Oops! I thought this went directly under the Amiga game forums but should have clarified in the title that this concerns the Amiga board only.

Hi!

Hopefully this doesn't come off as hostile, just found it interesting. Noticed some inconsistencies while practicing the speedrun and did a frame by frame comparison between the podium runs. You can already see in the first screen that the robot upstairs is running faster in the WR run compared to the 2nd and 3rd place. When the helicopter reaches the apex, the robot is touching the black pixels on the wall in all runs but at this point the WR is already like .5 seconds ahead. I would even argue that the WR falls behind the 2nd and 3rd place paces because the robot travels further by the time the player is exiting the screen.

While the helicopter jump is done more efficiently I don't see how this run is faster than the 2nd and 3rd place runs when it comes to the execution, especially when it comes to the last shot.

So as Voxandra stated on their run, the exact Amiga model/configuration, CPU and whatnot, should be standardized. From what I can tell, the 2nd and 3rd place runs run at the same speed and that's what you get when you use your run of the mill amiga_model=A1200 and fast_memory=2048 config on FS-UAE. I don't know what's going on with the WR run.

I agree that there should be a standard game speed/machine type for the Amiga boards. We had discussed it before, but since none of the mods were really all that familiar with Amiga were letting it slide until someone more experienced in the matter came along. Thank you for the input and we are taking your suggested standard under consideration now.

Thanks for the reply! The manual says that the game will run on any Amiga so it shouldn't be a chipset issue. I'm pretty sure it was designed for A500 but seems to run at the same speed on A1200. They do mention the A1000 in the manual as well so that supports my theory that it's universal.

It looks like Sarvets was using the Mister FPGA PAL Amiga settings... so that might explain the timing differences. Since the game basically just needs a 512KB Amiga to run, would simply specifying NTSC Amiga work as a standard?

if there was no standard to begin with, then split the category. literraly just ran what what on my mister and didnt think twice about it. Coldiron was viewing my runs live on a fairly regular basis and never mentioned anything about it so at worst we both didnt know. Splitting the category to pal and ntsc would prevent others from confusion when submitting any of the runs. They would just end up where they should. And to demi nah i dont take it as hostile at all lol. I just ran it as is

I don't think the leaderboards should be split any further than it already is. It's not exactly a competitive game, and I think that complicating submissions by splitting into categories like game version, game speed, game type, etc. will have more detriments as far as deterring newcomers than any benefits it would have. I don't want potential new players to have to care about what version of the game they're running to make sure they're submitting to the correct category, and I don't want to create a situation where we mods have to sit here and count frames or create a way to prove that you're running a certain version or reject runs because they're submitted into the wrong category. I also don't want to become a game where either you have to find a specific version of the game to get a wr or one where anyone can farm high scores by submitting multiple runs done on multiple versions of the game.

@demiurgi Don't take this as hostile, but I don't see any submitted runs from you, so either I'm blind or you're a potential runner. You're the first person to actually care about it, so if you want to do research and figure out which combination of Amiga version and game version gives you the best results, you should be rewarded for that effort. However, at the same time, I don't think we should punish people who just want to play a game by making them do research/homework that would just kill the fun.

EDIT: Oh look a run to verify. OOOHHHHH it all makes sense now

I'm doing some due diligence as a mod real quick. I'll be right back.

Personally I don't see a problem with limiting the runs to a specific configuration/version of the game because at least it would make it clear and arguably easier for new people to come in since you'd know what to use. When it comes to Amiga, you often have to make a call anyways because of loading times and whatnot. With this game at least the loading times and running speed seem to be consistent when it comes to the NTSC version.

EDIT: limiting the version and configuration is mostly an issue for OG hardware runners but that shouldn't be an issue with this game. Most Amiga hardware people use downloaded WHDload installations anyway nowadays it seems.

Oh and @Sarvets I didn't think you or the mods had noticed it or that you would have done it deliberately or anything, hopefully it didn't come off like that.

Nah demi it didnt come off that way at all. All good

Very good!

yeah this is a good solution. i dont think its ever gonna hit the point where its gonna need to be heavily picked apart. this should work fine unless something incredibly serious happens. im all for it

So when it comes to submitting, do we have to check the time from the frames? I noticed the other times got adjusted as well.

If you want to adjust them yourself, I left pauses to show the times I used for comparison. It can be a little tedious because there are some frames where the robot is rotating toward the camera, but I used the last and first frames where the robot is fully facing either right or left. Technically, the standard is arbitrary, but they can still be compared to each other as long as they are all compared to the same standard. If I'm the one verifying it, I'll probably double check it either way.

I concur with using the robot timing to normalize runs. We should have it written in the rules for how that works(ie make sure the first room is in the video long enough for the timing(in the event someone does something weird like on the DOS runs to go faster) and what the normalized times are).

I submitted an Amiga run of my own with an Amiga variant of the type of glitching that I introduced on the DOS runs. I included a view of Midnight after the run for the timing purposes, since I leave Midnight in the run before the timing can be done.