« first day (803 days earlier)      last day (4233 days later) » 
01:00 - 19:0019:00 - 21:00

Code Assassin
Code Assassin
1:00 AM
breaks silence with unnecessary text
 
 
1 hour later…
michael.bartnett
michael.bartnett
2:09 AM
ponders the syntax for italics
and guesses correctly
 
Code Assassin
Code Assassin
win
 
ToddersLegrande
ToddersLegrande
sneaks into the room with super italic skills
 
user4704
Combo breaker!
 
michael.bartnett
michael.bartnett
Super Italics Fighter
 
ToddersLegrande
ToddersLegrande
IV Arcarde Edition?
 
Code Assassin
Code Assassin
2:27 AM
Posting screens in a few minutes ^-^
 
michael.bartnett
michael.bartnett
sweet
Any Unity gurus in the house?
 
ToddersLegrande
ToddersLegrande
I've dabbled, definitely no guru, but I can pretend I might be able to answer
Actually... I've probably spent a total of 8 hours in Unity :P
 
michael.bartnett
michael.bartnett
That probably puts you in at least the 90th percentile :P I'm just curious if the only way to get surface normals of a collider is really by Collider.Raycast or Collider.ClosestPointOnBounds. They don't provide that data in the collision event data
 
ToddersLegrande
ToddersLegrande
I just don't understand half of these words xD
lol
Sorry
 
michael.bartnett
michael.bartnett
Lol no big, A for effort
 
ToddersLegrande
ToddersLegrande
2:32 AM
It's really just surface normals though. I've dealt with very little graphics related stuffs out side of 2D textures.
I got colliders working though! :D
 
michael.bartnett
michael.bartnett
Lol, good job! They're surprisingly tricky beasts
 
ToddersLegrande
ToddersLegrande
Yeah it felt silly what exactly had to be set
AND THEN
I could no longer use my position += method of movement and had to switch to a force based one -_-
 
michael.bartnett
michael.bartnett
Yeah lol, things go south fast if you modify transform.position directly. Unless rigidbody.isKinematic == true, but then you don't get OnCollisionEnter messages with static colliders :P
 
Code Assassin
Code Assassin
 
Raven Dreamer
Raven Dreamer
Anyone here familiar with Unity? I'm trying to draw a procedurally generated mesh, and my texture is not rendering (just a gray blob)
 
Code Assassin
Code Assassin
2:36 AM
@ToddersLegrande Have no idea what that means - but sounds awesome!
 
ToddersLegrande
ToddersLegrande
Hmmm.... I feel like I didn't get OnCollisionEnter to work but OnTriggerEnter (or whatever the trigger one is called) did work
@CodeAssassin We should team up and make a Twin Stick Moba together :P
 
michael.bartnett
michael.bartnett
@CodeAssassin Cool stuff, is the glowiness baked into the sprites?
 
Code Assassin
Code Assassin
@michael.bartnett yep
 
ToddersLegrande
ToddersLegrande
League of Legends, Dota, Heros of Newerth, Demigod, etc. in the style of a twin stick shooter (like geometry wars)
 
Raven Dreamer
Raven Dreamer
@ToddersLegrande Aha, that sounds brilliant.
 
michael.bartnett
michael.bartnett
2:38 AM
@RavenDreamer I haven't done much with procedural models and materials yet. Are you sure all your material props are being set correctly?
 
ToddersLegrande
ToddersLegrande
@Raven its an idea I've had for about a month now. I'm not sure how much fun it would actually be. I figured I'd make a 1 player version to test the waters and then go from there
 
Raven Dreamer
Raven Dreamer
@michael.bartnett Well, I have the same material on a regular cube next to it, and the one has the texture, (in the proper colors) and the dynamic one doesn't.
 
Code Assassin
Code Assassin
@ToddersLegrande cool - I would love to do it :)
 
Raven Dreamer
Raven Dreamer
@ToddersLegrande Does it being a two-stick shooter preclude button-based actions?
 
ToddersLegrande
ToddersLegrande
@RavenDreamer Nah. I figured I wouldn't even make it shoot automatically. Hook up default attack to a trigger or the A button or something. And then put special abilities in the other buttons
 
Raven Dreamer
Raven Dreamer
2:40 AM
oh, wait, maybe I just need to
 
ToddersLegrande
ToddersLegrande
@CodeAssassin do you have gtalk or anything like that?
 
Code Assassin
Code Assassin
@ToddersLegrande I have gtalk [email is cod3assassin@gmail.com]
 
Raven Dreamer
Raven Dreamer
if I'm not using lighting or anything, what are the "default" texture coordinates?
ah, blast. That didn't do it.
 
michael.bartnett
michael.bartnett
@RavenDreamer Are you assigning the material via an asset reference, or constructing a new material?
 
Raven Dreamer
Raven Dreamer
@michael.bartnett Asset reference.
public property of the script
drag and drop the material
 
ToddersLegrande
ToddersLegrande
2:49 AM
ლ(ಠ益ಠლ)
 
michael.bartnett
michael.bartnett
@RavenDreamer tried alternating between using renderer.material and renderer.sharedMaterial?
 
Raven Dreamer
Raven Dreamer
Oh, I'm not using a renderer, I'm making a graphics call directly.
> Graphics.DrawMesh(mesh, Vector3.zero, Quaternion.identity, pubMat, 0);
That might be it.
 
michael.bartnett
michael.bartnett
Sadly that's out of my knowledge scope. Should know more about that in a month or so, not that it'd be useful to you by then.
 
Raven Dreamer
Raven Dreamer
Ah, well, thanks anyway.
Ah, yes, it was with the UVs.
'course, now I got the texture on the insides of the cube. XD
Intruigingly, this also seems to have answered the question I was in the process of posting on the site. Double success!
 
michael.bartnett
michael.bartnett
3:29 AM
Nice!
 
 
3 hours later…
shihab_returns
shihab_returns
6:29 AM
@All: Good morning...
@all is here anyone to have knowledge in andengine..?
 
Gajoo
Gajoo
6:43 AM
you can ask, but I'm not sure if anyone can answer
 
michael.bartnett
michael.bartnett
sorry shihab, never done andengine, or any android. good luck.
 
snake5
snake5
andengine.. hearing a lot about it lately..
 
Gajoo
Gajoo
it's written in java, that's enough reason for me to call it bad!
 
michael.bartnett
michael.bartnett
i tend to agree, but the only other option on android is C++ with NDK. i like C++, but not the sound of NDK/JNI nonsense
 
snake5
snake5
I've come to think that no language in itself is bad, it's mostly about the crowd that is gathered around it
what's so bad about NDK?
 
michael.bartnett
michael.bartnett
6:58 AM
it's just tedious, IIRC, you've got to set up all the external calls and spend time managing your java/C++ boundaries. i haven't used it, so maybe it's not so bad, but i wouldn't relish working with it
 
snake5
snake5
what external calls?
even if it's required to setup function pointers in opengl style, there surely has to be a library that does it
 
michael.bartnett
michael.bartnett
the library setup itself may be a chore.
DllImport works really easily in C#, but that doesn't mean it's not a pain to use
 
shihab_returns
shihab_returns
7:16 AM
@michael.bartnett thanks for reply..
 
Gajoo
Gajoo
@michael.bartnett cygwin is the the only problem with NDK.
 
michael.bartnett
michael.bartnett
it requires you to use cygwin?
 
Gajoo
Gajoo
@michael.bartnett if you are going to create a game the only external call will be calling your main function
@michael.bartnett it used to require about 9 month ago, but I'm not so sure how it works right now.
a friends of mine migrated to linux just for that reason.
 
michael.bartnett
michael.bartnett
@Gajoo hopefully not anymore then. and for NDK there's no Activity class nonsense to worry about? im asking from having zero android experience.
 
Gajoo
Gajoo
try cocos2d-x
it might take a while (maybe a week) to get it compiled for the first time. but after that everything will run smooth
 
michael.bartnett
michael.bartnett
7:26 AM
then you're good to go on android and ios, right?
 
Gajoo
Gajoo
and windows and linux and mac.
also black berry if I'm not mistaken
for example I'm working on a game for iOS, in VS2012.
every week I just compile everything on an iMac device to make sure everything looks ok, and usually all debugging stuff takes less than an hour
 
snake5
snake5
anyone got any interesting problems for me?
 
michael.bartnett
michael.bartnett
7:43 AM
that is sexy, I'm going to have to try out cocos-2dx once i get a break from the main project
 
Gajoo
Gajoo
@snake5 udp data transmission for my game?
 
snake5
snake5
@Gajoo what about it?
 
Gajoo
Gajoo
I need my multiplayer support!
 
snake5
snake5
7:59 AM
..and?
where's the problem?
 
Gajoo
Gajoo
I don't feel like implementing it myself
at least not right now, and I don't know when...
 
snake5
snake5
hire someone then :P
 
Gajoo
Gajoo
you were asking if anyone can offer any job!
do you know what game I'm talking about?
 
snake5
snake5
8:31 AM
@Gajoo no, I was asking for interesting problems
jobs rarely offer any such problems
@Gajoo no
 
michael.bartnett
michael.bartnett
40
whoops
that's the line of my current compiler error
ctrl-g ftw
 
michael.bartnett
michael.bartnett
8:47 AM
@snake5 I've got a problem for you. given an array of vector3 velocities, most of them will have a similar orientation, but there will be one or two entries with outliers. how do you filter out the outliers?
 
snake5
snake5
find the average normalized vector in pass 1, check for outliers using a dot product in pass 2
simple
 
michael.bartnett
michael.bartnett
Awesome, that will work way better than this delayed queue thing I've got going on. 5am is a bad time for problem solving :P
 
Gajoo
Gajoo
9:13 AM
@snake5 still how can he filter outlines?
I mean average vector is not exactly in the same direction as the crowd
@snake5 battery.intellisenseless.com you can check older (2D) versions to play against bots.
 
snake5
snake5
9:38 AM
@Gajoo the crowd is guaranteed to have most vectors similar to others in the crowd, the dot product returns cos(angle) with 2 normalized vectors as input - no need to have the same direction, just similar
 
Gajoo
Gajoo
@snake5 the problem is how do you define similar?
 
snake5
snake5
@Gajoo angle under some threshold, most probably
 
Gajoo
Gajoo
it won't really help, I'm affraid
 
snake5
snake5
it appears that it did already help
 
Gajoo
Gajoo
 
snake5
snake5
9:44 AM
..so?
 
Gajoo
Gajoo
in this case, both blue and red vector are outliners
but the average is much more closer to blue vector compared to others.
 
snake5
snake5
no, they're not
there is no crowd in the first place
 
Gajoo
Gajoo
@snake5 crowd are black, yellow, and purple vectors.
 
snake5
snake5
outliers = <10% of data
probably even less
therefore your example is wrong
 
Gajoo
Gajoo
@snake5 how do you know about that?
also you must specify the threshold, before you can call your algorithm complete
 
snake5
snake5
9:46 AM
not at all
threshold depends on data
 
Gajoo
Gajoo
I mean you should define how one can extract threshold using data.
 
snake5
snake5
@Gajoo because that's how things are defined
 
Gajoo
Gajoo
for example you might say threshold is variance(data).
 
snake5
snake5
@Gajoo it is practically impossible to extract a threshold
 
Gajoo
Gajoo
@snake5 he didn't mention 10%,
 
snake5
snake5
9:48 AM
@Gajoo so? the exact percentage doesn't matter
 
Gajoo
Gajoo
@snake5 then it's impossible to implement your algorithm, we need a threshold in the process.
@snake5 it does, in my example outliners were 40% total.
 
snake5
snake5
@Gajoo it's not impossible
threshold is a value that you set depending on your needs
@Gajoo your example is wrong
 
Gajoo
Gajoo
@snake5 what's exactly wrong in there?
 
snake5
snake5
@Gajoo too few vectors to define a crowd
this feels like a chicken-and-egg thing
 
Gajoo
Gajoo
I can't understand what you are trying to defend. your algorithm is not wrong, it's just incomplete.
or inaccurate.
in some cases with some special conditions it's complete though.
 
snake5
snake5
9:55 AM
it's not incomplete
and what you call "special conditions", I call "the general case" or "the only case worth handling"
 
Gajoo
Gajoo
for example only if all the vectors are scatters with gaussian noise I can accept your algorithm.
 
snake5
snake5
what you drew there, that's a special case
@Gajoo you've misunderstood the original problem
 
Gajoo
Gajoo
@snake5 that's exactly why I'm saying it's incomplete, it can't handle special cases
 
michael.bartnett
michael.bartnett
You guys crack me up. I really feel obligated to implement this now.
 
snake5
snake5
@Gajoo that's not a reason for an algorithm to be incomplete
an algorithm is incomplete only if it cannot be implemented from the description
 
Gajoo
Gajoo
9:58 AM
it can either solve all cases or it can't. being complete means it can work for whatever input data you might have
 
snake5
snake5
@Gajoo yeah, right... so a json parser should also support all javascript code?
 
Gajoo
Gajoo
@snake5 also you didn't provide any measures for threshold value, it also implies incompleteness.
 
snake5
snake5
I've got tons of examples where this isn't true
 
Gajoo
Gajoo
@snake5 it doesn't, since JS code is not part of it's problem.
 
snake5
snake5
@Gajoo what measures? it depends on data which means "adjust until works"
@Gajoo ...
 
Gajoo
Gajoo
9:59 AM
but if JS interpreter can't load JSON data, it's called incomplete!
 
snake5
snake5
it's what I said about this algorithm - your special case is not part of its problem
hey, why can't your game work when I unplug my computer? it's incomplete!
 
Gajoo
Gajoo
@snake5 and why do you think so? 3 out of 5 is most of things in my measures!
 
Martin Sojka
Martin Sojka
To chime in a bit: I wouldn't use the average vector direction to determine the direction of the crows as a whole; I'd use the median (for each coordinate).
 
Gajoo
Gajoo
@MartinSojka that's way more better solution, except how do you sort data?
 
snake5
snake5
@MartinSojka there is no median
unless you quantize the angles
which is silly
 
Gajoo
Gajoo
10:02 AM
@snake5 if they were 2D vectors (which they're not) median was the most accurate solution (IMHO)
 
Martin Sojka
Martin Sojka
@Gajoo Natural sort for each coordinate; I don't mean dir = median(x, y, z) but dir = (median(x), median(y), median(z)).
 
Gajoo
Gajoo
@MartinSojka I'm not so sure it'll work, but it's worth trying, at least it works for my test case
and any other test case I can think of
 
Martin Sojka
Martin Sojka
@snake5 What the hell has median to do with quantization?
 
snake5
snake5
ah, sorry, mixed it up with mode or something
either way, what's the threshold?
 
Martin Sojka
Martin Sojka
@Gajoo It'll tend to generate values around (0, 0, 0) for any crowd which "explodes", even if it does so slightly unevenly, which I consider a bonus. :)
 
Gajoo
Gajoo
10:08 AM
@snake5 he doesn't need any threshold,
 
snake5
snake5
why not?
 
Gajoo
Gajoo
he compares and checks for exact equality.
 
snake5
snake5
...
that requires most vectors to be exactly the same
that's a special case
not the general one
 
Gajoo
Gajoo
1 hour ago, by michael.bartnett
@snake5 I've got a problem for you. given an array of vector3 velocities, most of them will have a similar orientation, but there will be one or two entries with outliers. how do you filter out the outliers?
 
snake5
snake5
"similar"
did you read it or just copied?
 
Gajoo
Gajoo
10:10 AM
I assumed similar means equal
 
Martin Sojka
Martin Sojka
@snake5 The median method is just to calculate the direction of the crowd as a whole. The threshold would still be calculated against the dot product of it with the crowd members.
 
snake5
snake5
@MartinSojka thought so
 
Gajoo
Gajoo
still variance (dot(data,median)) is a good threshold
 
snake5
snake5
so uh... what's the implementation of the median algorithm?
 
Martin Sojka
Martin Sojka
.. and I'd suggest making three pre-sorted lists (one by each coordinate) for that, if space is not an issue. You can eliminate a lot of checks by just looking at the fringes of it, and it helps with median calculation.
 
snake5
snake5
10:12 AM
3 sorted arrays of references to data?
hmm, I think I have an idea there
 
Martin Sojka
Martin Sojka
You can calculate median without sorting in one pass (examine each data as it comes and decide if you need to move the median depending on into which bucket - "less than" or "more than" - it falls), too.
 
Gajoo
Gajoo
just one notice, you must normalize all the vectors before sorting, or else things won't work out real good
 
snake5
snake5
no need for buckets, I think
 
Martin Sojka
Martin Sojka
@snake5 Well, in practice you'd just record the size of them. :)
 
snake5
snake5
I had the same idea but I wonder if it'll work correctly
what if the next passed vector is a huge leap over all data of the opposite side
 
Martin Sojka
Martin Sojka
10:15 AM
After a while, your median won't move much anymore, so you won't need to check which value to exchange it with so often.
@Gajoo The point is to not normalise it. People who run in the same direction, but way slower or way faster are not part of the crowd.
 
snake5
snake5
@MartinSojka "similar orientation" means you have to test for that, not "similar orientation and speed"
 
Gajoo
Gajoo
@MartinSojka it really depends on the problem specification, in his case, he wanted to eliminate those who walk in different directions.
 
snake5
snake5
from what I see, there has to be normalization
@Gajoo what's "variance()"?
 
Martin Sojka
Martin Sojka
@Gajoo If it's for a crowd separation, he'll need to separate the stragglers eventually. :)
 
Gajoo
Gajoo
sum(dist(M - Xi)^2)
@snake5 also there is wiki
 
snake5
snake5
10:21 AM
@MartinSojka why ^2?
that distorts the value
 
Gajoo
Gajoo
actually I meant sqrt(variance) but I couldn't remember the term used for that :|
 
snake5
snake5
:/
 
Martin Sojka
Martin Sojka
@snake5 The "^2" serves two purposes: for real-valued variables, it makes it so the deviation is independent from the direction of it (no negative values cancelling positive ones), and so that bigger deviations are given more weight.
 
snake5
snake5
cool
but I don't see how all that gives something that can be used as a threshold in the end
 
Martin Sojka
Martin Sojka
It's a measure on how spread out the directions in the crowd are. The lower the standard deviation (= sqrt(variance)), the more vectors are pointing in roughly the same direction.
 
snake5
snake5
10:31 AM
I understand that
but unless it's a measure in angles or distance or something real, it can't be a threshold
 
Martin Sojka
Martin Sojka
10:53 AM
@snake5 The standard deviation (also called "sigma") is of the same type as the values it is calculated on. For dot product, it means it's in cos(angle).
 
snake5
snake5
ah, good
 
 
5 hours later…
KitFox
KitFox
4:09 PM
Hello, hello! Writer's Chat, starting now in The Overlook. We're discussing how to do character sketches today.
 
snake5
snake5
4:35 PM
moderator spam, I suppose? :D
 
Code Assassin
Code Assassin
4:45 PM
-> Insert 'IsThatEvenLegalFace' <-
 
Stack Exchange
Stack Exchange
0
Q: Guidelines on asking game industry questions

ApoorvaJA lot of non-programming questions on this site are closed because they are "broad" or "open-ended". Game design is not a science and hence there are bound to be multiple answers. There is an informative thread about posting questions about game design questions here Similarly, the game industry...

discussion questions
 
snake5
snake5
illegal operation exception occured at .. here :D
 
Code Assassin
Code Assassin
lol
 
draconis
draconis
morning
 
michael.bartnett
michael.bartnett
5:11 PM
ohai
 
user4704
@Byte56 that publisher question is the only one I'm aware of as well
 
Byte56
Byte56
Same.
 
user4704
gamedev.stackexchange.com/questions/tagged/game-industry is full of stuff, but if anything I see more questions there that should be closed rather than the other way around.
 
Jimmy
Jimmy
what's a good way to phrase the publisher question?
 
ToddersLegrande
ToddersLegrande
Drop the reading materials list for starters
 
user4704
5:20 PM
Yeah, that's never a good sign.
 
user4704
For me, the big issue with that question beyond the 'reading material' issue is that the high-level operations of a publisher are basically already listed in the question.
 
Byte56
Byte56
Whenever it's just a yes or no answer and it's up to the answerer to elaborate usually isn't a good sign either. "Do today's publishers fulfill any other functions?"
Yes, depending on the publisher.
 
user4704
And consequently, the question is either "yes... and a list of context-specific details," or "no, and..."
 
user4704
...basically what Byte just said.
 
user4704
A question that was "what is the role of the 'publisher' in the industry" might be answerable with the same (or similar) list of high level operations already present in this question.
 
Noctrine
Noctrine
6:02 PM
Trying to properly phrase a game design question is more difficult than I thought :\
 
Byte56
Byte56
We don't pay you to be a mod because it's easy :)
 
Noctrine
Noctrine
Read as: We don't pay you to be a mod, because it's easy.
:p
 
snake5
snake5
:D
 
Byte56
Byte56
Yeah, just a semantics.
 
user4704
I imagine the daily tasks of a mod to involve (1) log in, (2) BAN ALL THE THINGs
 
user4704
6:07 PM
This is all I do at GDNet these days.
 
user4704
So it must be the same here.
 
Noctrine
Noctrine
Banning all of the things makes people complain on meta :\
 
user4704
Delete post.
 
snake5
snake5
someone always complains
 
Byte56
Byte56
(3) COVER UP ALL THE BANNING
 
user4704
6:08 PM
Hellban user.
 
Noctrine
Noctrine
There is no winning with users
 
user4704
I'm pro at this, trust me.
 
snake5
snake5
@Noctrine as soon as it's not productive enough, they stop
 
Noctrine
Noctrine
But the question was actually one of my own. But really it seems like a survey would be more appropriate, but no pleasing everyone so a survey would probably not work very well either :\.
Finally getting back into my virtual game now that work has calmed down a bit. Debating whether people's pets should die or not.
 
Byte56
Byte56
Yes.
They should.
But the Pet Semetary should be an option.
 
Noctrine
Noctrine
6:14 PM
Well if I went that way, I'd likely include a challenge or something that lets users bring back previous creatures and then something else that lets them stop them from dying. But I get the feeling that it will lead to people focusing more on building their personal character and neglecting the pet mechanics outright.
Really pitting my like of permadeath against marketability
 
Byte56
Byte56
I also like permadeath. But it does depend on the market you're targeting.
Those pansies and their fluffy games don't like permadeath.
 
Noctrine
Noctrine
its a slightly more advanced pokemon, more like the original digimon world (which was set up where creatures would die after some time, and would be prone to dying faster if they were mistreated)
 
Gajoo
Gajoo
is it possible to feed two textures to a pixel shader?
 
snake5
snake5
of course
 
Gajoo
Gajoo
then how can I do it in OpenGL ES?
 
snake5
snake5
6:18 PM
use different samplers
 
Gajoo
Gajoo
and how should I bind them?
 
Byte56
Byte56
I never played pokemon or digimon. Could you lose your pets to death?
 
snake5
snake5
glActiveTexture for the textures and there was some function to set uniform sampler id too
 
John McDonald
John McDonald
@Byte56 Not in the Pokemon I played a long time ago
 
Noctrine
Noctrine
@Byte56 In Pokemon no, but in the very first digimon world you could.
 
snake5
snake5
6:20 PM
@Gajoo more info here: opengl.org/wiki/GLSL_Sampler
 
Byte56
Byte56
@Noctrine Well, if it's kind of for people who played those games and want a more advanced version, I think permadeath would be an acceptable addition.
 
Noctrine
Noctrine
The better you were at raising them the better your creatures got in their next cycle (it was more like reincarnation). But it helped to space out gameplay, the world was generally pretty small. But when your creature died you had to grind the lower level areas again building it up before going after the bigger challenges that you cleared your way to earlier.
 
Gajoo
Gajoo
@snake5 you sure it works with OpenGL ES2?
 
Noctrine
Noctrine
@Byte56 Ideally I'd like that market as well as the younger market.
 
Byte56
Byte56
Well, there's always the option of turning on permadeath in the advanced options or harder difficulty levels.
 
Noctrine
Noctrine
6:24 PM
Its kind of a weird online multiplayer thing, like Phantasy Star Online, so that could create balancing issues.
Maybe if there is an incentive created for allowing them the potential to die?
 
snake5
snake5
@Gajoo not sure, but 1) most things should; 2) doesn't hurt to try
 
Byte56
Byte56
Right, permadeath server has better/faster levels or something like that.
It would kind of separate the noobs and the advanced players too. Some people may appreciate that.
 
ToddersLegrande
ToddersLegrande
@Noctrine Permadeath was pretty awesome in Monster Rancher (IMO) - you had to balance pushing your monster as far as it could go but careful not to let him die because the trick to really advance in the game was breeding. Combine two monsters that (should) result in a third monster better than the first two.
@Noctrine is it a persistent MMO like WoW or is it online multiplayer more like Diablo? If its the latter it could be like Normal vs Hardcore mode
Just separate communities
 
Noctrine
Noctrine
Not exactly, trying to not the scope too much. Single Player and Co-Op modes, mostly just mission based with some exploratory stuff. Then community chat/playrooms and battles.
It's already feeling far to large from just saying "Multiplayer would be a nice addition"
 
michael.bartnett
michael.bartnett
6:42 PM
is it bad form to pimp your question out on here?
 
melak47
melak47
depends. what are you paying?
 
Noctrine
Noctrine
They generally pop up in the ticker for people who are interested
 
michael.bartnett
michael.bartnett
even if they're a couple days old?
@melak47 considering a bounty :P
 
Noctrine
Noctrine
12 hours :p, but that means it's come and gone.
 
michael.bartnett
michael.bartnett
12 hours then they float down, or float up?
 
snake5
snake5
6:48 PM
what are the most efficient modern terrain rendering methods?
I'd implemented that mipmapping thing a year ago, there wasn't enough detail
far from games like BF3 (or even BF2), Fuel
 
melak47
melak47
@snake5 well, the modern thing seems to be to do tesselation and/or displacement mapping, and stuff
 
01:00 - 19:0019:00 - 21:00

« first day (803 days earlier)      last day (4233 days later) »