KungFuSquirrel

Not that there were many, but Prey was the best game on the Doom III tech (though Brink might technically take that mantle since it's still an evolution of that engine). Absolutely worth playing if you haven't. Can't wait for this one.

Very impressive environments. Not a fan of the animations

It looks like the traditional problem of FPS multiplayer animations - when they're just moving normally they can look pretty good, but bits where the legs are spinning around fighting the aiming of the torso... it just doesn't work right. Some of the moves they pull off still look pretty great, though.

Color me impressed.

http://www.pcgamer.com/2011/03/10/hawke ... -megacity/

My definition of streaming content is when I don't encounter a loading-screen for a very long time, which is what I noticed in Singularity

That's too narrow a definition when everyone's still streaming levels like mad You can only exist in the same persistent level for so much game.

Still hardly anyone takes advantage of the streaming content feature in the engine. The only somewhat recent game I can recall using it is 'Singularity'

Wait, what? Outside of UT3 and maybe some other MP-only stuff that keeps everything in the persistent level, I'd be shocked if there's a single game that's not using some amount of level streaming (and textures can stream automatically).

hey look at all those really obvious assets placed in the street.

Oh hell. Is it because cars are placed perpendicular to the flow of traffic?

Got what you mean by angled roads now.

Oh, no, no, that wasn't a comment on that - when I looked at the top-down, I completely missed the cluster of assets that should have made it clear to me that you were using the street and not the elevated side walkways

ah, I'm dumb, I was misreading the layout (hey look at all those really obvious assets placed in the street. ) I was for some reason picturing you spending more time on top and crossing over the street. Completely off base, then, don't mind me

(but I think the angled road suggestion might still help things a bit )

I realize this may be a little late to suggest it, but why limit yourself to a grid of streets? Even just taking a chunk of the layout and pushing it over at a 1:2 slope might help give you some of the visual doglegging you need. I don't know how deep it is, but what if you had a ruined chunk of elevated walkway that required you to go into a building and down into the street section, which could be a dark, claustrophobic mess of wrecked cars and debris and give you much more controlled sight lines - consider also the read of seeing a landmark/eventual destination poking up in the opening above the street, between the walkways. And of course the entrance and exit to that lower level provides a pair of very natural streaming transitions

It's been a couple years since I've been in UE3, but if I'm remembering correctly the move was best handled by 1) copying the objects to the streaming level, 2) cut/paste the kismet to the streaming level (which will then find the same objects from the original level), and 3) delete the original objects.

I'd recommend somewhere in the realm of B and C, though; things can get complicated real fast if you try to hack it up later. Depending on what you're using and how, you might also be able to alleviate some annoyance by keeping some of your scripting/script-relevant objects in the persistent level or a dedicated scripting sublevel separate from the geometry. Your mileage may vary, but I do recommend trying to plan for it to whatever extent is possible as early as possible to avoid stress later.

Bump!

I think we had enough support in the voting thread to do both the door challenge and the 3x 512 texture challenge (or some variation of a texture-limited one), but I had another thought I wanted to toss into the ring for future consideration:

The Design Test Challenge:

You have two weeks to build a 5-10 minute single player level in a given theme that incorporates at least one agreed-upon element (one that I did a while back, for example, required some example of physics-based gameplay). Themes ideas would be collected in advance and then one would be drawn randomly on the day the challenge begins to ensure no head starts

It's a bit high pressure, but won't take much time off the calendar and might be excellent preparation for those among us who are considering applying for jobs and may be taking tests in the future.

Feel free to toss out any other ideas you might have; we can always look at them down the road

Welcome!

I gotta say, great work on Vorkuta in particular. I'm not all the way done with Black Ops yet (I'm getting my ass kicked in Kowloon, most recently), but I think that level and the first chunk of Baikonur are my favorites.

I wrote up some thoughts on the trailer: http://blog.button-masher.net/2011/02/1 ... -of-games/ {shameless blog promotion, eh? )

As great as the trailer is, I really don't get the sense the game's tone is going to match. Which, I mean, the game sounds good, but I think I'd rather play the type of game that could come from the trailer than the real thing.

haha, what a comeback! 1024 is the winner, followed by the Doorway Challenge in second and the 512 texture challenge in third.

I still think there's enough support to line these up in sequence, so once the new 1024 has ended we can roll straight into the doorway challenge, etc. (after an appropriate couple-week breather, of course). Sound good?

I'll get a 1024^3 rules/discussion thread up shortly, though I think this is going to be a super easy to agree on.

Terrible.

Also can I please reiterate to keep the challenges well spaced apart (with adequate warning) and with 3-4 week timescales. I really want to participate but life is very busy right now, im sure it is for others too.

Absolutely. I could see possibly running another 3 week challenge on the door, but I'd support 4 for the 512 texture and 1024^3 challenges. I'd say the absolute minimum between smaller challenges would be 2 weeks, but a 3-4 week break is probably still the safest.

If the vote is close enough, we might as well just line up all three in succession over the next few months

the 1024 box feels like it's been done before.

It has; we ran it here in 07. Numerous other sites have done it too. But, it's a classic!

Let's vote! I grabbed the top 3 challenge suggestions from the discussion thread, as I imagine they'll still have the most support.

I'll let this poll run through the weekend, then we can sort out start times/rules/etc. Let's try to keep things a little simpler this time, eh?

I think the earliest we'd start anything would be Feb. 19th - that gives us almost a month since we wrapped the 20 Brush Challenge, and hopefully will be enough buffer time for everyone.

Discuss!

Edit: I suppose I should describe the three a bit better...

1024^3:

- Build an entire level contained within a 1024x1024x1024 cube

3 512x512 textures:

- Build a level using a maximum of 3 512x512 textures

Doorway gameplay:

- Using a pre-built door frame, come up with and script a clever way to get through a locked door.

You know what, a 10th-anniversary re-make with updated visuals and modern XBL support? I'm totally fine with that.

Post it in finished work!

Here's a brutal one for you - Triathlon. First week, Source/HL map. Second week, Quake map. Third week, Unreal map.

Speedmap stuff would be great to try again, but it's hard to sustain things like that. I wonder if we might have better participation/results by mixing those sessions in between larger challenges?

Please, suggest away! Small and manageable is always good.

I think the 1024 was one of the best we had, and I also have an interest in not failing to complete something like I did for the last one That might be a good one to roll out as the next one, but I think we should let things settle for a couple/few weeks before we go after anything else.

Sentura brings up valid points on voting; and I think we should try to simplify voting in the future. I also feel like more of an "approval voting" style might be better for future challenges - if we allow people to vote for up to, say, 3 entries, and then tally based on those totals. Voting for one in each category was hard enough, and voting for just one in a single category might be torture

Hey guys, just a reminder to cast your vote for the 20 Brush Challenge if you haven't already. We got some great entries including some great work by some new faces. Submissions are here: viewtopic.php?f=57&t=15843 with links to the voting threads.

I also wanted to see how people thought about trying to run more of these again? I know it's something we've done in the past with mixed results, but I had a blast on this even if I did just build 20 floating hexagons, and I think there's some fun options to run new ones. Some thoughts I've had:

1024 Challenge - resurrecting an oldie but a goodie: build a level inside a 1024x1024x1024 cube

Geometry challenge - All geometry, flat texturing. Like the classic Q3 GeoComp

Old School challenge - HL1, Q1, Unreal, etc.

Art Pass - similar to what Valve ran for TF2

I think in the future it might be neat to have guest judging panels and the like (I'm sure we can all find plenty of people to help out), but I do like maintaining the "Challenge" part of the title vs. "Competition" - no prizes, just being creative and working within fun limitations.

Any other thoughts? Good idea/bad idea?

...and moved this one over, too!