Commentaires
ScotlandReally_Tall4 years ago

This is only the demo for Super Mario 256. Once the full hack is out I'm sure it will be added.

VTMagno et J_duude aime ceci
ScotlandReally_Tall5 years ago

ROM hacks would have separate leaderboards: https://www.speedrun.com/asm3d

ScotlandReally_Tall5 years ago

There have never been any good category suggestions in the SM64DS forum, ever. Even if there were, there's absolutely no point adding a category if nobody has done a run. We already have many categories that explore the game in different ways and offer more than enough variety. And if there was, one day, a really great category suggested that had some runs and wasn't trivial, arbitrary, boring, or a healthy combination of the three... it could just go on the main leaderboard.

ScotlandReally_Tall5 years ago

If you want to see records beaten by small amounts there are already around 600 perfectly good ILs. There's nothing stopping you running this, but it really does not need a leaderboard. And you can communicate with the community pretty easily through the SM64DS Discord server: https://discord.gg/Dpg2cYH

afnannen136, WildGeffe64 et 2 autres aime ceci
ScotlandReally_Tall5 years ago

No. That's just failing moat skip once...

ScotlandReally_Tall5 years ago

You can view solely the N64 runs using filtering: https://i.imgur.com/PMx9dRn.png

Caseywho, HorhaytheDragon et 2 autres aime ceci
ScotlandReally_Tall5 years ago

You can! Emulator too if you really want to, but emu will be hidden by default. Same applies for full game runs :)

Imaproshaman et IanTEB764 aime ceci
ScotlandReally_Tall5 years ago

80 Star and 50 Star are the easiest main categories to start with. You can get help or other info in the SM64DS Discord server: https://discord.gg/Dpg2cYH

Pear et Imaproshaman aime ceci
ScotlandReally_Tall5 years ago

This glitch is extremely well-known, and used a fair amount in the B Button Challenge:

ScotlandReally_Tall5 years ago

Has anyone experimented with automatically removing loads ( https://www.speedrun.com/nsmbds/thread/soxwb ) yet? This would ensure complete accuracy if there are deaths or an alternate route is used, compared to adding a set time. It shouldn't be any trouble once it's set up initially, and means you don't need to do lots of work calculating loads for different categories, and could even apply it to misc. categories and category extensions. You could put all current and obsolete runs through the program as well so everything is consistent and accurate.

I didn't vote since I don't run this game, but if you can get that load remover to work then using it for absolutely every category makes sense to me. Hope it works!

ScotlandReally_Tall6 years ago

Option 1, adding high scores to speedrun.com, is the best option. Every other option boils down to creating a new site in some form, which is unnecessary. Firstly, there isn't anything that a high score site needs to do differently from what speedrun.com already has, aside from sorting leaderboards by score rather than time, which is an easy task. For the majority of games there's significant overlap between the playerbase and resources. It's clear there's already high demand for scores here, some games close to rivalling Cyberscore despite no official implementation. And using one account to submit everything in the same place is more desirable than constant switching. This also applies to moderation. Scores could be implemented on speedrun.com in a few different ways. Some users will want to see only scores/speedruns, others wanting a mix. A third tab after "Full-game Leaderboard" and "Level Leaderboard" called something like "High Score Leaderboard" as many have suggested should work, with settings available similar to those for IL runs so it can be ignored easily if desired. There could be a more distinct separation ("modes" for the site like Forstride suggested) or an extremely similar sister site but I think that's overcomplicating the whole thing and limiting the convenience greatly. A more representative domain name isn't worth the hassle for users, let alone the price.

Ideally, scores here would have some crossover with the other high score sites. Giving mods the option to import all scores from another site, only those that have some type of proof, only those with a video, etc, would work well and could be updated automatically when new scores appear on other sites. Imported scores could be marked as such with a variable stating the original site and a link provided. There could also be an option to auto submit scores from other sites, yet still have moderators here approve them. When it comes to moderation, for classic high score-focused games there would be conflict if the current speedrun.com mods for that game handled score moderation, so the distinction of speedrun mods and newly chosen high score mods would be good. For small games these would be the same person/people though, so the feature should be optional.

P.S. I made a thread about high score runs a while ago which has some interesting points: https://www.speedrun.com/The_Site/thread/5cadj

DarkEonMaster, SadEgg et 3 autres aime ceci
ScotlandReally_Tall6 years ago

I wasn't aware that InputRedirection worked while playing DS games, considering that the Rosalina menu is disabled. Sure you can use it!

ScotlandReally_Tall6 years ago

Paper Mario series server (with a channel for SPM): https://discord.gg/aBvMuxY

Tortree et NmFlash8 aime ceci
ScotlandReally_Tall6 years ago

I think the prevalence of cheating is being exaggerated in this thread. Sure, it's happened many times in the past, but that's across all games ever in the span of a decade. The vast majority of bounty attempts will be legitimate, or the kind of obvious trolls that are already commonplace on the site like 1 second runs: I can't recall any illegitimate claims to SDA or TASVideos bounties, even the massive ones like the old Barrier Skip bounty for The Wind Waker which was >$1000.

Despite this, there will always be people worried about the chance of cheating. I think as long as you give the person creating the bounty unlimited freedom for setting the proof required then everyone will be happy. Minimum levels of verification would be the run getting accepted to the site, with additional options such as having splits, a video, a mic, the run being livestreamed, an input viewer, a controller cam, the run happening at a live event, etc. The only problem with user-defined requirements is ensuring payout, which I'll touch on later. And those proof methods are for bounties for records being broken: with a glitch bounty the only proof necessary is another person replicating it. I'm not sure if you've considered TAS bounties too, but similarly all that's needed for those proof-wise is the input file, or acceptance to TASVideos.

Anyway, to me the biggest issue is not cheating but how you'll make sure the money is paid. The main problems with older bounty lists are that expiry dates were not set, so today nobody knows how many of them, if any, they'd get the reward from. Yet even when the bounties were new you had to rely on the word of another person with no other guarantees. If this site held the bounty money centrally and required expiry dates for all bounties then I think these problems would mostly be alleviated. The remaining issue is how to confirm the bounty has been met. With a simple bounty that requires a run under x time and acceptance to the site, the system could be automated. But when more proof than splits and/or video have been requested, or the user has set subjective requirements (run can't use any "big" new skips to achieve x time, the glitch must make the run more entertaining, I must give the TAS a technical rating of at least 7/10, etc) then it gets more complicated. If the bounty setter can deny a valid claim due to ambiguity then this would hurt the system. I think you'd need a small impartial group to moderate bounty submissions and confirm all requirements are objective with as little ambiguity as possible, as well as determine when the bounty has been met. You also need to consider what happens if multiple people help to find a skip, how bounties could be taken down, and so on... it's something I'd love to see on the site but there are a lot of logistical issues that need thought, and without speedrun.com holding the money while the bounty is active I don't think the feature would be worthwhile.

Tl;dr: Cheating isn't a problem as long as users can set the level of proof they want, there are bigger issues with creating a bounty system. The site needs to moderate bounty submissions and hold the money while the bounty is active or there isn't much point.

werster, Timmiluvs, et MASH aime ceci
ScotlandReally_Tall6 years ago

There's also a commentated English tutorial: Been meaning to try this out myself, but haven't yet.

J_duude aiment ceci
ScotlandReally_Tall6 years ago

Thanks for everything! ^_^

Otterstone_Gamer aiment ceci
ScotlandReally_Tall6 years ago

@trueace Drastic is just an emulator, right? DeSmuME is preferred but I doubt anyone will really care if you use it since emulators are hidden by default. Just don't use emulator-specific functions like speedup or savestates obviously. You could also run on DS/3DS/VC without recording if you'd prefer to use something official, video isn't required unless you get really good. But the ideal thing would be DS/3DS/VC with a webcam/phone camera, or even better, with a capture card. Hope that helps!

ByteWise et Imaproshaman aime ceci
ScotlandReally_Tall6 years ago

The old requirement was that a sub 33 minute Main Levels - Any% run needed a video. I overhauled the rules a week or two ago, explaining a lot of things that weren't specified previously (for example, what happens with timing when you retry a level, and good ways to retime runs). I also removed this limit, given that it's deeply unfair due to the variance in loads between different devices. I'm not sure if a new roughly equivalent limit based on the loadless time (sub 21, sub 20) would be best. Alternatively, we could accept all runs unless the times are blatantly fake, but require a video to calculate the loadless time (as this is necessary to place high on the leaderboard). MrCafecito also suggested a video cutoff based on placing, say top 5, but that doesn't translate well to the other categories which have little to no runners. Personally, I'm leaning in favour of requiring video for calculating all loadless times. With the current timing method it's near impossible to get an accurate loadless time without careful retiming, and while doing this I've seen massive differences between the submitted time and the actual time. It would also work nicely for every category, even including Full World ILs.

I think the obvious takeaway here is that you should find a way to record :P I'm sure there's an app that you can use. I'll deal with the run once a rule is decided, but I'm not sure when that will be. Panther asked to be unmodded months ago and hasn't done anything since (he didn't know how to unmod himself, I'm a regular mod or I'd have done it for him already), MrCafecito is barely around and grizzly is never around. So I was meaning to ask in the server to get opinions from non-mods, but have been very busy with coursework/exams. Sorry about the delay :/

I'd be happy to hear any runner's thoughts on video requirements and cutoffs, would be very helpful in deciding a rule.

jarod aiment ceci
ScotlandReally_Tall6 years ago

There are five runs currently pending of which two are yours. If we start falling further behind I'd personally be happy to add more mods, but I think we're good for now. That burst of activity was probably when MrCafecito finished verifying many runs at once, there's not really a flood of submissions at the moment.

P.S. I'm pretty sure I verified your W6 IL a while back, did you edit it? I'll verify that one and the other ILs today regardless, as long as I remember.

ScotlandReally_Tall6 years ago

Oh sorry, it wasn't in the Guides section for some reason. Added it now:

À propos de Really_Tall
Inscrit(e)
9 years ago
En ligne
7 days ago
Parties
778
Jeux joués
BOXBOXBOY!
BOXBOXBOY!
Dernière run 4 years ago
240
Parties
Super Mario 64 DS
Super Mario 64 DS
Dernière run 4 years ago
200
Parties
BOXBOY!
BOXBOY!
Dernière run 4 years ago
166
Parties
A Hat in Time
A Hat in Time
Dernière run 5 years ago
65
Parties
Not the Robots
Not the Robots
Dernière run 8 years ago
21
Parties
Super Mario Run
Super Mario Run
Dernière run 6 years ago
13
Parties
Mario & Luigi: Paper Jam
Mario & Luigi: Paper Jam
Dernière run 8 years ago
12
Parties
Shrubnaut
Shrubnaut
Dernière run 7 years ago
10
Parties
Jeux modérés
Super Mario 64 DS
Super Mario 64 DS
Dernière action 3 years ago
527
actions
BOXBOY!
BOXBOY!
Dernière action 4 years ago
166
actions
Toki Tori (Remake)
Toki Tori (Remake)
Dernière action 4 years ago
6
actions
Ooh, Aah & You: Fetch-A-Fruit
Ooh, Aah & You: Fetch-A-Fruit
Dernière action 6 years ago
1
action