This will help developers prioritize between performance and new/better functionality(as always, of course, this data is merely an indicator that will be considered for decisions, not anything “binding” us to do this or that.)
(This thread earlier held the actual survey, but someone pointed me to google survey, so thanks for that!)
2.) How fast does gameplay feel to you in 0.7.1, SDL build?
I noticed a perceptible slowdown compared to my previous 0.6 build. Travelling long distances feels more tedious in this build.
3.) Would cataclysm still be playable to you if it were slower?
A further drop in speed would definitely impact my enjoyment of the game.
4.) Would you be willing to trade functionality for speed, in particular:
a) View range
I’d sacrifice, say, a 20-25% drop in view distance for significantly higher speed. Could this be made adjustable in the options menu, to allow us to find the best trade-off?
b) Accuracy of what you can and can not see, or how well you can see something
I’m not quite sure what this means in the context of an ASCII game. It’s not as though you can drop texture detail levels.
5.) Do you have a multi-core CPU? If so, would you like cataclysm to make better use of your additional cores?
If it’ll help performance and isn’t too technically difficult, sure. No point having cores sitting there doing nothing.
Not incredibly important, but if you e.g. take this scenario: http://i.imgur.com/QFamZ3M.png Imagine the lines on the right where stuff is “blacked out” to be much more “blocky” and having more artifacts.
@CIB: For polls like this, you might want to think about using Google Docs (or Google Drive now). Participants don’t need to signing in or register and, more importantly, it would allow you to keep track of the results way better. Here’s one of the numerous guides about how to set up one.
Just submitted my responses there, but wanted to expand on one of the questions found there.
For me, performance is one of the top priorities in any game. No extra cool feature justifies a game becoming a slog to play with or, in extreme cases, for it to become unplayable. Although I’ve made suggestions around here that would inevitably have a negative impact on performance, I’d be happy to not have any of them if implementing them would mean that we’d go back to, say, the horrible performance after 0.6 pre-SDL (for Windows).
Also, and maybe I’m wrong here but this is my impression, I don’t think that trying to simulate virtually everything (with all the calculations and keeping track of things required) is that important. Creating a world that gives the illusion that it simulates everything is more than enough. In my opinion, that is. So, persistence, for instance, shouldn’t be mandatory for all the features (such as NPCs or critters), as long as they are there and behave properly when the player is around. Etc. Basically, and as far as I’m concerned, feel free to “cheat” with some features for the sake of performance.
As I view it, CataDDA should always be the light on resources but heavy on entertainment game that has been (mostly) so far.
The trouble is that it can get pretty important. Image this setup:
legend:
= wall
w wood
f fire
player base
====w===
fwwwwww
vs this setup:
player base
========
fwwwwww
Just one tile and it makes all the difference of whether your base camp is set on fire or not. So yeah, being “inaccurate” can be nice and all, but for some things it can be pretty devastating.
Yes, I can see that happening. But there’s probably other instances where you could, say, cut corners with persistence without any dramatic effect on gameplay. What I was saying was just that, if you see that opportunity with any feature, go for it.
–
EDIT: BTW, for those curious about the results of the survey, you can check them using this link. By default, you can only access them once you complete the survey. But maybe you just are curious about how is it evolving or finished it but didn’t save the bookmark.
I for one think that the game could cut performance and there wouldn’t be too terribly much of a difference. It is a turn based game after all. And the more things simulated, the better the game will get. Realism can greatly help a RL. Especially on things that can kill you, the more accurate they are in calculations the better.
First question in the poll has weird answers… Has the creator never heard of pentium 4’s at 3ghz? It is not a dual or quad core specific feature. Most last single cores were 2-3ghz such as mine. It makes a huge difference from 1ghz to 3ghz. Please redo and simply ask how many cores do you have. You wont see a dual core under like 1.5ghz
It did make a difference for me. I got rid of windows and installed linux just to get better performances on CDDA. The difference was HUGE. I wouldnt be able to play with the performance of windows version anymore. It was a horrible experience for me. Realism is okay but if the game starts to get slow to a point that you only do a turn per second when yure trying to move across 2000 tiles (2000 secs you do the math how lng it would take :p) it wouldnt be enjoyable at all for me.
Yeah, the game’s playable for me when walking on a move-by-move basis at the moment, but when driving a car, with continuous movement for several ‘steps’ between each ‘turn’, the duration taken for each of those steps becomes an issue. Right now it’s okay, but if the game’s speed were halved, I don’t think I could consider it playable anymore.
Take the GHz as the main factor for that particular question, with the descriptions being as a guideline. For cores take a look at the following question.
Take the GHz as the main factor for that particular question, with the descriptions being as a guideline. For cores take a look at the following question.[/quote]
Yup thats what i did. Anyway it could be interesting if multiple cores could be used for different tasks in the game like fire spreading or anything AI related. I think it could be interesting for a ascii game to use the technology of having multiple cores. My desktop only has a single core but i think the era of single cores is about to end, all recent games requires dual cores and even phone have dual cores now (some even have quads). It could help with game performanceif the team have the skill to code the game that way.
Uh, part of the reason some folks play open-source (hi!) is that it’s free…and that helps when you also don’t have the budget for the latest hardware. Requiring multi-core processors would keep me out.
(Granted, some of you may consider that a good thing…)
[quote=“KA101, post:17, topic:2694”]Uh, part of the reason some folks play open-source (hi!) is that it’s free…and that helps when you also don’t have the budget for the latest hardware. Requiring multi-core processors would keep me out.
(Granted, some of you may consider that a good thing…)[/quote]
There’s a difference between “requiring” and “leveraging”.
No application requires multiple cores, unless it is extremely poorly written; it’s all about multithreading and leveraging the multiple cores if present.
And things running slower on older machines, that’s just the nature of the beast.
This thread made me try out the SDL version, which is quite a bit snappier than the console version on my system (64bit linux). Good work! This makes a huge difference! Since you play so many turns even a delay of a fraction of a second is incredibly important.
My laptop has fallen way behind. 2.5ghz dual core processor, intel hd 4000 graphics card (which has countless annoyances), 8gb of ram, and 400gb of hard drive space (that is counting the 250gb hard drive I added).
My senior year of college I am hoping to build a top of the line desktop, but I’m not going to shell out 2 grand until then. The last thing I need is to not be able to pay for college.