During a climb animation, you can move the manual aim cursor with L1 + the analog stick, as well as the strafe buttons. The camera position will be stored even after the climb as long as you're still holding L1. This lets you line up vent shots during a climb -- and even still continue to move your aim after the climb finishes -- instead of having to move the cursor from a standstill, saving a few seconds each time.

Example 1: Example 2:

If I remember right, they specifically show up on the winding path that leads into the main Pharmacy room, or at least a bunch of them did for me

It seems like what's happening is that NTSC waits for VSYNC before advancing the game, so it runs at an even 20fps and updates every 3 frames. PAL doesn't wait for VSYNC, so there's screen tearing, but it updates more often and moves faster. The difference across the whole run is several minutes.

All currently active players are running on emulators that are confirmed to be slower than PAL PS2 with fast disc speed, so if you play on console you'll have the most competitive setup.

It could be added as a new category tab to keep everything under one game page for convenience (and I'm considering trying to run SE as well). I've seen a lot of communities do that.

I want to echo what Polari's saying, insofar as I think it's important for speedrun.com to acknowledge that inaccurate leaderboards maintained on SRC are at least as big a problem for us as submission lockout is for you guys.

There's no doubt this has become the centralized hub for leaderboards within the community and, in many cases, outside of it. When people want to find out about a game, they either go to youtube, or they come here. Coming here is a huge step up from going to youtube, where people have uploaded TASes or claimed WORLD RECORD and PERFECT RUN on something they didn't even do. I come here first when I want to know about a speedgame.

SRC being higher profile than MMLB makes sense, and it's clear you're aiming to be an increasingly accurate authority on every game. But it shouldn't come at the expense of functional, content external communities because you haven't had a chance to come up with a good solution yet, who have to come here and play ball or else deal with this inaccurate information being out there. People in the broader community even start to stick their nose in and tell us what to do, in a "why do you get to be special" sort of way. It causes a lot of unnecessary tension.

To that end (and I can only speak for myself here), if the API or Joka's automatic mirroring solution is unlikely, I'd be satisfied with game pages that

a) don't display any leaderboards b) feature a prominent redirect to MMLB c) still take submissions for the purposes of a complete SRC profile, and d) require those submissions to have been verified on MMLB first.

The important thing about this is that it prevents newcomers from stumbling on a leaderboard on speedrun.com that doesn't indicate that it's inaccurate or sporadically maintained, but still lets people round out their profile if they want. This wouldn't be ideal (that'd be the API), but it would be far simpler than trying to manually maintain two concurrent boards. We could check to see if the run is already on MMLB, confirm it's the same video link, and up it goes. Only the people who are interested in tracking all their PBs on their speedrun.com profile would use this feature, and it would be advertised as an option in the game page redirect notice.

Trying to maintain two leaderboards manually would be a lot of extra work and would, in my opinion, eventually lead to more confusion and burnout among the classic Mega Man community, who are broadly happy with the board we have and aren't interested in migrating or maintaining additional game pages here. The incomplete/inaccurate boards with no prominent links to MMLB (as well as the nonexistent boards) have also caused confusion for people who come to our streams and don't know what the record is or if anyone runs the game.

I realize that's because Speedrun.com is what most people know about. They think everything is on here if they're not told otherwise when they get to the game page. But that doesn't mean the structure of this site is the best fit for a community whose leaderboard predates SRC and is tailored to what we need, with features SRC is unlikely to have in the near future, or maybe ever.

If Joka's automatic mirroring from MMLB to SRC isn't possible, my ideal solution would be a prominent (not tucked away in the Rules tab) message and redirect like DKC currently has, but also allow people who wanted their Mega Man times on their SRC profile to submit them (and they'd be approved by whoever runs the board, after checking to see if it's been verified at MMLB). Times would no longer display on an incomplete leaderboard, but they'd be visible in your profile, which would solve the PB Tracker problem. Everyone who uses the site gets what they want this way.

If that's not acceptable, and SRC is unwilling to provide a DKC-style redirect for the Mega Man community, it seems like they're essentially forcing us to manually update these pages in addition to the leaderboard we prefer to use in order to keep them accurate. It's hard to see that approach as anything but antagonistic to external leaderboards and the runners who prefer them. If this impression is wrong, I apologize, but it's the impression I've gotten over the years.

I hope we can work something out that's better than the current situation, but I personally at least don't have any interest in leaving MMLB or doing anything that leads to its deprecation.

I like the idea of All Speedrun Levels as a category, I'd definitely watch runs of that. It'd probably have to be broken up into multiple tiers of glitchless categories like the Story Mode is, though. I remember the in-game leaderboards being ridiculously fast.

So, I don't understand what's happening at the start of this level where you drop off the edge, turn around, and warp to (I assume?) a later checkpoint. I've tried it a bunch of times but it doesn't work for me.

Also, is this really not a glitch? Cause it seems pretty glitchy.

FLY'N's performance is tied to its framerate, meaning that at 60FPS, it runs at 100% speed, at 54fps it runs at 90% speed, etc. Over the course of a 80min+ game, this adds up to minutes of lost time at anything less than a perfect framerate, and the game often dips below 60 in massive levels.

I first noticed this when comparing runs recorded in FRAPS to runs I'd done on Twitch recording only with OBS. I was consistently losing time on every split, even in autoscroller stages, and had no idea why. It turns out that while recording in FRAPS, with the "Lock Framerate while recording" box checked, FLY'N maintains 100% running speed even if the framerate itself drops below 60. You can even set lower framerates, like 15FPS, and the game will lock its speed to a consistent 25%.

FRAPS unfortunately isn't free, but it's the only thing I've found that can provide a consistent running speed for RTA comparisons of this game. I'd be interested to hear what kind of results other people get with and without using it, or if another method is found to lock the framerate. But if anyone pushes this game really hard in the future and consistently falls below my split times for no apparent reason, that's probably why.

I timed from character control to the SNAPSHOT logo appearing on the screen, but I'm happy to switch to entering the final door, it seems pretty easy to split there too.

Edit: It's been so long that I forgot why I originally did that, it was to mash through the final polaroid cutscene so the credits would play. But it was impossible to tell which input skipped the last one, so I just split on the first visual cue.

You need very quick hits on the first 3 phases to make the 4th possible, as the first balloon will block your jump if you're too slow. The third balloon prevents an even faster 4th hit by being right on the edge of the platform when you'd ideally reach it, so this seems to be the limit of this fight. You have to delay your jump between platforms 2 and 3 so you'll grab the edge of platform 3 after the balloon passes, but before the next laser.

The skulls in phase 4 aren't affected by the global timer, so you can do the same thing every time. When the second "chain" of skulls begins (inside the last large gap before the gaps shrink) I do a medium-delayed double jump followed by a quick double jump to land in the smaller gaps and get onto the first balloon platform. This took a while to get used to, since you can't see the gaps until right before you land in them. You have to know they're coming and stagger your double jumps in the same way.

I'm not sure if grabbing the second key earlier in phase 5 could speed it up any. It'd be a very small time save if so.

I was a bit put off running this game due to confusion about what my category would be as a result of what I was excluding, so I appreciate that NG+ "glitchless" (no mega, no wall clip) is a thing on this board. I agree that it's tough to clearly define the difference between manipulating the height of a jump pad boost and ¤profoundly¤ manipulating the height of a jump pad boost, but in practical terms of how it affects the route, it's a major difference.

I'll probably run and try to optimize NG+ Glitchless with as many skips as I can consistently do, but I actually have some thoughts about running a secondary route that foregoes most skips and just sticks to the designed courses. This would be a truly arbitrary distinction and probably not a good candidate for a competitive category, but what initially drew me to the game was how much fun the "intended" route was and how great it felt and looked to execute it quickly. With a game like this, I'd just as soon grind to produce an optimized playthrough of the most impressive-looking route I could find than focus on getting the best time I could. As I said, probably not a category for the leaderboards, but DeadCore inspired me to rethink what a speedrun necessarily "has" to be about, and it seems like it would be a lot of fun to work on.