« first day (2730 days earlier)      last day (2294 days later) » 

5:34 AM
0
Q: How to ask this question, which has a 7-year-old duplicate?

JayI want to post a question about creating a "game UI" with javascript and HTML5 canvas. I'm not sure how to keep it on topic because It's basically a recommendation question, It's already been asked and answered here. The existing question was asked 7 years ago (active 4 years ago) and the w...

 
 
3 hours later…
8:47 AM
 
 
5 hours later…
user92578
1:26 PM
 
user92578
Had to be done
 
user92578
Should probably add some waves or something while it's transitioning
 
nwp
Looks pretty good.
It is an opportunity to make speedrunners happy or unhappy, if you care about that.
Although I wouldn't know if a cycle that you can barely make or a cycle where you can relax for a second would make them happy or unhappy.
I think the first would make them happy because it lets them show off their skill while the second is just safe and boring.
 
user92578
Yeah I try to design levels so that if the player plays well, they can catch some cycles
 
user92578
I do believe there are some levels in the desert area where that's not the case though
 
1:36 PM
@Tyyppi_77 fantastic!
@Tyyppi_77 speedrunners will figure out how to break it
 
 
1 hour later…
2:48 PM
Played with mining a few cryptocurrencies, it really isn't worth it...
But I am playing now with Folding@home, which at least if I don't get anything, I know that I'm helping to do research with it.
 
3:23 PM
@PearsonArtPhoto yup, really not worth it. Unless you find a new one before the difficulty goes up AND you get lucky and it survives long enough for hte price to go up.
 
Yeah, I'm really realizing that.
The thing that kills me is I joined a pool to look for VCash, and I actually found 6 blocks.
But I haven't been able to cash in because I didn't hit the min balance.
And the difficulty has spiked since I started by a factor of ~100 or so...
 
3:45 PM
:D
your best bet on VCash is staking or incentives, but one is 0.7% annual and the other requires you hold at least 10k
I think RDD stakes at 5% and there is no more mining, so you could try to buy some of that cheap and stake it. Plus RDD stakes based on how long it's been since your coins have staked, not how long the client has been running. so you can run the client for a few hours once every two weeks and get your coins.
Beware anyone in crytpo who is too gung-ho. they're probably invested, underwater, and desperately trying to get others in to help increase the price.
"DASH could go 100x in the next month!" shit like that
 
Yeah.
CURE has a 4% stake, so...
I think I'll just use a bit of real cash and buy a few things.
 
nwp
I guess as long as you see it as a learning opportunity with some amusement it's fine. Just don't see it as an investment.
 
Yeah, that's pretty much what I'm doing.
With Folding@home I'm lending my computer to do Alzheimer's research, and making a bit of digital currency to boot.
And I wouldn't put any "real" money into it, just some Christmas cash that I don't know what to do.
 
nwp
Did you buy subnautica and slay the spire yet?
I bet they have a much higher fun/money ratio than crypto currencies.
 
LOL
I have a huge backlog of games to play already, so...
 
4:00 PM
:D
 
 
1 hour later…
5:26 PM
Anyone knows what's the use of a lock-free atomic variable as opposed to a non-lock-free in c++?
 
Thread safe nature?
 
nwp
21
A: Lock-free multithreaded programming?

Kerrek SBThe key in lock-free programming is to use hardware-intrinsic atomic operations. As a matter of fact, even locks themselves must use those atomic operations! But the difference between locked and lock-free programming is that a lock-free program can never be stalled entirely by any single threa...

They are both thread-safe. The difference is in performance, progress guarantees and implementation difficulty, but I'm fuzzy on the specifics.
 
@nwp Thanks!
I guess in my situation I can use either of these. 2 threads, read/written to once per frame..
or something like that.
 
nwp
I can tell you that you definitely don't want to implement this yourself. Supposedly even the great experts figure out after 5 years that their lock-free algorithm was never correct.
 
Oh, no, don't worry, it's for a BOOL/bool :P
 
nwp
5:37 PM
I think facebook has implementations worth looking at.
 
So I'll use std::atomic_flag
 
nwp
Just use std::atomic<bool>.
Should be the same thing but more expressive and consistent with other types.
 
@nwp The implementation is lock-free then?
 
nwp
It should be for any reasonable platform.
I don't think any implementation actually uses mutexes for std::atomics even though they should.
 
I gathered that they use underlying OS method calls for that instead..
 
nwp
5:43 PM
No, they are simple assembler instructions. And there is a high chance that you are on an x86-compatible machine where you get so many guarantees it doesn't even matter much if you use atomic<T> or T.
Check godbolt.org if you are curious what it compiles to. You should be pleasantly surprised.
 
Ah, yeah, ok. Well for the sake of documentation, using atomic<bool> seems the way to go, even if it doesn't have much impact on the code execution per se.
They don't have visual studio :P
 
nwp
They do.
Just need to figure out which year corresponds to which compiler version.
Wait, I swear they do.
 
Ah, yeah, didn't realize it was under the name 'cl'
 
(sorry, wrong textbox o_o)
3
 
nwp
x86 MSVC 19 2017 is probably the right one.
 
5:51 PM
Yep, latest version of visual studio.
My asm is old and rusty, but the atomic version still seems to have more stuff going on.
 
nwp
-O3 or it doesn't count.
/O2 for VS
 
That's optimization?
Arg
hold on
 
nwp
Yeah. Looking at performance of unoptimized code is not useful.
 
Yeah, I changed browser and forgot to update that
Lol, func2 is just obliterated :P
 
nwp
@AlexandreVaillancourt You can mouse over an instruction to get an explanation of what it does.
 
6:05 PM
Sweet!
 
nwp
Your functions don't return a value which might make the compiler go "It's UB anyways, may as well throw the code away".
You can use -O3 -Wall -Werror -pedantic to catch most of that.
 
Yeah, that's most likely what happens.
 
nwp
Then again sometimes C++ compilers do some impressive valid optimizations.
 
Yes. That's why it's no longer worth it to try and outsmart the compiler. ("I'll ++i this loop index instead of i++ it.")
I guess I'll have to try an actual implementation to see the difference!
 
6:23 PM
anyone here work with android emulators?
mine started crashing if I don't use software rendering, and that's fucking slow
I didn't change anything; it just stopped working one day
 
The default Android emulator is pretty poor.
Genymotion is a good choice, but it might not work well for games.
I think there might be i86 emulators you can use that might be a bit faster too.
You can always delete the emulator and start a new one, it might fix the issue.
 
ok i'll try that
this is an x86 emu
making a new one still doesn't work
must be some video card driver shit
man i hate PCs
the android 8 one fails, too (the other one is 7.1.1
 
6:40 PM
Genymotion is my biggest suggestion for Android emulators.
 
ok i'll look into that, thanks
I don't think I can use it; this is at work and they'd have to buy a license
looks good though
 
@Almo Heh.. some companies can't do math.
How much will it cost to buy the license? 100$. How much will it cost for our dev to wait for the thing to run? At least 200$.
(I'm just pulling numbers out of my head here.)
 
That's right, I've always used the free version...
But seriously, it's worth the license. Talk to your manager.
 
@AlexandreVaillancourt yeah
Failed on eglChooseConfig
Could not initialize emulated framebuffer
emulator: ERROR: Could not initialize OpenglES emulation, use '-gpu off' to disable it.
What makes me so mad is this was working last week
 
6:55 PM
How big is the company you work for?
 
around 400
 
$400 a year then for Genymotion. A bit pricey, but still...
 
i use it so seldom it wouldn't be worth
i can get by with the software rendering, it's just annoying
 
Yeah, that is true...
Android is really finicky, to say the least...
 
hahahah
to say the least
 
6:59 PM
I haven't played much with it recently, but I did a bunch with the early days.
It used to be a lot worse...
I'm not fully sold on Android Studio yet, but I suppose I haven't really given it a chance.
 
IGDA Developer Satisfaction Survey 2017 -- Summary Report (PDF): bit.ly/2017IGDADSSResults
4
 
7:38 PM
 
0
Q: How should we handle questions asking for specific legal advice?

Josh PetrieWhile questions that touch on the law are generally on-topic here at Game Development, there are subsets of that topic that I think we should clarify our position on. In this case, I'd like to discuss questions that ask for specific legal advice. By that, I mean questions that lay out an asker's...

0
Q: Should we remap the "legal" tag?

Josh PetrieContinuing the trend of law-related meta questions... is there anything we can do to the legal tag to provide better guidance as to its use and the kinds of law-related questions that are acceptable here? Or is it fine as it is?

 
@DukaZhou where art thou?
 
8:21 PM
@TheMattbat999 lololol (re: software developer)
@TheMattbat999 busy these days trying to work out the basic [M] game trees, demonstrate the 2x2(2x2) solution, and show how that solution applies to mxn(jxk) for all even m, so if you don't ping me directly, I may not come.
 
@DukeZhou it is true though. At least for me.
 
Are you a cannibal coder, like myself?
 
@DukeZhou I have been working on an DOS styled RPG lately. It is actually coming along pretty good for the progress I have made.
 
PS-- have you downloaded the Mbrane app yet? I'd really like to know what you think. People are saying either (a) the game is too hard, or (b) the game is too easy
(this makes me happier than I can say;)
@TheMattbat999 when you say "DOS styled", what do you mean specifically?
 
@DukeZhou pixeled and the overuse of billboarding.
 
8:45 PM
ic. sounds good!
 
8:56 PM
@DukeZhou Also will includes alot of content we planned for Dark Dungeon.
 
I'll definitely play it when it's ready, assuming it will run on Mac...
definitely interested to see how you're implementing the ideas
 
9:08 PM
@DukeZhou not much yet, but I don't think I will have the planned world size.
 
9:19 PM
Definitely a good idea to keep the scope as narrow as possible at first, to facilitate project completion.
 
nwp
What do you mean project completion? What is that? I've never done that ...
 
@TheMattbat999 (PS- you don't have to ping me once I'm actually on the chat--was just meaning to say I haven't been checking in regularly lately.)
@npw lolol
 
9:58 PM
Ok.
 
10:45 PM
Hello :) Who invited me to chat? :)
 
nwp
Destiny
 

« first day (2730 days earlier)      last day (2294 days later) »