EllipticGeometry
Maze generation with a hybrid of Prims and Kruskals algorithms
updated
The nervousness got to me after the second rail crossing. Not good with the sensitive thumbstick. I managed to contain the shakes enough to finish the run.
The extra digits in the title were extracted from the save file.
I’m surprised by the performance I’m now able to get out of this dodgy controller. I was wrong when I thought a wheel was required to be competitive in the Ferndale, although it would still be a huge help.
I adjusted my recording settings somewhat after someone mentioned the odd resolution I was using (960x600, further reduced to 480p by YouTube). This is 8:5 720p now, using sheer bitrate to afford a faster encoder preset. Should look a little better to you. Being that person to comment on it can work!
I would say the van is the easiest vehicle. Decent cornering without the Satsuma’s suspension launching you on any of 3000 little projections. Above all, it’s too slow to be on the limit most of the time. Just gotta watch out for that speed wobble when already otherwise unsettled on a tight dirt road. It didn’t take too many attempts to get this result.
Blueprint available here: forums.factorio.com/viewtopic.php?f=193&t=100189
The only emulator tool I used is a savestate at the top of the ramp.
Normal: Your car has normal speed. Computer drivers are still significantly slower, but can use items now. Rubber banding is engaged: the drivers that start in 1st and 4th place speed up once you overtake them. Not that they can keep up by any means.
Hard: They’re a tad faster. They may be a tad more aggressive with items, if still mostly ineffective. Note that although I didn’t get hit in this video, sometimes they’re quite mean. Imagine if they got to use the full complement of items, not just fireballs and bombs. Imagine if they wouldn’t only use items when you are near, although they’d probably drag each other down overall. Imagine if they didn’t drive on rails so you could mess them up properly. That goes for item-less collisions as well.
Professional: Items are disabled for everyone. Computer drivers are faster yet again. The driver in 2nd place barely managed to stay within draw distance when I crossed the finish line.
Expert: Even faster computer drivers. I could only enter 1st in the final lap. Rubber banding has them breathing down your neck. A few car/track combinations become a living hell. Captain Rock on Toon Island II is fine.
You’re rewarded for playing at higher difficulties: each tier unlocks one more goody if you win the championship.
http://www.notdoppler.com/supercrazyguitarmaniacdeluxe4.php
Can you tell what people were looking at?
The pedaling sounds you can hear seem to stem from a bent chain guard. I fixed the worst of it but apparently it's still quite audible when your ears are attached to the frame. Maybe I should get it looked at, although the remaining rubbing shouldn't cause real issues.
The hardest part about this game is definitely the bad controls along with that affront to physics. Both slow to react and really touchy. Makes it look like I intended to run into some walls if you don't know what's going on.
Unlike what some people have said, you don't really have to grind in this game if you get the right parts and use them well. I won every single one of the required races without having to do others to get more money. It only takes an hour that's about as varied as the game gets, which is why I ran it now that I learned that. This time should be in the ballpark of optimality as well.
This is the PAL version running on a somewhat modified PCSXR 1.9.95. No CD audio because something about the Mac version of PCSX or my custom image format/plugin (images weren't natively supported at the time I ripped it) doesn't support it, not to mention the physical disc was badly scratched. I was lucky the data track could be read in its entirety by a computer; a physical PlayStation often froze while loading. The music wouldn't be a good experience with all the interruptions anyway.
Most modifications are to graphics. This is basically DFXVideo with a new GTE Accuracy path and a homemade clipless homogeneous polygon rasterizer. The result is perspective correction, support for vertices outside the usual 2048x2048 guard band including behind the camera (edit: vertices were clamped to z≥0 here before being fed to the rasterizer), the equivalent of approximately 11 subpixel bits, and the occasional polygon looking ridiculous when different vertices end up in the same bucket or only some vertices are sourced from the GTE. And although it isn't literally perfectly precise like my custom 2D rasterizer which even applies the fill rule to edges between texels as if with infinite precision, it is perfectly robust. Unlike standard DFXVideo which isn't even accurate, I should add.
I also modified the video output to remove an unnecessary frame of latency of the Mac version (as if uploading on the order of 100000 pixels to the GPU warrants a multi-frame pipeline), and to add biquintic resampling in a linear color space. Biquintic probably isn't the best idea, but it's cool to two-up bicubic. It improves acuity (although not as much as windowed sinc with the same support) while representing cubic polynomials perfectly and improving smoothness to C². If the outer lobe were more pronounced than it is with coefficients focusing on these properties, ringing would be quite abhorrent.
Maybe I should've investigated the aspect ratio. Everything has a fixed aspect ratio in DFXVideo (usually 4:3), which only uses parameters to determine which pixels to use. This game should be 8:5 or so to have the least distortion. I can change that if I just think of it before starting the recording, the same way I didn't think to put my traditional rasterizer back.
Doing a full disclosure here to make clear that I haven't done anything nefarious. I didn't even bother to check if #define BIAS 1 works with this game. Would be nice not to have those sometimes severe drops in frame rate. Mind that the real hardware also had many of those drops; this emulation is fairly purist-friendly for something that isn't cycle-accurate. Either way, I'm probably slightly disadvantaged compared to real hardware, and this is too beatable for bickering.
I had to delete a rant on less relevant details of my rasterizer and ideas for changes. Now you'll never know. Thank YouTube for saying my description was too long. Thank people who fill their description with garbage tags for most likely being what encouraged the implementation of this limit.
The R22 has a marking at 102 knots. In testing, it fell apart somewhere between 120 and 140 knots as shown by data output. That's pretty difficult to reach during forward flight, let alone backward flight. You can even get weathervaning to pitch it over if you apply too much cyclic; it's not just yaw. I can do around 100 knots, tops. My solution is to gain some altitude, get 90–100 knots there, and start descending. That way I can build some more speed without asking more of the angle of attack and rotor system. It becomes ever more difficult as the speed builds, but as you can see it worked.
ANCHOR DEPLOYED!
Bonus material: at 0:44 you can catch a glimpse of a collision scene at the adjacent bus crossing to the right. Given the large entourage of emergency vehicles that must've been serious. I hope you're not here to see it in detail, because I wasn't. Witnessing a fatal car–cat collision as it happened a few days ago was gruesome enough.
1. Open a new tab.
2. Paste a search term. Doesn't really matter what it is.
3. Switch to a different tab to cancel suggestions.
4. Switch back and click on the field again, not triggering suggestions.
5. Hit enter and immediately switch to a different tab. It probably helps to use keyboard like I did.
If you're 'lucky', the other tab is entirely supplanted. You can't even use back to go back to what used to be there. It doesn't happen every time but it only takes me a few tries this way. A search term that hasn't been used before seems to be more successful, but isn't necessary; the search term I used here had been used a few times before I got a good take.
Here I'm running five versions at once to allow for quick comparisons. All are locked to 60 fps without vsync. I'm going through a routine in each:
* Look around horizontally and vertically and see how the selection box updates. 1.7.10 and both suggested fixes track perfectly. 16w38a is behind by a tick. 16w40a merely kills interpolation, so every 20th of a second it jumps to the crosshair and stays there even if the crosshair moves out of that box before the next tick.
* Do the same thing in third person, so you can see the head model and how the selection box is supposed to be obscured. The head's pitch always tracks the crosshair perfectly. In all versions but the one with my second suggested fix the head's yaw is behind, corresponding exactly to the selection box in 16w38a.
* Sprint-fly (unfortunately slow in 1.7.10) while looking in the same direction. In all versions but 16w40a the selection box is right underneath the crosshair. In 16w40a it jumps around and leads the crosshair at times. That's an artifact of killing interpolation.
* Build a wall and place blocks on it after/during a quick diagonal movement. No matter how I try to time it, 1.7.10 and both of my suggested fixes won't budge. Most of the time it works. If my timing is off the block appears where my crosshair actually used to be, or doesn't appear if I clicked on air. 16w38a and 16w40a are identical: the block appears at the pitch of the current tick but the yaw of the previous tick.
My first suggested fix is like 1.7.10. My second suggested fix doesn't change getLook in case they want to keep current mob behavior (which seems to be an improvement, although I haven't investigated it). The one side effect is that the head model's yaw now also tracks instantly. In a third permutation you could copy Entity's getLook implementation to EntityPlayer, reverting EntityLivingBase's version.
I'm not sure how Grum doesn't see what's going on. Then again, this bug has existed for over two years before it got any attention at all.
2:17 Some sharp turns, the second one on slippery sand.
2:37 This light tends to detect you in time but you can pretty much ignore it regardless. At head height you can see there's nothing there.
3:43 This cycle path is very new. It used to be more of a mess, with the part at 4:22 being little more than a narrow sidewalk/pavement providing access to the gardens, where you could legally cycle. An improvement.
4:56 I wish I could see a thing.
6:48 Rough asphalt wasn't bad enough?
8:09 Watch the red van. At 8:32 I thought I'd better brake given the blind spot, but this is a breed of driver that pays attention. Hats off to him!
9:11 Train!
9:47 I was expecting this guy not to crawl along. Why would you want to be stuck behind a girl?? :-)
10:00 Cyclists dismount. This epidemic has got to end.
10:30 This driver got close to wrecking me. You can tell why I don't trust them.
10:43 This is a cycle path as well as access road. It confused me with the way drivers are passing on the wrong side. Not such a good idea after all.
11:29 I came to a full stop only to find I could be slotted in immediately. Why can't we have timely detection? There is even an induction loop at 11:26, and then again at the stop line. But no, gotta press the button.
11:49 Stopping for 56 seconds and almost no conflicting traffic. The guy I overtook at 11:00 caught up and jumped the red light at 12:28 with no issues. I can't blame him.
13:03 But he's a smoker with a dirty moped alongside, so I maintained some distance. At this point I was tired of stopping so I got ready for a mild sprint. Not that it could last because I knew what's past that tunnel.
13:41 That's right, a cycle roundabout. No mandated structure; you could go around the wrong way. I don't think it even helps cyclists that much, let alone in this form. I sure didn't like those curves. You can't just build something out of context. I guess it beats a level crossing with the large road above, but it has more prestige than thought to it.
14:00 What's this doing here?
The puzzle: http://fivethirtyeight.com/features/can-you-bake-the-optimal-cake
6:18 I wonder what was up with that magpie.
I'm not sure I want my channel to be full of these. I already made a report unlisted. What do you think, subscribers?
Before I started recording, DangerousBoy and WesternWizard stopped in front of me. WW soon rammed me upon which I reported in-game. The same fate befell DB moments later. Recording started 2016-01-08T20:06:05Z (that is, UTC).
I don't know about Miracle. Miracle was the one spanning lanes and tailgating at 1:15. Again tailgating at 3:20, then going off-road in lieu of braking. Note my right mirror at 3:36: that's Miracle. Miracle may not have actually wanted to hit me but wasn't acting like it. You be the judge of that.
My lane change at 3:21 was when I saw the lane I was going to be in was clear (and in fact, I was close to getting hit in my current lane): I didn't really cut in front. I'd delayed that lane change in an attempt to send WW down the wrong road.
http://www.marops.net/portfolio/trailer-reversing-game
I recall it ran surprisingly well. I just had problems with coasters breaking down. You only need one to break down to halt the whole thing. Before every run mechanics always had to fix a bunch of coasters.
In the second I was casually spinning out while descending the Hollandse Brug in the other direction. Hit a peak of 54 km/h without trying. I climbed it as fast as I normally ride on flat roads, with less effort.
Some people think it's hills that make cycling hard. That cycling can only work because this is such a flat country. A flat country, open to winds that can occasionally slow your descents to nil. Yeah, right.
Footage is slightly old. Note it was recorded in medium mode, which doesn't really look faster than it is, as you might be used to from wide footage.
0:16 Look how much faster.
1:29 Look where you're going!
1:49 Moped is turning left. Easy to see in person, hard to see on fisheye footage. These things are rarely used for longer trips.
2:44 Look, there's a reason you're only supposed to ride two abreast. Also, the "Begrip voor uw dank" (understanding for your thanks) sign is still up. There used to be a random "except 🚀✈️" sign at 1:02. They must've thought they were exquisitely funny. I'd appreciate more seriousness when it comes to cycle infrastructure.
3:54 These puddles are practically permanent. Fix your roads!
6:05 Relatively clean now thanks to dry-ish weather. It'd be a cesspool.
6:20 Why is this suddenly very narrow in the middle of cornering? Note how I felt the need to get off of the relatively smooth asphalt. It's good that it's an option, but you shouldn't have to use it under normal circumstances.
6:41 One of those vehicles meant for disabled people, but that anyone can use anywhere without a license. Also approaching two of the 'slow' mopeds. I'm holding back because people cross without looking all the time here.
7:33 Stupid 90° turns. A serious cesspool due to those working on the rails. Again relatively good thanks to recent cleaning and cooperative weather. It stretches all the way to 9:24 or so (where you only had a couple decimeters of 'good' space a few days ago). Also note the driver on the cycle path. Wha?? Stay on your coarser but higher-quality network!
9:18 Smoker. Such horrible air around them.
10:30 This place also attracts a lot of mud as well as leaves. I now habitually take this corner at the wrong side if I can get away with it, because it's a better line and it doesn't get as slippery as the other side.
11:00 I'd almost forget, but why do we still have these bumpy bridges?
11:06 Of course I'm overtaking you.
11:35 This detour is seriously annoying. Major cycle routes are fair game when anything is happening nearby.
11:52 Battery ran out. It now lasts 45 minutes in this mode. I'll probably get new ones at some point.
Again a list of complaints. Don't get me wrong: I like cycling. This won't quite stop me. But I wouldn't respect it if I didn't point out the problems.
My goal was to cruise at 30 km/h as a nice reference speed. With the heavy gusty crosswind that was a little too ambitious. I still got very close in both directions so it's a fair comparison.
This is intended to be reference material. If you don't want to see 92 minutes of mostly nothing special happening, simply don't watch. I gather some people find it relaxing, though. I don't care what you do. Just don't expect it to be action-packed.
The group had been quite slow. I wanted to check out the front but was held off by the very wobbly girl in front of me. There was finally some room to overtake properly here. It turned out Ymte (of Intercitybike, co-located with Tempelman), who was leading the group in his green Strada DF-XL velomobile, also had plans on that stretch. I know he wasn't trying, but by the time I caught up we'd left everyone else way behind. He may not even have pedaled after reaching the top of the bridge. He certainly wasn't when I rode alongside him. These velomobiles are legendary, not just experiencing very little drag but also maintaining some speed in a crosswind.


