United StatesDonkey Kong Jungle Beat (gamecube one) can't be searched, for whatever reason.
[quote=Ewil]ILs should be timed manually without loads and cutscenes should be included, if the game doesn't have IGT or it's broken.[/quote]
It does have IGT, but it's only accurate to the minute, which isn't very reliable. I am currently timing ILs to include cutscenes since 1) They're all fixed durations so they don't make and difference between runs and 2) They could theoretically be skipped.
And by hidden, I mean I'd like for viewers to be able to toggle their visibility.
I agree. It's mostly for the sake of individual levels, since game time isn't accurate enough for them, and calling the primary time 'time with loads' wouldn't be very accurate, since it would differ from full-game runs. Unless the series gets really popular I'm able to manually time things myself.
Since 'real time without loads' is the primary time, recording 'time with loads' and 'in-game time' is fairly arbitrary. If there was an option to hide those columns by default, I would.
I've set 100% as getting 5 Stars on every mission as it makes the most sense. There's four in-game requirements to get them:
- Beat the mission under a certain time
- Complete all secondary objectives
- No reloading from saves
- Play on Commando difficulty
It may or may not require 5 stars on all four categories to have 5 stars overall. I'll check this when I can. If it doesn't, the definition could be changed to 5 stars on everything (since that would fit '100%' better). Depends on what the runners think.
For now, I've made 'real time without loads' the primary time. Until I or someone else makes an auto-splitter they'll have to be manually timed. 'Time with loads' and 'In-game time' are available as entries, but they're not required to be filled out.
• Real time without loads is the sum of times for each level, starting when the HUD appears and ending when the mission locks at the end or the last frame the HUD is visible. Time wasted on reloading or restarting are not subtracted, but this can be changed if the runners want to. It's required for submissions, but the runs can be timed by moderators if necessary. • Time with loads starts on clicking the campaign difficulty and ends when "Mission Complete" appears on screen after the final mission. Not required for submissions and ignored for individual levels. • In-game time is the sum of the times displayed at the end of each mission. If there were any reloads or restarts, this entry can't be filled out. This too can be changed. Not required for submissions.
Anyone looking to run the game shouldn't be too concerned about getting the right time. So long as a timer is visible I can check the times manually.
I'll probably apply this method to the other games in the series when I have the time to re-time runs.
Sure, you'd be checking when you do and do not have the achievement/item. It may take some time if it's a Boolean value.
How are Real Time/Time With Loads and Time Without Loads organized in the backend? When I added Time Without Loads as an option to Command & Conquer: Renegade I noticed the then Real Time values moved over to Time Without Loads. Wouldn't that cause a swamp of inaccuracies for runs that aren't manually updated? Why not have 3 time records (RT,GT,NL) and change the column titles based on which columns are visible?
This does explain how this run is considered first place on my end, but then how does the database know that the 0:21 is Time Without Loads and not Time With Loads, since the API reports it as realtime and not realtime_noloads?
It's not COMPLETELY different each time you make one; you'll generally go through the same process.
You can find a hook by using Cheat Engine's ability to watch memory. You continuously test to find a certain value (like, say, money value) until you nail down where the value is being stored in the memory. How exactly you then get the hook I can't recall, unfortunately, it's been a while and I'm not properly equipped to check.
Ah okay. Auto-splitters have to be made for specific games in order to work. It monitors memory values to know when to initiate a commands. If you wanted to make one yourself, I'd suggest getting a program like Cheat Engine to help find the hook for the memory values, which you then add to the autosplitter.
For specifics, Cryze and Wooferzfg can help.
What you're thinking of is cartridge flashing. I know they do it to Yoshi's Island cartridges so they can play ROM Hacks with Super FX.
There doesn't appear to be any N64 flashers that I could find. You may have better luck.
If you have some programming knowledge, here a reference: https://github.com/LiveSplit/LiveSplit/blob/master/Documentation/Auto-Splitters.md
Made a list of black hole'd runs (runs that don't exist on any leaderboard, regardless of obsoletion): https://docs.google.com/spreadsheets/d/13PkXwKL-lm_dnz6pkU1TKh8nJ4OACPWN4rKZKx6uIYU/edit?usp=sharing
Not sure if this thread was the right spot.
[quote]Delete pls[/quote]
Surely you jest.
It probably existed for consistency. A fresh file would go from the start to end. Bouncing around to different missions 'doesn't seem right'.
If I remember correctly it's mostly transitions between screens that's faster on WiiVC. Resetting takes longer but the transition speed makes up for it.
Couldn't tell you about region differences.
The site staff refrain from interfering with other games' leaderboards and rules. Try making a thread in their forum about it to discuss allowing it on their leaderboard, either together or separate from the current Any%.
Twisted Insurrection and Mental Omega are Modifications instead of ROM Hacks.
kirkq mentioned there are daily backups, so I'm pretty sure it's possible to roll something back. If they can't, they can at least list all the runs that are gone.
How does the database rank leaderboards? I'm guessing it's PrimaryTime>RealTime>GameTime>NoLoads>RunDate. I've found some runs with all of the above being the same, this being the worst offender. This one has same times but different dates, which explains the obsolete runs. But wouldn't it then rank by Dates on leaderboards too? ie for two runs with the same times, the older one is ranked higher. How do you get around that?
...also how do you query it efficiently? Takes me like 15 minutes to generate positions for all runs. I also get the floor number instead of ceiling for runs/players with identical times.
