Famidash is a homebrew NES demake of the 2013 PC/mobile game Geometry Dash. Released in 2024, version 1.1 demakes the first 13 levels of the game (as well as 7 community levels). This TAS improves the current publication by 201 frames by swapping the final level with Clutterfunk, which has a long ending that doesn't require input, allowing input to be ended early. Two of those frames were saved during menuing after Cant Let Go, although I'm not really sure how. 6 frames are lost altogether in order to do the different level route, compared to doing them all in order.
To add to this, I also spent a good amount of time trying to find various "swag routes", a term Geometry Dash players use to describe unintended jumps in levels. This sometimes leads to me humorously jumping on the edges of spike blocks, flying in straight lines, gluing the UFO to the ceiling, and just generally going places you’re not supposed to go. One really interesting moment in this run is near the middle of Electroman Adventures (62350f), where I use the mini UFO to skip a ship portal. It doesn't save any time, but it's a cool find.
Important note: The final rerecord count is 3107. I had to figure out why the SHA1 hash wasn't correct, so I copy+pasted the inputs on an empty file.
Last thing worth mentioning. The game is going to have an update soon that adds more levels. Not sure on the exact amount but it should be around 46 levels. I intend to make a TAS on the new version. I bring this up, because the TAS would be longer as a result of the newer content.
DrD2k9: This run is an acceptble improvement over the current publication of the same branch.
Assuming the v1.2 run is as optimized as this, it will likely be acceptable; but I'm not sure how the site will deal with the significant addition of content between the releases. It's possible that v1.2 may obsolete this run for being the version with more available content (assumign all the content of this run is also in v1.2). It's also possible we may keep both versions. I'll have to discuss this more with other staff.
For now though, this run can and should be published (if for no other reason than historical record keeping/obsoletion chain purposes). Accepting.
Procedural notes:
The movie syncs on BizHawk 2.10.
The submitted movie file had an extra frame of blank input; replacing with a properly truncated version.
Did bizhawk throw up an error window when the run finished? Because the embed video has a really alarming stop/stutter right after the 19:20 mark
It did. Sorry about that. I fixed that issue since then. Working on a new encode now.
Edit: I'm still getting the error window. Don't think I'll be able to figure it out right now.
since I think the original publication is the worst TAS to ever be published.
That is a rather strong opinion with no backing evidence or reasoning. Since you didn't comment (or vote or review) the previous TAS, could you elaborate why you feel this way? How does this improvement change that opinion?
Joined: 1/7/2023
Posts: 89
Location: Somewhere on planet earth probably
Bigbass wrote:
Mikewillplays wrote:
since I think the original publication is the worst TAS to ever be published.
That is a rather strong opinion with no backing evidence or reasoning. Since you didn't comment (or vote or review) the previous TAS, could you elaborate why you feel this way? How does this improvement change that opinion?
Ok, it probably isn't the worst on the entire site, but it's definitely the worst I've seen, and I have my reasons for that:
* The game is an autoscroller, where the only optimization comes from menuing and ceasing inputs as fast as possible.
* The game itself is just a clone of an already existing game, meaning there's no reason to not TAS the original, unless this version had substantial changes, which it doesn't (this is the biggest problem in my opinion).
* The TAS itself has no "playaround", you can see from the creator's encode that they just made the movie almost entirely on recording mode. You don't gain anything from watching the TAS that you wouldn't get by just seeing a walkthrough or just playing the game yourself.
* Remember how I said there were only two elements with optimization. Well, one of those elements isn't even optimized! There was no consideration for what stage should be played last in order to end inputs as fast as possible, which I wouldn't care about if it wasn't one of the only two things you could optimize.
I think this TAS is better, but not by much, it's still "meh" to me (whereas the original was a big no), but it improves on the shortcomings of the previous TAS, such as adding actual playaround and taking the last stage into consideration. Also, Walgrey says that there will be 46 levels, which is more than the original game, meaning there will be new levels, which would negate the second issue I mentioned earlier.
"He who is tired of TASing is tired of life" Something Homer Simpson definitely said
* The game itself is just a clone of an already existing game, meaning there's no reason to not TAS the original, unless this version had substantial changes, which it doesn't (this is the biggest problem in my opinion).
Geometry Dash does not have a native linux release. Unless there's some other way to TAS it that's acceptable by this site, going with a TASable demake seems reasonable to me.
Mikewillplays wrote:
* Remember how I said there were only two elements with optimization. Well, one of those elements isn't even optimized! There was no consideration for what stage should be played last in order to end inputs as fast as possible, which I wouldn't care about if it wasn't one of the only two things you could optimize.
I don't think it's fair to so harshly judge past submissions for improvements that weren't discovered until later. You didn't provide this feedback on the previous submission, and the judge didn't think of it either.
Thanks for elaborating. In the future, I'd strongly encourage you to provide more detailed feedback upfront, instead of bluntly saying something is "the worst TAS" ever. If you believe there are improvements that can be made to a submission, please offer them constructively.
I've been asked to list the differences between versions 1.1 and 1.2 of this game, but I couldn't find a nice way to integrate it into my submission notes. Hopefully this suffices.
From what I can tell and according to the game's changelog, the only noteworthy differences are changes in the game's physics and more/different levels.
Note: The game's changelog mentions 63 levels, but only 43 are actually in version 1.2. The rest are in a separate ROM, called the "B-Sides", and this is currently being TASed separately.
(It's really weird because one of the levels in the B-Sides is just a test level of seemingly infinite length, so it's more like 62 levels, actually.)
The following is a table that has all the levels in versions 1.1 and 1.2.
Main Levels
