« first day (594 days earlier)      last day (2868 days later) » 

12:01 AM
I've got a question that's gone unanswered for about a week. What are the sanctioned ways of promoting it other than offering a bounty?
Begging, pleading, murdering, stealing
Is it the audio sampling position one?
I have 0 audio experience
if you use a separate channel, it still uses the same output buffer?
@BerickCook, yeah, that one
It sounds like you're essentially having to predict what audio will be played and when, well in advance of when it will actually be played, yes?
@Jimmy, the channels are my own design concepts -- they all use the same "output" buffer
@BerickCook, yeah that's the idea
12:11 AM
What kind of game is it?
@BerickCook, it's a shmup
Q: How should I adjust starting sample position of audio track to match next buffer fill?

stepheltonI've written my own audio mixer. For Linux/Windows, my game uses SDL. On a small handful of Linux machines (including my laptop), SDL has a bug where it allocates a ridiculously large (about 120 milliseconds) audio buffer. While I'm not to the point that I care so much about affected machines, it...

in case anyone else is following
120 milliseconds is an eternity in a game like that. Is it for all audio? sfx and music?
all (there's no difference in the two except that music is streamed from ogg format)
yeah, 120 ms is nuts =P
I can't think of an easy way to do it off hand. Especially considering the player can tap the fire button, pause the game, etc
Yeah, that's the kind of thing I keep running in to. The only solutions I've thought of would be pose performance issues
The way I see it, I can think of things as either samples or time. The two are interchangeable, but when things like pause occur, the natural unit of measure is time.
If the AudioManager basically kept a log of any event that affects the timeliness of audio (pause, increase / decrease playback speed) then i could replay through it whenever i needed to figure out exactly where a track should be
and maybe that's the best i'll do... not exactly simple, but probably cheap enough to not be a performance issue
12:20 AM
That's a lot of work for a special case bug
I admire your dedication
well it affects everything
the notice-ability is directly related to the output buffer size
But only on a "small handful of Linux machines"
i can tell even with a smaller buffer... in particular, if two of the same sound are fired at nearly the same time (two of the same gun being fired, for example), they can both end up playing at the next buffer-fill, which effectively doubles the sound of that one clip
that problem could be solved in a very different way, though
also, if i were to have a buffer underrun, this problem would really amplify the result
12:43 AM
Sounds like a fairly flawed system. Is there nothing better you could use to handle audio?
1:06 AM
Not one that would support all of my target platforms. Also, my system seems quite capable other than this issue. Fixing the issue is likely a lot less work.
And i'd miss the chance to learn something...
15 hours later…
4:14 PM
So, I've looked over all the questions here about level editors, and most are for tile based games, I found this one: gamedev.stackexchange.com/questions/8600/… for non-tile basted games, but the only editor linked has not seen an update in over 2 years. Has anyone else got any tips? Or is this the type of tool I would be best suited for building myself, and making it build the exact levels I need for my game?
I should point out that I am looking for a non-tile based editor :)
Is this the kind of thing you're looking for? gleed2d.codeplex.com
That's probably the project that hasn't been updated often eh?
September 2010
Yeah, gleed is listed in that question... :(
:/ yeah
Yeah, that is close to what I want...
there's a few games that have non-tiled editors in them, but there's not really any general case editor. you probably would be better off making a custom one for yourself, or at least expanding on an existing editor that's out there and is open source
4:21 PM
Gleed is open-source
That is what I was afraid of.
I was thinking about forking it and making it compatible with XNA4.
What's missing from gleed?
There's already a fork for that
I'll have to check that out.
hmm, if you fork a project on Github, doesn't it start a "new" project for you?
These look very attached to the main project
yes, they do
I'm not sure about github
4:25 PM
Well, on Codeplex, there doesn't seem to be a project page for this guy's "AnimationsPhysicsObjects" fork of Gleed
It's just a sub-page of Gleed
Yeah, it looks like it's tied to GLEED.
Not sure if that matters much... if it does what I need. I'm defiantly going to check that out.
yeah, cool
Steve Dunn has been updating Gleed on Github rather recently. I got some really basic stuff working with it. github.com/SteveDunn/Gleed2D/wiki
Wow that's really awesome.
I don't know how I missed that!
I've been poking around for ages, and I only recently got serious about it.
I only ever found the codeplex site and some old msdn forum posts linking to a broken site
yay! everybody wins!
4:44 PM
Is that a legitimate reason to run around GD.SE and update all the links to Gleed?
I just posted a comment to the original answer on the question I linked...
should we update the actual answer?
I donno. I'd vote yes because the content and meaning of the answer doesn't change
How does one decide when it is appropriate and when it isn't?
fair enough
The new github project even points back to codeplex
Actually.. there's only the 1 candidate for updating, and it's the one you added your comment to
5:04 PM
I believe it, but I'm surprised there aren't more links to that editor.
Actually... there could be a few more. Search fail
5:18 PM
Well... I've polluted the activity list with a few updates
always fun
alright, I've done my good deed for the day
Shoulda changed the titles to questions while you were at it. That bugs me, now I have to edit them all.
heh, yeah. I wasn't even looking at that
I'm glad we have more active users with editing powers
5:27 PM
heh, "Excavator" = earned
and yeah, I'm sure it helps to take some pressure off you and Tetrad
1 hour later…
6:29 PM
I was useful! :D
Yeah, the googlefoo is strong with this one.
I forget how I came across that but I was in the same situation as you for a bit and then it was like hey! Look at that!
I even reported a few issues that he fixed promptly which shows promise
That's great to hear. If I end up using it, I'll probably jump in on it too. So far it looks very promising.
7:11 PM
Q: When should I decide what language and platforms I will use for a project?

MikalichovDuring your development process, when is the recommended phase to decide what platform to aim for, and what language to use? I figured that it would be better to choose at the beginning what would be your target platform, as it would affect your design decisions, and limit the language you are s...

I like how this is a meta-question for "what platform should I use" but it's much less irritating
heh, true
Yeah, I wasn't sure I should bother answering that.
Especially since, imo, the answer is pretty obvious.
I faced that exact problem last summer. My personal opinion is very similar to your @JoshPetrie
If a broad language will cause the project to fail, you should pick a narrow language that you are familiar with
7:19 PM
This is also a bit sketch:
Q: How do I get the source code from a Google Code game project?

BluFireI'm trying to get the Hedgewars source code. When I went to the downloads tab, it doesn't specify which is the actual game. I tried downloading it using the SVN Checkout on Tortoise, but it seems like it doesn't work on the browse section of Source. (Hg>project_files>Android_build>SDL-android-...

hah, yeah. I was like rly?!
I like his comment on my answer (which he accepted): "thanks but I get an error." Full stop.
yeah, nice
I hope he didn't copy it by hand
ho wait, he said he hg cloned it
hg clone https://hedgewars.googlecode.com/hg/ works perfectly fine for me
7:27 PM
hg was "unresponsive"
He's got it now
technology... sigh
Some of these open source projects take a lot to build yourself though. For example, Mumble's source scared me away
oh, is that all.
You probably have half of that stuff installed and configured for IsoTower
it's the open source projects that require a sacrificial animal that scare me
7:33 PM
hmmm are there good dependency managers out there
only thing i have installed for Isotower is SFML libraries, and the GCC compiler :P
and Code::Blocks, I guess
You don't have Boost?
I thought all C++ developers used Boost
i don't. i think it's got some useful functions, but it's also got a lot that I don't need.
i'd rather keep dependencies low
7:36 PM
Boost is not one library, it's a huge collection. The dependencies can be as small as a single header.
most of the really useful libraries have to be built separately, though
I just keep a copy of all the binaries I need instead of building dependencies every time
@thedaian Is that a problem?
but then again, I'm distributing dlls instead of statically linking code
shrugs probably isn't a huge problem. it's probably not that much different than using SFML or SDL or anything else, really
i think the last time i tried looking at the boost libraries, i was confused and thought that all the boost libraries were included as part of it
7:42 PM
Yeah. You're right that less dependencies is a good thing, it's just that Boost is too good to pass up.
Boost isn't a library, it's a philosophy :P
that too. i am a rebel, against the tyranny of boost
shrugs there's really not much I need boost for. smart pointers might be useful, and the multiplatform file system thingy seems useful, too, but beyond that, SFML provides most of the same library bits, and everything else, i can pretty much work with
and the only thing i really feel is missing (super easy, javascript/python string manipulation) doesn't exist in boost anyway
and by "missing" i mostly just mean "it would be really nice to be able to write Console.log(new std::string("this " + 45)); and have that magically work
Smart pointers are in std now, don't really need Boost.
yeah, if you're using the fancy shcmancy new version of C++
Also, why not Console.log() << "this" << 45;?
7:55 PM
hmmm, i wonder if that would be feasible...
Just have log return an ostream. Or you can have it construct a temporary which works better for multithreaded logging.
the thing is, Console.log() outputs whatever string is passed to it to my in-game console
You can use stringstream with temporary then
8:16 PM
i'm assuming it would work like how iostream's stuff works, but i have no idea how those functions actually work...
8:38 PM
Here's a quick example: ideone.com/blF9i
Actually, that doesn't flush right.
@Jimmy I have no words for this picture.
yeah, that's kind of what I thought, it ends up requiring another function to wrap around the output
@Jimmy that would get it to flush right :P
9:25 PM
@Pubby your class seems fine to me but that's not anything like how I would implement a logger.
well, I don't think he's really implementing a logger, just implementing a function that can do console.log() << "this" << 45;
10:22 PM
Hey, guys. I'm building my first component-based system engine, using the outboard system approach.
Could you take a look at my classes and tell me if I'm headed the right way?
Here's http://thunked.org/p/view/pri/b7onr08ba (Render System)
http://thunked.org/p/view/pub/lmsu6u7gq (Physics)
and http://thunked.org/p/view/pub/43i9sm4nq (Entity stuff)
10:35 PM
@Damir I am not much of a C++ programmer, but what I see does not make my eye bleed, so for whatever that's worth...
@Nate You've worked on component systems?
@Damir Not really, I really meant (for better or worse) what you've got doesn't look bad to my eyes, but as I'm really not an expert, that may or may not be a valuable bit of information.
@Damir I've done some work with XNA, which is somewhat component based... that's about the extent of my experience with it.
@Nate Oh well, Thanks for taking a look, Nate :)
@Damir No problem. Good luck. My philosophy is that if what you have is working, run with it, because you'll get somewhere. Don't worry about building a perfect system.
10:53 PM
@Nate Yeah, that's a good tip.
Though I tend to get carried away :D

« first day (594 days earlier)      last day (2868 days later) »