Looking for ideas for a before-and-after benchmark test before upgrading nw.js

HankB

Actual size
Veteran
Joined
Mar 5, 2021
Messages
115
Reaction score
64
First Language
English
Primarily Uses
RMMV
EDIT: (tl;dr: I'm stress-testing RMMV, so what are some commonly used functions/event commands that seem to use a lot of resources and slow things down?)

I'm getting ready to upgrade my MV installation with a newer version of nw.js (as per this post). But before I do, I'd really like to do a "before and after" comparison, based on FPS rates, so I can get a somewhat accurate and measurable indication of how much the upgrade actually helped. A lot of people report that the upgrade helped a lot (like a lot lot), but I'd like some actual numbers.

The general idea is that I create a map that gradual becomes more and more of a resource hog. As that is happening, I constantly check the framerate, and whenever it drops below 30, I record a log of exactly at what point that happened and then start over again. I do this 100 times and then average the results together.

Here's what I have so far: I'm using Galv's Event Spawner to gradually spawn in events into a blank map. The event that is spawned in is just a dragon that flies around randomly, but it's a parallel process that is also doing some math and checking a conditional branch, like:

Control Variable : #1 = Random 1..2000
If variable #1 > 1000
Control Variable : #1 = Random 1..2000
else
Control Variable : #1 = Random 1..2000
end

and it just does this over and over. Obviously, as each new event is spawned in, it becomes a heavier load on the game. Whenever enough dragons are spawned in that it causes the FPS to drop below 30, the player is transferred to another map, which logs the number of events that were spawned in, and then transfers the player back to the testing map. This happens 100 times, and then you are given an average of how many events were spawned in per cycle.

Currently, for me, I'm averaging about 170 events spawned in before my frame rate drops below 30. Of course, it will be different for different computers, which is kind of the point. I want to create a test that others can use before they upgrade so that they can see how much of a difference it makes on their own systems.

I feel like this method would probably give users a pretty good idea of how much of a performance increase they are getting after the upgrade, but the problem is that it's completely based on the number of visible, active events that are on the map at once. And, yes, those events are moving around, and doing some math and conditional checks, but I don't think that those things are very intensive.

What are some other, more greedy event commands I should check? Any advice would be greatly appreciated. Thanks.
 
Last edited:

Syberduh

Veteran
Veteran
Joined
Jul 26, 2021
Messages
31
Reaction score
19
First Language
English
Primarily Uses
RMMZ
Displaying/moving/scaling pictures?
 

Frostorm

[]D[][]V[][]D
Veteran
Joined
Feb 22, 2016
Messages
2,370
Reaction score
1,985
First Language
English
Primarily Uses
RMMV
Did you ever end up doing those comparison tests?
 

Latest Threads

Latest Posts

Latest Profile Posts

I have been enjoying a game on my iPhone, called Wonderbox. It is just a little game where you create adventures, share them, whilst also being able to play others. It's a bit zelda-esque. I have started some youtube for those, gonna start with RPG maker things soon as well. So if you wish to see, do check it out :)
SF_People3_3 added!
index.php

ScreenShot_1_3_2022_1_32_12.png
I feel QSprite for MV is underrated. I basically am able to use my chibi sideview battlers I drew on the maps and have all the motions available. this'll help a lot with cutscenes.
I've figured out how to reasonably tall-ify chibi sprites on my editor. I like that making tall sprites gives me control on expressing how tall characters seem on the field relative to each other as well as better differentiate adult sprites from child sprites.
doing some wacky experimental stuff in MV right now :kaoluv: I cannot wait to drop what I've been working on

Forum statistics

Threads
118,561
Messages
1,116,951
Members
155,601
Latest member
Syncope
Top