First of all, we would like to apologise for the long delays people have been experiencing in having their runs verified. It's a busy time of year, and many of the moderators of the leaderboards have been unable to spend any time verifying runs. This calls for more moderators to be added to help lower the waiting times, a proposition we have discussed for a long time. While it is important that people have their runs verified faster, it is also important that the integrity of the leaderboards are upheld and cheated runs do not get accepted. Now that we have set up some mechanisms for bringing mods onto the team, this should be much more possible.

The mod team has decided to make some more changes to the way the leaderboards are run. These changes relate to the rules for submitting runs and the people who will be verifying them. Below are the new rules which will be used for verifying all runs after the time of this announcement. Here, the any% rules are shown, but similar rules will be applied to other categories as well.

Timing is done using the SDA standard: 
Start - The frame "400" appears in the top right corner after selecting 1 player. 
End - The frame all sprites are invisible after touching the axe in 8-4. 

Approved hardware/software (please specify in the run comment):
- Original NES, Famicom or FDS
- Original NES or Famicom with Everdrive or Powerpak
- Wii, Wii U and 3DS Virtual Console
- FCEUX, Nestopia or Bizhawk
- NES Classic Edition
- Analogue NT Mini

Important rules (please read before submitting):
- All times require a video.
- Turbo/Autofire and left+right/up+down are banned.
- Times below 5:00.00 are treated with suspicion. It is highly recommended that you stream all of your attempts and provide split files to have your run accepted. 
- Times below 5:00.00 require an input file if you are playing on emulator.
- Using savestates to the title screen is prohibited. You must use the reset function on your emulator or console and the reset must be shown in the video.

First of all, the timing is exactly the same. As a reminder, times of 5:03 and above for any% or 19:15 and above for warpless are not given millisecond precision on the leaderboards due to the large amount of time each mod would have to spend to accurately retime these runs, so it is recommended to leave '000' in the millisecond field.

As you can see, there is now a list of approved hardware and software for runs to be accepted to the leaderboards. It includes almost all hardware except for any clone consoles such as the retron, which are banned. The list of emulators provides accurate emulators which have the movie recording functionality for the submission of input files. Until there is an official resource for it, feel free to ask the mods or people from the community how to use this functionality. Across the three emulators, there is support for most popular operating systems, including Linux, FreeBSD, OpenBSD, NetBSD, OS X and Windows.

The biggest change to the rules is that video is now required for all times submitted to the leaderboards. The 5:21 cutoff for video was very arbitrary, and led to many people submitting runs which had no proof whatsoever. There are no official requirements for the equipment used to record the video, so as long as it is clear enough and captures the gameplay feed entirely, we are happy to accept a camera pointed at a screen, a cheap capture card, or any other method as proof. For lower times, such as sub 5 in any%, we expect the feed to look good enough to easily determine the bowser pattern.

One reason that there is a list of approved emulators is the ability to get input files for high level emulator runs. An input file is a recording of all inputs given to the game, frame by frame. By making input files mandatory for runs under 5 minutes on emulator, the moderators have another form of proof to determine if a run is legitimate, speeding up run verification times.

The final announcement is that we have recruited valued community members to help with the verification of runs. The moderators we have added so far are Bismuth, goofychocobo and SuperSonic71087. Please don't overwhelm the new moderators with requests to verify your run. We will get to them all eventually.

Happy speedrunning :)

• Super Mario Bros. mod team

If you would like to request a new category for these leaderboards, please reply to this forum post with a message of the following format:

Category Name: Runner(s): Category Rules:

Extra Information:

Where the name and rules are as they would appear on the leaderboards, runners indicates who has run the category so far (the higher this number, the more likely the category is to be accepted) and extra information is a short paragraph explaining what the category is about and why it should be added to the leaderboards.

People can show their interest in a category by liking the forum post for that submitted category. Please keep this thread concise and formal so that as many submitted categories can be reviewed as possible. Any arguments about whether a category should be added should be taken elsewhere. If your category hasn't been added within a week, the request has most likely been rejected.

Would people prefer to have one Any% category for all difficulties or a separate category for each difficulty?

There are two new versions of supertux, which are 0.5.0 and 0.5.1 and haven't had categories for a while. We now have to work out whether categories should be made because a 0.5.0 run has been submitted. Categories weren't previously made because of the "Development Version" message that shows on the title screen of these versions.

Things We Know:

• 0.5.0 is slower to run than 0.4.0 (from testing so far) but does have some significant changes
• From my knowledge, no one has done testing of 0.5.1 for the speedrun
• 0.5.0 runs have historically been disallowed from the 0.4.0 category
• The 0.5.X line seems to still be in development, as evidenced by the 0.5.1 release, so making new categories for each version could make many categories which are unnecessary due to this development.

So, what should we do about this?

Hi supertux runners!

I'm interested in running this game, but the official version of supertux on Arch encounters a segmentation fault at the level 'A fork in the road', and similarly for some other levels later on. The git version, however, doesn't have this problem. Would it be okay to speedrun the game using the development version? If not, that's okay. I'll attempt to contact the team and see if I can get the bug fixed eventually.

Thanks.