New Sauerbraten SP map - need feedback & mapmodels

by Morosoph on 11/28/2006 17:49, 11 messages, last message: 12/07/2006 14:33, 2030 views, last view: 05/04/2024 18:49

As my map's not quite complete (it needs playtesting and mapmodels), I've not yet posted it at quadropolis.

I've taken a few risks. If anyone wants to try it, I've posted it here:

http://www.geocities.com.nyud.net:8080/tim_wesson/

If that address doesn't work, try stripping out .nyud.net:8080 (Coral Cache) from the web address.

Feedback would be appreciated. Is the labyrinth too hard, for example? I have less time now, so might not be able to build the mapmodels (I need to get to grips with Blender), but it would be a shame to leave it unreleased on that basis.

The mapmodels that I need are two halves of a stone tablet, a convincing sceptre, a crown, and a large jewelled cup.

   Board Index   

#1: Oops!

by Morosoph on 11/28/2006 17:51

I forgot to say. If anyone want to give me a hand with the mapmodels, you can share the limelight :o)

reply to this message

#2: Re: Oops!

by da-old-mitaman on 11/28/2006 23:20, refers to #1

Hey there! Good to see another person working on an sp map.

I did a quick run thru and I'd like to give some observations:

1. Watch those frame rates. In the large open area with monsters it dropped down below 10fps. Large maps work ok in mp mode, but when monsters are added, they "suck-up" lots of frame rate. Consider breaking up the map into several pieces (smaller map sections instead of one large map). I had to do that with my ARMY OF ONE map to make it playable.

2. Concentrate on adding detail with in-game geometry and textures, not just mapmodels. I know it's a WIP, but the map has an overall "boxy" look to it.

3. Give the player some "surprises" like monsters triggered behind you, or traps, not just a room full of monsters that overwhelm the player.

4. Always give the player some "breathing room" to enjoy looking at the map. Sort of mini-breaks between each battle.

Hope you find this helpful. SP maps take lots of thought and planing. It should be more that just constant onslaughts of monsters.

My 2 cents worth - MitaMAN

reply to this message

#3: Re: Oops!

by Pxtl on 11/29/2006 02:05, refers to #2

Hmm - I wonder why monsters slow the game down so much? It's not like they have horrifically complex AI or freakishly detailed models. Anybody profiled the monster code in those situations? They have some confused pathing or something?

reply to this message

#4: Hmm.

by Morosoph on 11/29/2006 03:30

10 FPS is a major problem. My machine is over spec. for this kind of task (Opteron with nVidia 6600), I'll have to develop on an older machine, as I'm not seeing these frame rates, even at 2048x1536.

Also, strategy should avoid facing too many at once (don't take the teleport that is straight ahead at the start until those monsters are cleared. At least jump in with a Quad boost!) Stick your nose in, then retreat. Let them come after you.

It would be a shame if I had to seperate the level: it is meant to be open plan. I'm not sure how to break it up.

The oasis gave a big problem on a friend's machine. Perhaps I should just remove monsters from that area?

The beginning is more boxy than the rest (guess what I did first!) I'll have to think on have to improve that. The game design improves as the game progresses, although I'll have to confess: no original textures. Another friend suggested that some of the lights in the labyrinth should be "broken" so as to give the player some feedback as to where they are.

I'll take in feedback; I posted this in part because I'd stopped refining the level. Maybe if I get a second wind...

Thanks for the feedback, MITAman!

reply to this message

#5: Mass Or Class

by MeatROme on 11/29/2006 12:08

Some nice basics in there ... maybe you should have concentrated on just one/a couple of the scenarios you've sketched into this - fleshed it out true and full ... then you just concatenate these single-theme-maps in a row of SP maps.

In case of your concept this isn't all-too-easy, but you can get somewhere close w/o much a-do. Anyway - with just spawnpoints to save your progress, the vast amounts of dreary sameness of space to cross in your map and the amount of time required to find all triggers in that big labyrinth ... well ... how would you define your target audience?
"I-play-for-42h-straight-each-day" !?!
:-P

But from the work done you could copy-n-paste into a newly aligned central map and do a couple of inside maps going out from there - compare AdamParker-Day1 and the ReVamp-Experiment (both on quadropolis.us) for ways to handle this "the sauerbraten way".

HTH

reply to this message

#6: Thanks, MeatROme

by Morosoph on 12/02/2006 00:27

My life's become a lot busier, so I don't promise anything any time soon. However, my level is released under a Creative Commons licence, so anyone can run with it, as long as they give me fair billing!

At a minimum, I won't release the level on quadropolis until those mapmodels are done. However, I'm not stopping others from adapting sections of the level, or indeed the whole thing. I may even do this myself :o)

As for "no saves", I agree that is a problem. I started the level (default size) in order to get some practice at level design, and continued until I filled the space! Along the way, I made it into a bit of an...epic, including its Indiana Jones moments.

The labyrinth has a fast way through it, tripping all of the levers, and the level solved in order doesn't involve too much desert-crossing. But I conceed that these have to be learnt through exprimentation. The level takes me about 1H 20M to solve, so it's not exactly short...

I found a bug that after about an hour's play, some of the projectiles stop working. Or maybe it's after the presence of too many monsters at once (eg. the cave). They stop working for monster and self alike. The guns work fine, as do granades, but not the rockets. I ought to submit this, but I will need to reference my level as proof. That would be one reason to complete the mapmodels to release Enigma of The Desert on quadropolis.

reply to this message

#7: too long

by kurtis84 on 12/02/2006 15:40, refers to #6

Aside from the fact that it's just too big, it also takes too long to play. As others have mentioned, this really needs to be split up into a few smaller maps. IMHO, an sp map should not last more than 30 minutes.

Don't try to use mapmodels to fill in for lack of details...build everything from cubes, until the map pleases your eyes, THEN add in some mapmodels.

These are just my opinions...do what you feel is best.

reply to this message

#8: Mapmodels

by Morosoph on 12/02/2006 16:20

I hear you, kurtis84, and everyone else.

I ran away with this one. I think that it makes a good "epic", but you won't want to be playing it too often.

The mapmodels are simply for plot: two half-tablets in the dunes, filled with runes. A sceptre in the caves. A crown and Grail at the end. They won't affect gameplay, but only completion as a concept (a great quest).

I agree that this level doesn't have the Sauerbraten "look and feel", but I do feel that its size is it its own way a virtue. Perhaps this level could be combined with a "save game" mod?

Taking part(s) of this level to make a smaller level(s) is a good idea, and I might do this, if I can find the time.

I have licenced the level so that anyone else who might want to do this can do so. Anyone who wants to make a derived work has a relatively free hand (by-nc-sa).

I'm grateful for all the feedback, BTW, and I'll certainly keep it in mind for any future levels!

reply to this message

#9: Re: too long

by spentron on 12/07/2006 02:53, refers to #7

I liked it a lot more than I expected from this thread. The size is cool in a lot of ways, although parts are too plain. The cave work was very impressive, I didn't know Sauer could be made to look that natural, curvy and anti-grid.

I got a steady 9 fps, it didn't even seem to improve as I killed monsters. Still playable, but partially because baauls, ogros and knights were disarmed.... Monsters have a lot of polies and I think are poorly occluded -- a smaller number of smarter monsters could perform better, the AI isn't much of a load even with lots in the current design (I'd like both kinds).

I think 1.5 hrs for a SP level can be good if it keeps you busy the right way, but savegame (save on exit?) becomes important unless you don't mind leaving your PC and the engine idling 24 hrs between sessions. This I went on more like 3 and finally gave up. Find-the-key puzzling needs to be limited to avoid the 90s-ish/RPG-ish make the game go on unnaturally longer syndrome.

The size could allow some interesting monster techniques, for example wake up monsters or teleport them in on the far side of the level and you still wouldn't see them for 5 minutes. See "monster machines" on my cube page and levels for some ideas ...

BTW I haven't pointed to my Cube 1 page lately for the newer folk, I have four (included one collaboration) SP levels in the last release but for the rest you'll have to look at http://www.intergate.com/~spentron/cube/index.html

reply to this message

#10: um

by Aardappel_ on 12/07/2006 08:15

did anyone get past "the pyramid" without cheating? I had good fun until that point, and tried finding a way out to the point where I actually had the maze memorized, but had to give up. I only flipped each switch once, don't EVER make puzzles that require a certain number of flips from multiple switches out of view, that is just diabolic.

reply to this message

#11: Re: um

by Morosoph on 12/07/2006 14:33, refers to #10

Aardappel:

You flip 'em all to green to open the first door, and red to open the second.

Perhaps the doors should stay open?

I thought that the labyrinth might be too hard. Maybe I'll litter a few more clues around the place (I lit the doors to the main rooms red and green for a reason).

I agree that the labyrinth is the most contentious part. That you need to drop down a hole to get from the lower maze to the upper is another factor...

Interesting fact: apparently the (real) pyramids were made out of concrete!

http://www.geopolymer.org/category/archaeology/pyramids/

I'll have to admit that I got carried away designing this. The bits I most like from the level are those that I made last: the caves and the castle. Perhaps I could reuse those elsewhere...


spentron: Thanks your comments! The caves took a lot of work. I hand-tuned everything, and if you fly inside the caves, you'll see that I shaped there are hidden gaps along the exit passage that allow more flexibility in the visible shape (close the gaps to see what I mean).

reply to this message

   Board Index