Incline SHELTER update thread

[shihonage]

I've been reading Fallout3 LP and noticed that the writing in the final product was worse than the placeholder writing we had in Shelter before scrapping the level with that character.

______________________

shipped version of Fallout3:

______________________
pre-alpha version of Shelter:

 

[soggie]

Shihonage, I got a question. Are you going to show different equipment on the characters? How are you going to achieve it? Render every permutations or use layered sprites?

 

[shihonage]

It is very likely I'll have to use layered sprites.

 

[soggie]

Facing 6 directions? What about using 3d models instead for characters and NPCs?

 

[Kosmonaut]

n/m

Huh?

 

[shihonage]

Facing 6 directions?

8.

What about using 3d models instead for characters and NPCs?

Not doing that. However when it comes to your project feel free to make your own decisions. Fallout Online 2238 uses 3D models now, they may be a more "progressive" role model.

Huh?

Nothing; I was just gloating about Fallout 3 writing compared to Shelter pre-alpha writing... but then got self-conscious and erased the gloating

Nothing self conscious about that. It was an entertaining read.

Retroactively restored!

 

[soggie]

Facing 6 directions?

8.

What about using 3d models instead for characters and NPCs?

Not doing that. However when it comes to your project feel free to make your own decisions. Fallout Online 2238 uses 3D models now, they may be a more "progressive" role model.

Huh?

Nothing; I was just gloating about Fallout 3 writing compared to Shelter pre-alpha writing... but then got self-conscious and erased the gloating

Nothing self conscious about that. It was an entertaining read.

 

[shihonage]

This week has been a mess of work and personal issues, and not much has been done at all. We've fleshed out a couple more of plot/setting/endgame points of the game and one combat mechanic (sort of), and that's about it.

Work time seems like a great opportunity to work on Shelter, but there's only so much I can do when a Pavlovian "startle" reflex kicks in upon hearing the screeching of my manager's chair, amd my hand jumps to the familiar ALT-TAB combo. From work, I cannot sustain focus on 23114 lines of code... most of the time.

Anyway, the aforementioned combat mechanic would reduce accuracy when you change targets (or move), and regain it when you stay on the same target.

Hope to pick up speed now. Next update should be more interesting.

 

[shihonage]

* We have outlined a near-exact plan on how to finish Shelter. This plan will involve a feature freeze and cutting of things that don't work. And our sprite selection will be really... how do I say it without putting e negative spin on it... "SPARTAN". Hey, Spartacus was cool, right? Cough.

The game must be finished, and it will be finished - on an undetermined Friday (an undetermined Thursday was already taken). The plan is to have Shelter appealing enough to help us make a sequel with more experience and better budget.

Shhh... don't tread on my dreams

* I upped the minimal horizonal resolution to 1024. Shelter no longer runs in 800*Y resolutions. The minimal resolution for Shelter is 1024x600, which coincides with a number of netbooks.The optimal resolution is 1024x768.

* I made a workaround for a DirectX bug which crashed Shelter at resolutions exceeding 1600 pixels width. If you're into microscopic sprites, you can run it in any resolution your monitor allows. But you will need a serious 3D accelerator for that - onboard will crawl.

P.S. I really hope ATI fixed the bug in 9.5+ Catalysts which made DrawText() calls in DirectX8 applications slow everything down to a crawl. Otherwise we're screwed.

* After wasting many hours (and sad tears) I finally made Shelter restore itself properly after an ALT-TAB or Ctrl-Alt-Del ! Now it's a "mature" application!

* I started preparing to implement many different missing engine pieces in random order. Now I am working on game saving. Other things pending are "shift-clicking quantity of items", barter, cover system, character attribute screen, handling leveling up... oh, and encounter system isn't finished yet.

May Morbo help us all.

 

[MasterSmithFandango]

* We have outlined a near-exact plan on how to finish Shelter. This plan will involve a feature freeze and cutting of things that don't work. And our sprite selection will be really... how do I say it without putting e negative spin on it... "SPARTAN". Hey, Spartacus was cool, right? Cough.

Obviously you have to make those decisions or you end up with... too much shit, but any ideas on what type of things you're planning on cutting? Also, slightly disappointed by fewer sprites, obviously, but you have to do what you have to do.

 

[shihonage]

We don't plan to go all Vogel and eliminate walking animation, but we plan other shortcuts. Nobody sane will make 20,000 unique person sprites for us, so we have to get inventive.

The game will definitely NOT look as good as FO:T did, but it will run at higher base resolution (everything 28% bigger).

As for features... we'll burn those bridges when we run into them. For instance, I'm not sure we'll bother with low-intelligence dialogue options, or in-game time/calendar management. Anything non-vital may get cut.

 

[MasterSmithFandango]

Well for a small project, I wouldn't worry about it looking as good as FO:T (which was a beautiful game). If you cut out the low-int options, I'll honestly be disappointed, but at the same time I get it.

Keep fighting the good fight, BRO.

 

[shihonage]

Thanks.

This is why I'm considering cutting it out:

One of the unique qualities of Fallout and Fallout 2 was the difference in gameplay caused by creating a low intelligence character. If you have a low intelligence, many NPCs will simply shrug you off, while others will insult or even attack you. It is possible to complete the game as an idiot, but most side-quests are impossible to do, and character growth is limited, as is the story involvement possible - while you can certainly bumble through the main quests, almost all plot details and background information will be left missing. Furthermore, due to limited-at-best NPC interaction, almost all side-quests are unavailable. Because of this, playing as a low intelligence character is only recommended for experienced players looking to role-play, or as a humourous replay option.

Efficiency. Effort/Result ratio.

 

[soggie]

Twiddles finger waiting for the next update...

 

[eklektyk]

looks nice
progress in development gives hope for positive conclusion

 

[soggie]

looks nice
progress in development gives hope for positive conclusion

Speaking of which... shihonage, how many months have you invested into the game in terms of development?

 

[shihonage]

It's hard to say. Sometimes I spent 15 minutes a day on it, sometimes hours.

It was born in January 2000 (I can't believe that site host is still up). Shifted focus several times from Shelter to being a sequel to Monsterland back to Shelter back to another project (counter-terrorism game).

I have somewhere an iteration of it as being an overhead view shooter with working coop (and terrible handling).

Shifted back to Shelter and became isometric in 2005. Shifted to Direct3D sprite interface in 2007. NMA picked up our first video at the end of 2007...

 

[Unradscorpion]

Heh, Monsterland, you linked it before, I should give it a spin sometime.

 

[soggie]

It's hard to say. Sometimes I spent 15 minutes a day on it, sometimes hours.

It was born in January 2000 (I can't believe that site host is still up). Shifted focus several times from Shelter to being a sequel to Monsterland back to Shelter back to another project (counter-terrorism game).

I have somewhere an iteration of it as being an overhead view shooter with working coop (and terrible handling).

Shifted back to Shelter and became isometric in 2005. Shifted to Direct3D sprite interface in 2007. NMA picked up our first video at the end of 2007...

Still a one man show for now?

You wrote the engine right? Any plans to help out the FIFE team or go open source yourself?

 

[shihonage]

Still a one man show for now?

For programming, yes.

You wrote the engine right?

Laundry day. Nothing clean. Right.

Any plans to help out the FIFE team or go open source yourself?

I like the Carmack formula. He doesn't mind people modding the games from the start, but releases source only after they've become technologically obsolete.

I like that formula. Minus the releasing source part. Because, Shelter competes technologically on a level that does not get obsolete as quickly as graphics. In fact, most games appear to be going backward on that level.

Fallout 3 engine, for instance, doesn't appear to be able to handle "catch the water thief", "assassin pops up in the middle of Killian dialogue", "raider takes girl hostage" type of situations from Fallout 1.

 

[shihonage]

One of the features tested recently was the ability to _beat_ information out of a character using the non-lethal combat system, and complete a quest that way.

But, it revealed a jarring lack of consequence for beating them up.This is one of the reasons why I don't know if non-lethal combat can actually stay.

I upgraded the script system, so now people now can permanently _refuse_ to talk to you after you beat them within an inch of their lives. It is checked through a highly intuitive ShelterScript line: [GETMINDSTATE[SELF`M_MYBITCH]=TRUE]|BLURB[SELF`Go away, mean Player]|STOPTALK

This upgrade also allows the famed "barkstrings" to work properly, i.e. when walking around a settlement, you click on some "Loser" NPC, and they will blurb something that may have relevance to a recent event (or mudcrabs).

DISCLAIMER: no mudcrabs in Shelter.

This week I've been cleaning up the setting once again (some elements were extraneous garbage). I also fixed an elusive bug (this is a frequent occurrence).

Now I am finishing the loader for random Encounter system. Such mundane tasks comprise 90% of development and are about as exciting as my real job in a software company. I should really consider a change of profession to something much more lively and exciting, like, being a coroner or something.

Anyway, Encounters are circles of various radius on the BIG MAP, and they have different "weights". This system only works when there's more than one encounter overlapping the same map cell, and so it shall be. Otherwise, the lowest probability encounter will keep happening because there's no high-probability counterweight.

Calculations involved are similar to the test they gave me at Flagship Studios interview in August 2006, which involved calculating random loot drops with memory and CPU constraints of a 1980's electronic watch.

Let's finish this update with a random screenshot. Look ma, items on the ground!

 

[MasterSmithFandango]

Is it done yet?

 

[hoverdog]

DISCLAIMER: no mudcrabs in Shelter.

 

[Unradscorpion]

...[GETMINDSTATE[SELF`M_MYBITCH]=TRUE]|BLURB[SELF`Go away, mean Player]|STOPTALK
...
DISCLAIMER: no mudcrabs in Shelter.
...I should really consider a change of profession to something much more lively and exciting, like, being a coroner or something.
...

 

[shihonage]