The main problem with flat allowing any name code is that it would eliminate any way to see how the run was done apart from comments. Normally there is at least something to filter by.

I think the good options are:

• Variable like with the Maria categories, segregates
• Field that is displayed on a run indicating the name code that was used, allows filtering

Another option is to preserve current Any% NSC and change to Any% NSC No code, but that's ugly.

Unfortunately it seems like the time youtube provides with their api is not accurate, so I've had to add a framerate field in order to properly adjust milliseconds. This reduces ease of use but I'm hoping maintains the usability of the tool while being accurate. I didn't expect the api to spit out a fraction that is straight up incorrect.

The current time is taken directly from youtube's information for that frame, so it all depends on how the video is transcoded and timed by them specifically.

EDIT: It seems like the way they track time internally and provide it via API might have some inaccuracy. Will see if I can account for that.

Oh that is done on purpose, assuming that the segments would be spaced out by at least a frame.

Hope it is helpful, feel free to message me somewhere or submit an issue on github for problems, bugs or feature requests.

For a while we've used frame-accurate timing for top times in the SotN community. A while ago we also started removing loads for certain categories to combat performance differences.

I've made this tool to make the process of doing that faster and easier without the need to download anything. But unfortunately the process is not ideal and not precisely accurate.

You can find the tool here: Segmented Timer

And a short video instruction here: How to use the Segmented Timer tool

EDIT: I made an issue request to the youtube team and they responded that they will not adjust their API to increase accuracy of the reported current time. This means that the only way for the timing to be accurate is to request the runner to provide the exact framerate, since they both don't have a visible framerate in their API and even if they did it would be rounded. My desire was for this to replace the need for yt-dl and avidemux, but it seems like it won't happen.

I imagine either nearest or just removing the last digit.

Well no everything that has milliseconds should be rounded to hundredths is what I'm saying. No need to retime.

I suggest some slight changes to run timing conventions and definitions. I've edited this guide to my proposal: https://www.speedrun.com/sotn/guide/eho60

This includes a change in how we define loadless timing to better suit how we do it. Instead of defining the loading period, I think it is better to define the gameplay segments.

I've made a tool that can hopefully help make the process of timing runs with or without loads less cumbersome: Segmented Timer

Additionally I think it makes sense to round milliseconds to the second decimal point since encoding and capture methods make the thousands effectively meaningless.

Since it turns out that there is significant performance differences between different versions of the Mednafen Saturn core it seems reasonable to adopt this methodology for emulator runs, while encouraging runners to use the latest version available.

I made an example of timing a run:

The first frame is wrong, you can double check by looking at the times on the spreadsheet with avidemux.

Routing refers to how URLs are mapped on the website. Currently you can only link to a board's main categories, but not to subcategories.

It has been years since the feature was added and a majority of games are using it to organize leaderboards. It is downright inexcusable that there is still no routing set up for them, considering how much this impacts navigation on the site.

I think he meant mobile, not just Android.

Bruh, it's called a jump frame, because you have to jump on that frame :D The frames are fine you just weren't doing the right things. Also you really shouldn't be doing this trick at all at the start, it doesn't save that much time and only does so if you do it really quickly. There are many much more important things to focus on at the start.

That is the good jump frame, your second picture shows you already landed on the stairs though, so that is way too late for untransform. Just those 2 frames don't really tell the whole story, if you post a video with input display I can give you a more specific pointer and tell you exactly what is going wrong.

Probably a good idea to put in the rules that the game has to be ran at 60hz/fps since photn can run at whatever you can set your display properties to.

Oops I forgot to publicly share this when I left. Here are the old verification guidelines I made a while back for reference. https://docs.google.com/document/d/1d71jr1UkLyGC7j63syY3kmzYzBTzxWzSkeNlRXwSbp0/edit?usp=sharing

Os emuladores permitidos são Mednafen e Bizhawk. Existe uma versão do jogo para celular que você pode rodar, embora por enquanto não seja rastreado nos fóruns, mas você ainda pode postar sua corrida no fórum.