also like how you've done the Kamikaze dudes, they're exactly like how I did an enemy for my first game project. More bullet types, variations in flying styles for the aircraft are a must though, they're quite boring as it is now. For that you should be able to get away with doing color edits of the existing sprites and doing a different path and speed pattern for them. It's also difficult to dodge considering there is no hitbox either (or at least it seemed to be like there was no hitbox).
For v0.1 this is good progress though, I had no issues running it on the best PC in my house (I should hope not since my family bought it in 2008). I'm interested to know though: Is this a C++ Program or is it made with something like Game Maker?

One other thing: Spotted this after I played the game:

Code: Select all

2010-Jul-05 14:15:41 | info:    Log system is running
2010-Jul-05 14:15:59 | error:   d:\air task\airtask\Registry.h(47): Throw in function class stable_ptr<class AirTask::MissionProgress> __thiscall AirTask::Registry::Get<class AirTask::MissionProgress>(void) const
                                Dynamic exception type: class boost::exception_detail::clone_impl<class AirTask::NotAvailableError>
                                std::exception::what: class AirTask::MissionProgress is not registered
                                [struct tag_function *] = __thiscall AirTask::EnemyShip::~EnemyShip(void) throw()
2010-Jul-05 14:18:15 | error:   d:\air task\airtask\Registry.h(47): Throw in function class stable_ptr<class AirTask::MissionProgress> __thiscall AirTask::Registry::Get<class AirTask::MissionProgress>(void) const
                                Dynamic exception type: class boost::exception_detail::clone_impl<class AirTask::NotAvailableError>
                                std::exception::what: class AirTask::MissionProgress is not registered
                                [struct tag_function *] = __thiscall AirTask::EnemyShip::~EnemyShip(void) throw()
2010-Jul-05 14:20:37 | error:   d:\air task\airtask\Registry.h(47): Throw in function class stable_ptr<class AirTask::MissionProgress> __thiscall AirTask::Registry::Get<class AirTask::MissionProgress>(void) const
                                Dynamic exception type: class boost::exception_detail::clone_impl<class AirTask::NotAvailableError>
                                std::exception::what: class AirTask::MissionProgress is not registered
                                [struct tag_function *] = __thiscall AirTask::EnemyShip::~EnemyShip(void) throw()

I think I saw the game lose a frame after going into the grass section but besides that the game ran perfectly and had the log not existed I would never have noticed these errors. Keep up the good progress.

Good feedback, exactly what I wanted. Thanks a lot!

The game is fully written in C++ using BOOST, SFML and other third-party libraries. The missing MissionProgress is not really an error. I was just to lazy to handle it in a proper way I'm happy that you found this out. Good testing style!

[BPzeBanshee]

Good feedback, exactly what I wanted. Thanks a lot!

The game is fully written in C++ using BOOST, SFML and other third-party libraries. The missing MissionProgress is not really an error. I was just to lazy to handle it in a proper way I'm happy that you found this out. Good testing style!

Thanks, man. I been doing testing on various platforms for Thunder Force IV Rebirth as well as some experimentation with Xeno Fighters R using VirtualBox on Mac OS X. Eventually I'll put your game to the VirtualBox test as well though I considering what you've used to make it I can't see there being any issues.
For the next release, would it be possible to put a FPS and object-on-screen counter for a debug mode or something? Not really a necessity since the game clearly works using a recent computer but it'd be a good addition to work out the performance level of the PC required to play it. For example, my game project (built in Game Maker) runs at 60 FPS, and begins to lag on my oldest machine with about 380 instances on the screen while my Mac (dualbooting Windows, which I tested Nordenfelt on) has well over 5000 objects before it begins to lag. It gives me a clear minimum of just how far I should be going and considering the high quality 2D stuff you're putting into it so far you it may be a good idea in the future to find out the bottleneck of the program.
Also, I notice there's some lag involved at the boss when it animates - is it just me or do you get that too?

[hermitC]

FPS and object count are already available in test builds. So there is no problem providing them for you in the next version.

I'm noticing some lags to. They are on my todo list now.

[worstplayer]

No technical problems here (except that I can't move up and left when shooting, since spacebar blocks some aroow key combinations on my keyboard, use Z please).
Oh, and turn off antialiasing when rendering sprites. White "halo" around player's ship looks horrible.

[genecyst]

what is the minimum required system to play this game?

[BPzeBanshee]

what is the minimum required system to play this game?

At this point we're not sure, that's why I asked for FPS and object counter to work out an estimated benchmark. The game does not appear to be visually taxing (I mean that like fancy draw rendering and 3D tech) so I would guess it should run even on Legacy computers at this point.

[hermitC]

No technical problems here (except that I can't move up and left when shooting, since spacebar blocks some aroow key combinations on my keyboard, use Z please).
Oh, and turn off antialiasing when rendering sprites. White "halo" around player's ship looks horrible.

The "holding space blocks diagonal moves" problem seems to occur on other machines to. Thought it is just my laptop which has this problem. Does anybody know what's the reason for it?

Antialiasing is needed for sprite rotation. I will have to test switching it off for not rotated sprites because I don't know how mixing aliased and anitaliased sprites will look like.

The player's ship will be replaced anyway. The graphic style is still in experimental phase but the white edges won't survive

[rtw]

The "holding space blocks diagonal moves" problem seems to occur on other machines to. Thought it is just my laptop which has this problem. Does anybody know what's the reason for it?

Ghosting or rollover

http://en.wikipedia.org/wiki/Rollover_%28key%29