View Issue Details
| ID | Project | Category | Date Submitted | Last Update | |
|---|---|---|---|---|---|
| 0014548 | The Last Federation | Bug - Gameplay | Apr 23, 2014 11:30 pm | Apr 24, 2014 5:29 pm | |
| Reporter | tbrass | Assigned To | keith.lamothe | ||
| Status | resolved | Resolution | fixed | ||
| Product Version | 1.0! | ||||
| Fixed in Version | 1.005 Beta (Holy Smokes It's Big) | ||||
| Summary | 0014548: Graphical lag in late game makes TLF unplayable | ||||
| Description | A couple of us on Steam have started discussing how late-game graphical lag in the strategic view makes the game nearly unplayable. We have no problem in the tactical view, and early to mid-game is just fine. But with thousands of moving ships, the whole thing slows to a stuttering crawl, especially when trying to fast forward. Potential fix: settings option to reduce displayed force levels when forces exceed X. So instead of displaying a thousand individual ships in a fleet, use 100 10-ship icons instead. | ||||
| Tags | No tags attached. | ||||
| Internal Weight | New | ||||
|
|
I suspect its not graphical lag - its cpu lag. Think of something like DF's collection of thousands of cats bringing the entire fort down to <10 frames/s. DF in general isnt exactly gpu intensive. |
|
|
As a test, if you zoom way in on the solar map looking at basically nothing, does it still lag for you? If so, then it's a GPU thing. If not, it's a CPU thing. Either way, if you can provide your specs and a savegame that exhibits this, then we can look at a solution. Thanks! |
|
|
I also had some major lag in lategame.. but only in the diplomacy screens as for my system.. I have a 6 core AMD 1100T CPU, 16Gigs RAM, and an ATI R9 290 4Gig video card and a Vertex4 SSD |
|
|
|
|
|
Hi Chris, Save is uploaded, I've zoomed all the way in and it still lags when sped up (by lag, I mean, jittery, jerky performance where nothing will happen followed by 2 months disappearing and a dozen notifications on the screen). I'm running a laptop w/ 8gb ram & an AMD A10-6400M APU (2300 Mhz, 4-core). Ajnin makes a good point in that there is also lag in the diplomacy screens (I will select an option, click execute, and there is noticeable time before the next popup comes into existence). Thanks for looking into this! |
|
|
Got it, thanks! Sounds like a CPU issue for sure, then. |
|
|
All my games that manage to go up to 3040+ or so with billions of inhabitants on planets and hundreds, if not thousands of armadas, do indeed get laggy *when sped up* However, on normal speed the game has (so far) managed to stay playable (not sure if this would be the case if I'd ever reach 3080 or so, also not sure if getting that long in the game is even easy anymore with all these ground combat and infrastructure affecting patches!) - most of my experiences are from 1.001 actually. Mah system specs be about 5 year old core i7 920 @ 3.6GHz, 12GB RAM, Geforce 750Ti. Despite my positive outlook on things, I do suspect that the simulation engine could use some optimizations too. ;) In tactical battles I've had no problem whatsoever, except on those turns when a ton of small ships spawn at once (such as when breaking open a Hydral space hangar, or what are they called) - but even then there's usually a hiccup for one turn (while the entities are loading, I'd presume) and after that things get normal. EDIT: I see, the tilde character does something wierd in comments. Edited it out, it caused some funny stuff to happen. |
|
|
For 1.005: === Solar Map Performance Improvements === * While on fast-forward the little "non-sim decorations" for armadas in orbit, satellites, etc are only updated every 250ms, rather than every single sim step (particularly since not all of those sim steps are even visualized on fast forward). * Now the completion of a new armada no longer immediately causes a race to do the full "reevaluate fleet priorities" logic, but causes it to do so the next time the solar month changes. * Some building-construction related checks have been moved back from per-frame to per-solar-month (and only then when needed) since they didn't really need to be done that often and were causing signficant CPU load at that frequency. * The auto-resolve combat now uses coarser calculations for individual armada-vs-armada shots when dealing with larger power differences between armadas (the computation involved something that was done X times, where X was basically the numeric difference in strength between them), saving a great deal of CPU cost that was happening every game-second and thus causing spikes in lag (from 20ms frames to 80ms frames, for example) for just the first frame of each second. * Overall, in the test case we used for this: ** Pre-changes: *** Normal speed solar map - each frame took about 25 milliseconds. *** Fast Forward solar map - each frame took about 55 milliseconds. *** Ultra Fast Forward solar map - frame length isn't the main thing here, but it took about 35 seconds to simulate 50 months. ** Post-changes: *** Normal speed solar map - each frame took about 12 milliseconds. *** Fast Forward solar map - each frame took about 21 milliseconds. *** Ultra Fast Forward solar map - about 30 seconds to simulate 50 months (less of an improvement here, there's just so much math going on, etc). Thanks :) |
|
|
Thank you, Keith! |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| Apr 23, 2014 11:30 pm | tbrass | New Issue | |
| Apr 24, 2014 12:14 am | Lancefighter | Note Added: 0036611 | |
| Apr 24, 2014 11:09 am | Chris_McElligottPark | Note Added: 0036653 | |
| Apr 24, 2014 11:09 am | Chris_McElligottPark | Assigned To | => Chris_McElligottPark |
| Apr 24, 2014 11:09 am | Chris_McElligottPark | Status | new => feedback |
| Apr 24, 2014 11:17 am | Ajnin | Note Added: 0036654 | |
| Apr 24, 2014 11:17 am | tbrass | File Added: for_review.save | |
| Apr 24, 2014 11:23 am | tbrass | Note Added: 0036655 | |
| Apr 24, 2014 11:23 am | tbrass | Status | feedback => assigned |
| Apr 24, 2014 11:26 am | Ajnin | Note Edited: 0036654 | |
| Apr 24, 2014 11:48 am | Chris_McElligottPark | Note Added: 0036659 | |
| Apr 24, 2014 11:48 am | Chris_McElligottPark | Assigned To | Chris_McElligottPark => keith.lamothe |
| Apr 24, 2014 2:04 pm | Kalpa | Note Added: 0036682 | |
| Apr 24, 2014 2:05 pm | Kalpa | Note Edited: 0036682 | |
| Apr 24, 2014 3:55 pm | keith.lamothe | Internal Weight | => New |
| Apr 24, 2014 3:55 pm | keith.lamothe | Note Added: 0036713 | |
| Apr 24, 2014 3:55 pm | keith.lamothe | Status | assigned => resolved |
| Apr 24, 2014 3:55 pm | keith.lamothe | Fixed in Version | => 1.005 Beta (Holy Smokes It's Big) |
| Apr 24, 2014 3:55 pm | keith.lamothe | Resolution | open => fixed |
| Apr 24, 2014 3:56 pm | keith.lamothe | Note Edited: 0036713 | |
| Apr 24, 2014 5:29 pm | tbrass | Note Added: 0036731 |
