Traditional LCDs do a "zero-order hold" to fill the theoretically infinite time between frames:
http://dspguru.com/dsp/faqs/multirate/interpolation
But interpolation then requires filtering. Are the 24 to 60 FPS judder artifacts the result of the lack of filtering because it's just "upsampling?"
Are frame interpolation algorithms just too expensive to produce a transparent (no judder or artifacts) sequence in real-time? I imagine that each component of every pixel would need to be treated like a separate waveform, with zero-stuffing and then filtering.
In contrast but analogous to LCDs, plasmas use impulses of light instead of the zero-order hold of LCDs. So a plasma might quadruple each frame of 24 FPS video to produce 96 FPS. Isn't this basically a less-perfect zero-order hold with zero-stuffing and no filtering? Somehow that's perceptually better?
Does the plasma method simply look transparent enough to not require a filter to make it true interpolation?
Obviously video requires a lot more data than audio. Is that the only reason for such different approaches to sample rate conversion?
Also, I suppose shutter speed influences how much blur is contained within each frame, which I guess is like the noise floor. Yay photography.
edited 8:51 PM EST January 12, 2017
edited 4:14 AM EST January 14, 2017
BUY THEIR PRODUCT! Just kidding, slurp the intellectual commodity from their pores.
screenshot of final histogram
Maybe I'll go back and finish Infinifactory... Looking forward to whatever Zachtronics comes up with next!
Now if only Marc ten Bosch would release Miegakure...
edited 4:23 AM EST February 5, 2017
Fine, set my TV's RGB range to match. I notice that Super Smash Bros. U has colors that aren't saturated enough. This is a constant nuisance during gameplay. This is the only game where I've noticed this issue. Yay 1080p, 60 FPS (except in some dumb menu transitions; somehow that's difficult to perfect), with color that's inferior to Brawl and Melee. Something about having cake and not being able to eat it.
If you want to see something particularly disgusting, every single video in the game has a compressed (too high) black level, which makes the video look greyish and washed out. Nintendo botched the transition to HD color space in multiple ways. When you choose a design scheme centered around the primary colors, I should be seeing the primary colors. There are multiple uses of grey outlines in the menus interspersed with various levels of black, which makes the outlined things... un-poppy and muted. I hope that hacking will disable the 16 - 235 conversion, at least.
Nintendo: "Oh my gosh 720p+! Let's make the extra detail provided by the extra resolution irrelevant by bloom lighting the hell out of games that should have nice, vivid colors, like Mario Kart 8 and Super Mario 3D World!"
If Wii U hacking ever becomes stable, I look forward to desperately searching for a code that disables bloom lighting. Like the no bloom code for Gamecube Twilight Princess. Stop filtering my damn image. My retinas should burn with delight.
Got 5 of the 14 achievements, might go back and look into the others but probably not... A weird experience.
Should be able to play The Legend of Zelda: Breath of the Wild some time tomorrow, I'm hearing good things though I'm trying to avoid spoilers.
Also played Mushroom 11 and Mini Metro yesterday, I got these with the recent Humble Freedom Bundle. I'd been waiting to play M11 since I attended a talk the designers gave on their level design and tutorial philosophy. It's a pretty neat mechanic, you kill off part of the "mushroom" and it grows back elsewhere, this plus physics makes for a lot of possibility.
edited 8:04 AM EST March 4, 2017
On a serious note though, I don't get all the hate I'm seeing for the game on youtube and online I mean IGN gave it a 10 rating but then again you can never trust those guys either. My experience on the other hand has been fantastic so far, the graphics and art style are simply beautiful and the combat system is amazing! However I have to admit, it takes some getting used to but once you get the hang of using the flurry rush moves to their fullest potential then you can pretty much conquer this game.
The music is quite wonderful in it's own aspect and fits perfectly into the world of hyrule. The chemistry link has with everyone he encounters on his journey is absolutely splendid and so is the story and all the nitpicking homages and clues to previous zelda games and their lores.
I have to say this is without a doubt the best zelda game I've ever played in my life and on my top 10 list of favorites:
1:Breath of The Wild
2:Twilight Princess
3:Wind Waker
4:Majora's Mask
5:Ocarina of Time
6:Skyward Sword
7:Spirit Tracks
8:A Link to the Past
9:Minish Cap
10:Link's Awakening
edited 8:37 PM EST March 5, 2017
2. Ocarina of Time
3. Majora's Mask
4. Oracle of Seasons
5. Oracle of Ages
6. A Link to the Past
7. Wind Waker
8. Twilight Princess
9. [Imaginary hack of Skyward sword that replaces motion controls]
10. Four Swords (with bonus single player mode which I haven't played because necesito emulacion)
If only Nintendo would release a gimmick-less, high-quality console once again. The Gamecube, if you will. Will you?
Nah, gimmicks sell.
Been playing a lot of BotW, not sure how I feel about it, besides wow that's a lot of art. Easy to immerse myself in. The bite-sized dungeons are weird... not my favorite aspect, even being a fan of physics puzzles in general. Haven't got the hang of combat yet. I'm beginning to feel bad about stripping all the natural resources from an area, then coming back to see that the game remembered.
edited 5:32 AM EST March 6, 2017
I'm sure you'll get used to fighting enemies, really as long as you always have your shield up and raised, have plenty of weapons collected and make successful dodge maneuvers you can defeat any foe.
I actually like utilizing the games surroundings and some of the world items do respawn after you restart the game so it's no biggie.
Regarding music so far, I've been frankly disappointed, though I guess it fits. I've had the Talus Battle music stuck in my head, though.
I agree. From what I've heard of the soundtrack(out of gameplay videos, mind you), it tries to have this "minimalistic" feel to it but instead of actually achieving it they made the whole thing feel moody just from the piano compositions alone.
The game deserved a better soundtrack than this. It's no wonder I stopped caring about the game early on, especially since the Zelda series as a whole is uninteresting in my opinion.
I'm at 80 something shrines now and already completed the game, I do hate that the bosses of the beast dungeons were lack luster. I wish nintendo sticked to the traditional style of zelda boss fights like Twilight Princess or Wind Waker had, manifestations of calamity ganon are boring to me since all 4 are no different from each other in terms of attacks, strength or shapes and forms.
The shrines are occasionally brilliant!
It's probably the case that the old school melody-heavy music wouldn't have suited this game, but I'm still disappointed. Many of the towns have hints of classic themes, but I wasn't able to detect that in Gerudo Town or Kakariko Village.
But I'll argue that there are good reasons to have ww2ogg, for direct lossless conversion to a format that many players handle natively these days. Still would be good to have in vgmstream, though that'd again require additional metadata in most cases (or the decoder would have to try different things until it decodes exactly the right number of bits in some large number of packets without any errors), and I'm not sure how to handle codebooks (I guess they could be dumped to C? Or would it be better to just keep the binary packed_codebooks with the meta supplement?)
Part of the deal with MTA2 is that it is like MPEG in that it has more than just audio in the file (subtitles, lip sync, video), and there can be multiple audio streams, so you usually want to demux it to get just the audio for our purposes.
To be honest, I've never imagined MGS4 could use a format that did more than just audio(especially since "MTA2" is actually a newer version of the MTA format that originated on MGS3 as "MTAF" - the only changes from the former to the latter was an entirely "new" codec alongside more capabilities as the ones you've stated).
Perhaps that`s why your tool could only extract audio from MGS4 archvies alone, compared to sdt_demux.py by Nisto which could extract nearly every filetype imaginable based on .sdt-like headers(so long as a single file is accepted), but I doubt I`ll modify the script for MGS4 anyway considering I`ve been stuffing my HDD with a few EA and Ubisoft games for research purposes.
"I was mostly agreeing with you. Any butthurt is due to an uncomfortable chair."
Well, I stopped reading after the first paragraph for the first time for fear that I would provoke that reaction. Really.
Well, speaking of .wem files... they do have loop notes, though. There is a "smpl" chunk in a .wem header, so why can`t ima_rejigger5 support that too?
edited 8:14 PM EDT March 19, 2017
In other news, just had my ass saved by install steps 11-16 of here; I tried to install Ubuntu on a Mac where I was already dual-booting Windows 10, after that Windows wouldn't boot. Turns out I had to kill the hybrid MBR which apparently the Ubuntu installer had added, all 3 OSes work fine now. Whew!
I thought it was fun being the only person who didn't jump on each of Apple's latest incremental OS X upgrades (which provided no significant new features) in an Apple workplace, and then pointing out the crashes of oh-so-friendly Apple software as the no-longer bleeding edge OS version became stale.
Safari with its uber-smooth scrolling, I'll admit, is the best way to browse the Internet, though. Too bad Microsoft doesn't take a hint.
You mentioned before vgmstream's MS IMA decoding may miss some samples, so out of curiosity I tracked down Microsoft's C implementation.
What they do is write the block header sample, but discard the block's last decoded sample (while vgmstream ignores the header sample and keeps the last sample).
The last block sample and the next header's sample may be the same, or vary slightly. This means vgmstream misses the very first header sample (so in essence the file is +1 off) and some samples are (unnoticeably) off.
Could be fixed but not exactly a super important.
Is Wwise using this feature somehow? From the ima_rejigger readme it kind of sounds like this but I'd appreciate some pointers before I delve into the code.
Wwise doesn't use the final decoded sample, the final nibble is always 0. Maybe that's the same as what MS's official implementation does, but I doubt it. When I tried to "rejigger" wwise (it's also interleaved differently) and set codec 0x11, every decoder I tried (sox and ffmpeg) had artifacts.
edited 6:01 PM EDT March 21, 2017
Though it would make sense if this was different in some Xbox-specific stuff, the above is about RIFF WAVE. Maybe it would make sense for vgmstream to use the header sample and skip the last sample. Especially if it turns out to be always 0 as in wwise, though I'd expect some noticeable artifacts from that that I'd hope we'd have noticed.
edited 6:10 PM EDT March 21, 2017
It's og Xbox's (cough XDK cough), but the docs say it's a wrapper around Windows MS IMA with a fixed block size.
I see the decoder parses "cSamplesPerBlock - 1" but in all honestly I didn't build it myself to test it.
However ToWav, which I think just uses MS's libs, also agrees with this and writes the very first sample.
It's not unthinkable every decoder out there is incorrect, I guess, since the difference is very small.
Incidentally I don't think any PS ADPCM decoder is HW accurate either (I've seen several slightly different implementations), but I don't know enough to evaluate and fix it myself :/
It would be informative to try out some sample data with and without the fix you've proposed, or add a check for the last nibble == 0 to see if this holds up.
edited 6:39 PM EDT March 21, 2017
http://www.eurogamer.net/articles/digitalfoundry-2014-bayonetta-wii-u-face-off
Remember, the people at Digital Foundry are paid to point out technical aspects and issues in today's increasingly buggy ("unpolished" if you're feeling generous) games. Please, spare me the "wah graphics is hard defend defend stand up for your imaginary friend."
From a couple hours of casual play, here's a huge issue which they completely missed: In gameplay and cutscenes, the top third to sixth of the screen has an OBNOXIOUS, CONSTANT, BLUE BLOOM LIGHTING BAND. IT'S MORE OBNOXIOUS THAN THESE CAPITAL LETTERS THROUGH WHICH I CHOSE TO EXPRESS OUTRAGE, BUT HOPEFULLY YOU CAN ASSOCIATE THE SEVERITY. !!IT FOLLOWS YOU AS YOU TURN THE CAMERA!!
Figure 1: In this particular screenshot, the blue band isn't as noticeable, but look along the horizontal line near the bottom edge of the combo display.
[http://images.eurogamer.net/2013/articles//a/1/7/0/9/8/2/6/wiiu_5.bmp.jpg
See? I told you.
Figure 2: These guys got close to unearthing the (it really isn't subtle as long as your eyes are open) bug. Scroll down to the over-saturated images to really see it. Not present on Xbox 360.
http://forums.sega.com/showthread.php?517493-Bayonetta-2&s=be2870e3de145565f5f28b42dc0d18a8&p=8812482&viewfull=1#post8812482
If only I had a better screenshot of darker buildings/in a tunnel/in streets of the city. It's bad.
It ruins an otherwise beautiful (graphically) game.
Smaller bug: Some fullscreen graphical effects don't actually extend to the edges of the screen, so a pixel edge displays brightly against rest of screen. And this artifact is asymmetrical (two out of the four edges). Do people pay attention anymore when they develop games? Or are they the "artsy" types who can't be bothered with the details?
Subjective, review-type comments:
Still in the second chapter of the game, not impressed. Gameplay is frustrating. Free camera system is awful/gets stuck in corners. Audio balance is off: voice overs too low. Game feels quirky for quirk's sake/attention whoring. They say "Fuck."
Devil May Cry: great. Viewtiful Joes: awesome. Bayonetta: my sources say the popularity comes from uniqueness/lol Japanese behavior, not quality, difficulty comes from bad design. technical mess. jury is still out.
Oh, and if I die at a boss, let me restart right at the boss. Don't add loading between the continue point and said boss. K? Every game will be good from now on. Go!
Bayonetta wasn`t "the" kind of game I could play either, but given Hideki Kamiya`s game design philosophy I guess this is to be expected.
Back to vgm, Mark Brown's thoughts on BotW's score. Shouldn't surprise anyone here, though I'll admit I did miss Gerudo Town's callback, and I hadn't made the connection with the shrine music. Clock Town Day 3 (or Last Day) would probably have been a better example for MM.
The banner jingly thing is from Link's Awakening dungeons. Hyrule/Ganon's Castle sounds cool (haven't been there in game yet), maybe better than what they did with them in Twilight Princess. (I don't think anything compares to Farewell Hyrule King from WW, though).
AnonRunzes, I think you mean the opposite of "gloss" (as in "gloss over"); perhaps "obsess," as in me "obsessing" over little graphics details instead of seeing the "big picture" (pun-tastic, and, well, 1/3 of the screen is considered a big deal by leading carers [citation needed]). Video games are the #1 art form, and I want them to be right. Negligence bugs are frustrating, and they forever stain otherwise impressive works. Passion and such.
edited 6:46 PM EDT March 24, 2017
That's the type of argument you've created, more colloquially known as "putting words in my mouth." I'll play along. After all, the Internet's just a big, unfiltered debate. I attract 'em like flies to a dung heap.
https://i2.wp.com/s2.quickmeme.com/img/8b/8bce00e2853c6e2114fa905d9e8feaba2ba1bc0753ab4c60a0ff55c99fb337a5.jpg
Uncontrollably evocative!
To answer your question, I'd play well-made games instead. I wouldn't have purchased the game if I had known of the glaring bug beforehand. Not that the creators would have noticed. I'm the first person on the entire Intertubes to report it, so go me? I feel better, plus this has got to earn at least a few Internet points. Bugs this severe shouldn't make it to production, like the good ol' days when video games weren't simply a cash grab. Unlikely that my report is now miraculously respected, but what can you do.
Last word? Hi-ya! (As in the martial arts grunt, not the greeting, because that would sound nerdy.)
I don't know whether to recommend it, though I don't regret having played it, and it was worth the price of admission to me. Odd that I should feel so hesitant.
---
Something that has bugged me when trying to understand an abstract concept, like category theory or the concept of monad in particular, is the absence of a negative example. Ok, this and this and this can all be a monad because of abc. But is there anything that is not a monad?
Try Googling "What is a monad" and you'll find tens of thousands of explanations, but "What is not a monad" is mentioned (until now) exactly once, without any discussion. I'm sure it's come up somewhere but I haven't run across it explicitly yet.
I feel like a negative example is needed to really throw the boundaries around something. If everything is a foo, then a foo doesn't really have any distinguishing properties.
The trouble may be that a monad is both a set of things and what you can do with those things. Integers are considered a group, for instance, but only when taken together with addition. But functions, and the pairing of a function with a domain of objects it operates on, are objects we should be able to discuss.
Using the group example, we can point to examples of candidates which fail every axiom. Knowing how the assumptions can be violated points towards why those assumptions are interesting:
Closure: The set of all integers but 1 and -1, with addition, fails here. We need 1 because -9+10=1, etc. Indeed removing any one integer will (at least) break closure. (But all even integers works with addition!)
Identity element: Just the positive integers with addition are not a group, as there is no identity element (0 for addition)
Inverse element: The non-negative integers with addition fail due to all positive integers lacking corresponding inverse elements (-x for x); similarly the integers with multiplication (missing 1/x for all x besides 1 and -1)
(At first glance if there's no identity element the inverse element axiom must always also fail, but if the set is empty the inverse succeeds vacuously, yet there is no identity.)
Associativity: It's tricky to find something that only disobeys associativity, this is apparently called a "loop". Seemingly the primary example is the nonzero octonions under multiplication.
Earlier I had integers with subtraction here, but that misses identity as well: There's no i such that i-0=0 and i-1=1
---
This table sort of addresses this point. What fun, category and monoid are on there, and as we all know, "a monad is a monoid in the category of endofunctors, what's the problem?"
I may get this yet.
edited 6:21 AM EDT May 2, 2017
edited 8:06 AM EDT May 2, 2017
Gettin bogged down
Like Cashel Man
edited 9:08 AM EDT May 2, 2017
Huh. Let me check that out...
>it`s yet another "narrative" walking simulator
Come on, who likes games like this anyway? I can count a bunch of games that adopt this kind of boring gameplay such as Dear Esther, Gone Home, Everybody's Gone to the Rapture, The Vanishing of Ethan Carter(although that one really looks interesting)... really, there are too many of them!
edited 9:50 AM EDT May 2, 2017
I liked it, it worked pretty well for me, but I realize it isn't for everyone.
But I doubt there's any work of art that everyone will enjoy, or find interesting.
edited 5:50 PM EDT May 2, 2017
Maybe it was sloppy to use such a loaded term. The intention was to cover many different media, such as music and film, where it's understood that there is no accounting for taste.
At least as far as I got, it's not impossible that it gets really interesting later on. But AFAICT there is no reason to play it if you already know the basics of this stuff.
I TA'd a Computer Architecture course once, though there we used a graphical circuit layout system (Logisim). That had its own challenges, but overall I'd say it was a lot more fun to screw around with than MHRD was to play. MHRD is actually considerably more efficient to work with, I think, though you probably have to work things out on paper or with a lot of comments if you are figuring it out.
So I may just be too familiar with these basic exercises to enjoy plugging it into a game which has no other charms, YMMV.
I had fun with Human Resource Machine, which was at heart just a set of simple assignments to complete in assembly language. The most complicated problem was a sort, IIRC. But it had many highlights:
- It's fun to mouse around, even if it is probably less efficient, particularly in an attractive interface that Tomorrow Corporation (and formerly 2D Boy) does exceptionally well
- Having a "par" for a level to beat is a powerful motivator for me (even better if it's a histogram like Zachtronics is known for)
- There's even less story and characterization here than in World of Goo, but it helps to make things a little more interesting
- Animated execution is hugely helpful for debugging and interesting to watch
- Fun music!
I think if the Tomorrow Corp folks did exactly the same digital logic exercises with their trademark level of polish, it'd be a lot more fun to play. Maybe this makes me shallow, but I don't think I'm alone.
Of course better still would be better problems to solve!
MHRD does have a gate count for each solution, and apparently it compares you to the best globally and among your friends, though this is hidden in the settings menu for some reason. I don't think this works as well as either a fixed offline par (which the designer can arbitrarily choose to be not absolutely optimal if optimal is boring) or a histogram (for those who aren't going for #1 but still want to know where they stand), though with problems with solutions as simple and standard as these there's bound to be clustering around the high end anyway.
---
I don't like to be so negative, MHRD really does seem to do what it sets out to do pretty well. I just wish it was a bit more ambitious.
edited 5:08 AM EDT May 5, 2017
Tsunami Marijuana
"Get to higher ground"
---
I read through Unsong the other day after Scott Aaronson mentioned it was finished, it's a take on "what if all the weird stuff in the bible is actually there for divine reasons", or at least most of the jokes are. Lots of numerology and puns.
*** minor SPOILERS ***
There's a minor character whose name is never spelled the same way twice, which was hard to notice due possibly to "typoglycemia", which I thought was one of the best jokes in the book when I realized it.
The main backstory is that the universe used to run on god's design which permitted supernatural happenings, but an angel hacked together a mathematical physics system to run on top of it in order to deny the devil control over reality. Then the system breaks when Apollo 8 hits the crystal sphere surrounding the earth... and magic of various kinds returns.
Recommended! If a bit self-indulgent and eye-rollingly meta much of the time.
---
Funnily, I had just read The Traveller In Black the day before I started reading Unsong, which has a similar idea going on: magic/chaos is being replaced with science/reason/order. The Traveller is tasked with exorcising the demons of unreason which corrupt humankind, though I can't say I fully understand why his "single nature" is what it apparently is: essentially a genie who mostly uses your wishes to do whatever he would have wanted to do anyway.
I enjoyed struggling through the archaic language, it isn't perfect but I'd recommend checking it out.
edited 5:38 PM EDT May 31, 2017
From what I've heard, people hate the "smell" that is associated with them, because apparently, the operators and everyone who visits blazes up right there on the premises, 24/7.
They also have to work in cash, and never use bank accounts, because MJ is still federally controlled, so any business involving it is against federal law, so they can just seize your assets and freeze your bank accounts at will. Thanks, Nixon.
edited 10:17 PM EDT June 3, 2017
We also have an election for city council member going on, by mail only, and one of the replacement candidates is running solely on the platform of stopping a "monster warehouse" deal from going through. I can sort of see the reasoning behind stopping that sort of progress, since our local roads and highways are already clogged with 18 wheelers at all hours of the day as it is.
I`m starting to learn MIPS through this mips-iv.pdf doc I found around the internet so any tips are appreciated.
I could use a "XBE" preset and set the processor to something like "Intel Pentium III" but then the program crashes. And I don`t feel like actually buying one of the most expensive disassembles ever so I`m waiting for some group to crack the newest version there is if available.
I've also been working on a PSF2 ripping guide and some accompanying tools for some months now, but I haven't had that much priority on it lately since there doesn't seem to be much interest anyway. I'm hoping I'll finish it soon though..
edited 6:30 PM EDT June 4, 2017
The concert was very enjoyable, I look forward to the next CD coming out, they did wonderful arrangements of at least FF Chaos Temple and FF5 Home Sweet Home that aren't recorded yet.
---
While I was there I picked up a library card for the Clark Library on campus, and took out a copy of ENIAC in Action about the life and evolution of that computer. I've long wanted to read this. One of the interesting things is that while it was likely the first stored program computer to run (following modifications to its initial hardwired sequencer), this wasn't considered a revolution at the time.
---
I just turned on the key outlines for the virtual keyboard (Gboard) on my Android phone, it's amazing how different this feels now that it gives me clearer targets to aim for. I don't remember if I stupidly turned this off myself or if it was the default.
... And now I can't find that option again to turn it back off for further comparison. I wanted to turn it off for a quick typing speed test.
Oh, it's "key borders", and you toggle it when choosing a theme. Looks like it was off by default according to a few articles I've read. I think this may have been one reason I found someone's iPhone keyboard much better feeling.
edited 8:54 AM EDT June 5, 2017
I'm also catching up on my favorite tv shows on the CW over the summer, mostly The Flash, The 100, Supernatural and The Originals (only 3 episodes left of this season!) and finishing up some miniscule video game remixes in Mixcraft 8, has anyone used it yet as well? It just released earlier this year and I love it! Especially now since Celemony Melodyne comes pre-installed with it. :)
See this PDF which spells out the conventions.
edited 11:42 AM EDT June 7, 2017
No money to buy the awesome program (there was a full page of Squall Leonhart's scarred face over his gunblade! Lost forever!). The venue was awesome. The sound quality was great. I heard Koji Kondo perform the New Super Mario Bros. theme on piano before the game was released. Was his performance a world premiere? Yasonori Mitsuda was there. Nobuo Uematsu was there. And Silent Hill guitar guy. And others.
I attended Distant Worlds a few years later. Arnie Roth (has he played a video game in his life?) was more confident this time. The amplification at the venue was absolutely awful. Clipped to hell. When the audience was prompted, I yelled to indicate my preference for arranging "Battle on the Big Bridge" next. The populace yelled for "Jenova" louder... Or was it "Dancing Mad?" Don't remember.
The real reason I decided to get up and post today:
Rogue Galaxy
Dragon Quest VIII: Journey of the Cursed King
Both games developed by Level 5. Level 5 did great work developing screen-tearing-free Dark Cloud 2. Even previously quality-minded Square Enix produced Dragon Quest VIII.
Rogue Galaxy has screen tearing in the battle system. Otherwise pretty game.
Dragon Quest VIII has screen tearing... the whole damn time? At least while walking around the first town.
Now the Internet knows. Points for me. This report is based on playing the games on the PS2 from an internal IDE hard drive using Open PS2 Loader. I don't think the faster-than-DVD-drive loading is causing the screen tearing. I do think the faster loading is exposing duration-locked streaming music code, because entering and exiting the menus in both games makes the music hiccup. When faster hardware causes bugs, that's absolutely attributed to poorly written software. Faster than minimum hardware should only make software run even better.
Summary of huge bugs in aforementioned, lauded games:
1. Rogue Galaxy: Battle system screen tearing
2. Dragon Quest VIII: Whole-damn-time screen tearing
3. Both games: Faster storage medium causes streaming music skips during menu transitions
By the way, yes, screen tearing is absolutely a bug (it's objectively bad). Developers have two (funny how that number pops up so often in logic) choices:
1. Optimize your renderer to eliminate screen tearing
2. Render less to eliminate screen tearing
Option #3: Screen tearing, is invalid. Stop choosing it when you write code. K?
edited 12:27 AM EDT June 9, 2017
As for relying on timing, I don't know... With streamed audio on consoles things tend to be tightly tuned because the platform is standard, there's little reason to test what will happen with super fast seeking if there's no expectation it could ever happen. I expect even the dev units duplicate this timing even though it works off a hard drive, so that bugs don't only appear on retail hardware.
That said, Level 5 did do their own audio system for those games, so it likely didn't get as much care put into it as the SDK streaming system, which I suppose is more robust with this kind of thing if you're only noticing it in these two games.
Anywho not trying to argue too much, just wanted to bring up a little context.
It's funny how there's kind of the opposite problem with most USF rips (or so I am told): because PJ64 completes a lot of DMAs instantly, the tracer strips out the code that handles waiting for the DMA, so when it is run with real delays it crashes from that code being missing. Oops.
---
Play sounds like an awesome experience! I missed the first Video Games Live show at the Hollywood Bowl by a day when I was out there visiting, and even went there for a different concert! Still regret that to this day, even though I've seen VGL shows 3 times since and am kind of tired of it now. Speaking of obnoxious MCs...
edited 10:35 AM EDT June 9, 2017
Yeah, emulators and tight timing. But, like you said, that's the opposite problem. Emulators need to emulate the "minimum requirements" (AKA the hardware). I don't see how faster hardware would make correct software break, unless busy waiting/sleeping is inefficient?... Sounds like a software problem.
LazyUSF2, based on Mupen64Plus, has to disable lots of these timing delays in playback because of rips that will go haywire if timing is realistic.
I don't blame the rippers, either, since they used the tools available to perform the trimming. I sort of blame the trimming software, which is why I wrote my own coverage routines based on the Mupen64Plus emulator core in my library. When using coverage mode, it will enable all delayed timing as closely as the emulator core supports. It also forces full low level RSP emulation, even if that is still designed to execute and return instantly. Maybe that can be improved upon in a future version, if upstream hasn't already dealt with it?
(Or a new format made with hindsight that just runs a ROM...)
Sorry about the tracer/stripper/sparse tool. Even between the modded PJ64 used for ripping and 64th Note there are significant differences​. I should have known I had more work to do when I had to add a setting to include code within N bytes of any accessed word.
---
While I've got some edit time left, I've made myself a monument to stupidity over the last two days. I bought The Little Prover some months back, yesterday I got it into my head to really dig into it. So I spent about a day putting together an interpreter for the subset of Scheme the book uses, in JavaScript. Then I spent way too much time today manually typing in the proof assistant from the back of the book, when I knew the files were available online.
Fortunately when I was done and things didn't work I came slightly closer to sensibility, I grabbed the files directly, and things were still broken. I cross-checked with MIT Scheme and the code seemed to work OK there, so there was something I was missing.
That something turns out to be that quoted stuff is actually stored as (quote x), so (eq? 'quote (car ''x)) is true. This is similar to how all lists have 'nil hidden at the end, but I had known about that one already.
In any case I went back and checked the stuff I had manually keyed it, it was broken anyway so it's a good thing I didn't waste any more time on it. I'm glad my interpreter works (at least for the very few things I tested it on), but it's pretty slow. It's triggered by the 'change' event rather than 'input' since it was too slow to update on every keystroke, 'change' tends not to fire until the textarea loses focus.
edited 3:02 AM EDT June 13, 2017
Over the last week I also polished off an old pixel editor and worked on a triangle/quadrilateral sketch simplifier/recognizer for mouse/touch.
edited 3:06 AM EDT June 13, 2017
edited 3:08 AM EDT June 13, 2017
The key was that a profiler (always a good idea to use!) indicated most of the time was spent in name lookup. There are a lot of shortcuts we can take because of how simple the language is: no nested functions, functions aren't used first class, no redefinition, the only variables are arguments. So now most lookups are done ahead of time and I save an index to use when running.
What I'm down to now is most likely more closely related to the overall architecture, lots of recursion and most JavaScript engines still don't do tail call optimization.
[edit]
Actually, the performance may not be my fault after all. MIT/GNU Scheme runs it in about the same time (slightly slower) on my desktop: 734ms, vs 648ms for Firefox 54.
Of course that doesn't count the whole page load time which I didn't measure and which probably makes it a somewhat unfair comparison. But being within an order of magnitude of the official implementation in a real Scheme system is gratifying.
edited 4:18 PM EDT June 14, 2017
Will this catch on?:
https://projectreactor.io/
ecola
edited 1:23 AM EDT June 19, 2017
I have been thinking about my puzzle game dwim
again (mouse/touch only version here).
In the interest of coming up with some more puzzles along that line I wrote up a similar but simpler system in PuzzleScript. Currently Untitled, it has a few puzzles in place. Move the red block to the yellow block using only the sequences of moves available in the list.
There's an interesting board game/puzzle called Code Master that I picked up a while back and found pretty fascinating, don't think I mentioned it here yet.
edited 12:52 AM EDT June 21, 2017
edited 7:10 PM EDT June 21, 2017
I picked up a calculator today at Goodwill (Sharp EL-531W) that has a pental mode, along with the more standard binary, hex, and octal. I guess it is used in some Indian languages?
edited 2:56 AM EDT June 24, 2017
The design was a cool revelation to me and I'm pleased with how much I got out of it. It's the closest I've ever gotten to the Zachtronics approach of just coming up with interesting levels and then seeing how hard they are to solve.
I put ecola up on Show HN over the weekend, it briefly hit the Hacker News front page.
edited 2:45 AM EDT June 27, 2017
edited 8:56 AM EDT June 27, 2017
Second part of my two-part series roasting Bayonetta (and Digital Foundry). I recently finished Bayonetta Wii U. The lighting bug covers the top third (even up to one half) of the screen. We need rigorous reviews to avoid supporting buggy, unfinished games.
A decent aspect of Bayonetta is the music. I mean, it's okay. Maybe buy the soundtrack instead of the game?
I booted up Bayonetta 2 to (lol) see if the lighting bug still exists. Yes. The graphics programmers were incompetent. What a shame, too, since the Bayonettas' entire shtick is impugning (again) God, portrayed as just as fallen and petty as man. Videophile anti-Christians be disappointed. Don't worry, there are also plenty of audio bugs to go around.
Bayonetta 2 even has a scene right at the beginning depicting Santa Claus. The climax of which features the turd character screaming "I believe!" Subtle.
Play the God of War series for a better experience. One of those rare occurrences where Westerners made a better game.
New topic: Sly Cooper 1 has a bunch of race-condition-type bugs that are exposed when the game is loaded from an IDE hard drive using Open PS2 Loader. The first level is impassable (at multiple points, depending on your luck). The following are games that I've encountered which are unplayable using Open PS2 Loader:
1. Jak and Daxter
2. Madden 2002
3. Sly Cooper and the Thievius Raccoonus
I realize that these bugs could be attributed to Open PS2 Loader, but consider the following observation: Most games run fine (actually better thanks to faster loading times). I tend to play Japanese games because they tend to write better code (not always, lieutenant, not always).
Kingdom Hearts II: Stellar performance (albeit a few frame drops in certain situations). Kingdom Hearts II is no less complex than the above, broken games. Notice how the three problem games above are Western?
Faster hardware does not break correct software. The prerequisite is that the concept "correct software" exists, which is difficult enough to get people to believe. Ah, we've come full circle, now, haven't we?
Just starting to make real progress through the Frame-based editing paper (paper PDF here) that I noticed a few weeks ago. I like the treatment of break statements (figure 6): they have a little port through the margin connecting them to the scope that they break out of, which is a really nice idea. Not fundamentally different from IDA drawing lines from jumps to their targets, but in a mostly-structured program (as shown by the frames) it doesn't have to fight with as much other control flow.
edited 9:20 PM EDT July 2, 2017
hydlide 3 killed god three decades ago
Pyramids and Doors 7ED3-0000-034B-827B
edited 3:27 AM EDT July 29, 2017
The Fields
It's designed around a touch interface, mouse works as well but keyboard does not.
edited 3:02 AM EDT August 13, 2017
I'm making gravy without the lumps
edited 6:56 AM EDT August 13, 2017
Surf's up
edited 9:19 PM EDT August 13, 2017
Weave
Don't Fall
edited 10:28 AM EDT August 19, 2017
down with government
edited 7:22 AM EDT August 25, 2017
This time with a demo on YouTube
edited 3:34 PM EDT September 3, 2017
you got style
edited 3:35 PM EDT September 3, 2017
Random JS game of the day is Jetpack Joyride or similar mashed up with moving around a maze: Feat!
edited 12:41 AM EDT September 10, 2017
edited 1:07 AM EDT September 10, 2017
"What's going on?"
"Why aren't I dying/getting a game over? I haven't done anything."
"Am I supposed to catch/hit all those arrows?"
Oh, look! The blue box moved. Huh, why did it do that?"
OH! When an arrow hits the red circle, the blue box moves. How do I move the circle?"
"Spacebar. Ha, look at that. Now that I can collect arrows, what do I do next?"
"Wait a minute. There's a yellow box as well. I wonder if that's the 'goal' here?"
"Wow, this is annoying, hitting all these arrows. I can certainly see how some people could enjoy this (flappy bird comes to mind. I couldn't stand that game, and yet, somehow, it was a huge hit)."
"So, when I hit the yellow box, with the blue box, using a rather awkward control scheme (which I'm sure is the point. Obviously an easy game isn't fun either), the yellow box moves. Well, it is a 'proof of concept' type of game, after all. Even Tetris didn't have a 'point (literally. No point system was in place)' on its first go-through (the programmer almost didn't finish the game, he was having so much fun)."
Fun idea, for some. I'd grow tired of it far too quickly (my 5ish minutes was enough for me). Though, it *was* fun to discover how it all worked. I had a lot more fun figuring out how to play the game, than I did actually playing it. It reminds me of my first time playing MineCraft. I had a lot of fun figuring out how to play the "game". It was the discovery that drew me in. It was the exploration of the MineCraft world (again, a different kind of discovery) that kept me coming back. Until my PC couldn't download the game anymore (and my internet connection [aka - unsecured neighbor] moved away :( ).
Not for me (since I don't enjoy the control scheme, which as I said, I think the control scheme is the point), but if you could figure out how to keep the background grid stationary, I think that'd help others keep track of what they're doing easier. Nonetheless, nice idea, for those who enjoy this kind of thing. Mouser X over and out.
The long version essentially details a different game. The more I thought about what I'd do to fix the stuff I don't like, the more I realized my changes make it a different game altogether. For example, If I were designing it, I'd use a joystick, or some other pressure sensitive control method, to allow for "hovering." The "on/rise" or "off/fall", with no "hold/hover" is what gets me. I might try using the numpad, or any 2-key combo, really, where 8 is "on", 5 is "hover at current position," and no buttons/keys pressed is "off" (a PC mouse might function the same way). I'd have to test it, of course, but I don't see an easy way to implement a control scheme that doesn't frustrate me, using only on/off buttons (aka, most touch screens, and keyboards). Even if you had an analog button (Gamecube L/R, or Xbox L/R buttons), that would probably work. I've heard there's pressure sensitive touch screens, but I haven't encountered one myself (or, none of the apps available to me took advantage of it).
As I continued to think about it, I tried imagine what that might be like (a two-button control method), and tried to compare it to other games I've played and enjoyed. Asteroids came to mind (it has one button for thrust, as does "Feat!"), followed by Gradius, or Rygar. When I thought of Gradius, I realized this game is very similar to Gradius, but on a planet, where you have to deal with gravity always pulling you down (I have no recollection of Gradius ever having a gravity effect). When I thought of it that way, I imagined flying a ship (airplane, in this instance) where you always had to be pulling up, to fight gravity's pull. Which lead to me thinking, "What if, with the mouse (two button combo), one button(thrust)==one gravity, and two buttons(thrusts)==2 gravity, thus making you rise?" The more I thought about it, the more the idea intrigued me, until I realized that, at that point, it's a different game from both "Feat!" and Gradius.
LIke I said, regarding your game, I doubt I can offer much to make it fun for me.
Really, I just don't think "Feat!" uses my kind of control method. Obviously, there's people out there who like it. I'm just not one of them. On that note, I'm excited for the Switch (I'm buying one for myself for Christmas), where you *do* have pressure-controls available to you (at least, through the joysticks). Please understand, I think your game "Feat!" is a novel, creative idea. I just don't have fun with that type of control. And, while my 2-button idea intrigues me, I'm still not 100% on whether I'd actually enjoy it. But the longer I think about it, the more interested I get. Hmmm....
Also, I played your "PrograMaze" (I think that was the name). It was a lot of fun! I got to the final phase, when I had to stop. It was hurting my head too much, and I needed to go to bed (I still haven't gone to bed, but without such taxing thoughts on my brain, my headache has gone down. I really do need to get to bed though). I really like the idea of PrograMaze. I might try it again, when I've had a good night's sleep. Mouser X over and out.
Still won't make Mouser happy, though :)
Updated with a new scoring system, makes things a little more exciting, though there's still no game over.
edited 1:55 AM EDT September 11, 2017
That said, I found that there's a group of people that are attempting to create an "update system" (of sorts) that uses Winamp 5.666 as its base getwacup.com. I wanted to sign up on their forums (to help with beta testing), but I discovered that I can only login to my hotmail email account, and not my Yahoo account.... When I try to login, it asks for my login name (hit enter), then a password (hit enter), then it says it needs to secure my account. At that point it loops back to the login name. If I click "trouble logging in", it just keeps reloading the page. And it does this on my PC and tablet. :( I mention it here, because that's how I've been spending at least 30 min., though likely more.... Mouser X over and out.
edited 8:21 PM EDT October 8, 2017
Best on touch screens, though playable with mouse.
edited 7:57 PM EDT October 10, 2017
Two more intermediate levels to hopefully set up the last one better, last is now lev 6.
edited 8:53 PM EDT October 10, 2017