That's because of how the SNES handles transparency: Selected mainscreen layers are added/subtracted with subscreen layers (hence "translucent against subscreen") and this is what happens: Usually only layer 2 is on subscreen whereas layer 1, 3 and sprites are on mainscreen. The solution: Switch the level mode to 0E (used in the first room of the Ghost Ship) which puts layer 1, 2 and sprites to subscreen.
In addition to those above, Fireballs and the cape can interact with blocks as well. Those use only one interaction point.
There also is a separate block interaction routine used by line-guided sprites. It works with four interaction points, each four pixels alway from the sprite position.
Originally posted by Thomas
Now, the reason why Layer 3 interaction works a bit differently from Layer 1/2 interaction is that it doesn't process the interaction points for it. Instead, Layer 3 has its interaction hardcoded; for the tides it just checks if Mario's Y position is below a certain line or not.
Actually, the tide is secretely a foreground layer 2 level which is why you can change the tide's acts like setting with Lunar Magic since a couple versions ago and why you can use only half of the available screens.
There also is the layer 3 smasher which uses a classical hitbox which is why it doesn't act like a block and more like a sprite.
Three entries... I would have hoped more. Never mind, the number of entries were stagnating anyway, not to mention they are interesting and quite fun to play, that's the most important point. Here are all the entires entries for round 2:
To remind you on how to vote: You post the order of the entries where the highest ranked entry gets the most points. Don't focus on their names, presentations, graphics and audio but rather on the entries themselves i.e. how did they impress you, how much interesting did the bosses become compared to the original, how fun are the entries to play and/or what are the entries' quality. Participants cannot vote.
You have around three days to submit a judgement. That means, you can submit your votes for round 3 up until November 20th, at 18:00 UTC!
Is it a coincidence that there are as many entries in this round as there were in the last round of last Mad Scientist which was about bosses as well?
Disable the echo i.e. remove lines 40 and 41 in the Toad House music. The echo is one of the most common reasons to crash the music engine and while AMK usually fixes most issues (by checking how much space the music engine takes), it still has got that limitation within global songs.
and this round's winner is... Darolac
with the entry of Custom and score of 60 points!
This round surely shows how much a difference modifications can make to bosses and each in different ways: Darolac for customisations, KevinK for an interesting twist and NerDose for giving a new spin (\o/) to an overused boss. I won't force you but I surely hope you'll release your modifications at some point.
I'll reveal the winners and open up a feedback thread in an hour, please be a bit patient.
Since four years, there hasn't been any proper ASM contest (and before that, there were rare either if they did exist at all). As such, it's a good idea to have some feedback to improve on (possibly) later ASM contests.
How frequent should Mad Scientist be?
Do you think we should continue with public judging much like the music contests or should have have specialised judges similar to the other contests?
Should we go with multiple rounds (idol styled) or with a single large round (*LDC styles)?
If the former: Should we give a prize for the most scoring user per round or count the score in total and give prizes for the top three?
If the latter: Should we go with a specific theme or can we allow everything ASM related with a broad theme (e.g. sprites)?
Do you think we should have different types of ASM contests such as ASMLDC or 24hoASM/72hoASM alongside Mad Scientist?
(To the participants) How much fun did you have?
Is there anything else you want to add?
Something things to note:
I'm aware that 10 days (7 days in the first round) aren't much. However, the event calendar was rather tight so I was forced to use smaller rounds. I would have used two weeks of creating and one week of voting if we had time.
I may not be the best host and I admit I could have handled it better. However, I was the only one of the staff team who had shown interest to host it, not to mention this is the first time I hosted a contest (save for an inofficial contest year ago which was cancled since those contests were dying).
Since four years. As such, it's a good idea to have a feedback to improve on (possibly) later ASM contests.
On place three, we have got:
with a score of 1,163 points
On place two, the silver coder is
with a score of 1,956 points
And the grand winner of Mad Scientist is...
with a score of 3,983 points
The score may be low but remember that they're scaled logarithmically. I used the following formula: ln(n) / p where n are the total entries and p the entry's placement in that round. Using that, we'll get the following placements:
You will all receive a trophy for participating except the top three who receive a dedicated bronze, silver and gold trophy, respectively and can pick a list from our
You can also write a feedback here if you have any.
About judges: I went on with public judging because that's what the first Mad Scientist did (and in hindsight, it's also easier to host since you first have to think of the scoring cathegories and which judges to pick, though the staff team generally helps you out with it) but I also have considered to use dedicated judges next time (and if I don't host the next one, I surely hope the next host considers it).
There surely are advantages and disadvantages to both methods such as how everyone has got a different opinion on the entry but scoring is a bit caotic in public judging whereas with dedicated judges you can show off your resources.
However, something what will never happen: I don't see any point in mixing dedicated judges with public judging because you'd end up combining their disadvantages without using many of their advantages. Darolac brought up a good point why you should't bring a judge with public judging.
In addition, Mad Scientist is about creativity and doesn't care if the code is unoptimised. Aside from the time issue, that isn't much the point anyway of Mad Scientist. Same thing as with Idol and Super Famicompo: I'm very sure most submissions have got a bad MML since you can listen to the SPCs but don't get their source MMLs so there is no desire to waste energy for that.
This doesn't mean we can't host a contest which focuses on coding itself (such as code golf and obfuscated code) but Mad Scientist isn't one of them.
Keep in mind that the music section isn't the first section to remove packs as the ASM section removed multiple packs of more or less unrelated resources as well (e.g. romi's sprites, various SMB3, various SML and various YI sprites).
They have their advantages when you want to use a single soundtrack (the current Yoshi's Island soundtrack allows for that including global samples), of course, but the more I think about it, using a full soundtrack (including global songs) is something you have to know what you're doing.
Mainly through hacks like Brutal Mario (come on, everyone was inspired by it), the VIP series, , A Strange Mission, etc. and by various people like Ersanio, Roy, Yoshicookiezeus, Sonikku, etc. which inspired me to do SNES ASM (and programming in general).
I don't remember when I learned ASM, though, but I think it was with Ersanio's ASM tutorial where I made my first advancements + tutorials such as Iceguy's tutorial.
The second advancement was to read the SNES Developer Manual but I don't know when I read that. It must be either by learning how CC-DMA works (where the SA-1 readme redirected to) or by the second ASM Workshop (which contributed to my ASM skills a bit as well). That made me learn about stuff like about Mode 2 which made me realise how certain games managed certain effects (everyone talks about Yoshi's Island but the train stage of Tiny Toon Adventures makes use of Mode 2 as well).
Starting a couple years ago (when I started to used BSNES+ as my primary debugger), I was interested in how games handle certain effects which is why I investigated them with a debugger (and still do). The result: I learned how Yoshi's Island handles its colour gradient and parallax HDMA (which brought birth to "Scrollable" HDMA Gradients and Parallax HDMA Toolkit), I learned how SMB3 (SMAS) handles transparent water graphics (colour half-addition on layer 3 and sprites and layer 1 are both on mainscreen and subscreen) and how Super Metroid handles dark rooms (though I have speculated that it used colour subtraction especially since I noticed those rooms use two foreground layers years ago).
You have to also use the correct level mode i.e. 0E (layer 1 level) or 0F (layer 2 level) as layer 3 is otherwise only transparent to layer 2.
Sure, you still can use UberASM but it's really the same method just in different ways.
The issue about Piranha Plants and other sprites which go behind the foreground isn't easy to solve, though, as that's a byproduct of the transparency. There are Piranha Plants which work better against such effects but no other solution for other sprites exist.
The tutorial you linked refers to the UberASM patch. However, we use UberASM Tool which works different to the patch. Most importantly, you use a list.
To use the tool, simply extract UberASM (preferably in its own folder), put the ASM file into "levels" and open up the included list.txt. You'll see below "level:" that two codes will be inserted, one for level 105 and one for level 106. They're example codes so they can be freely removed but these also explain how to insert codes to levels with UberASM and I'm pretty sure you'll figure out on what to enter to list.txt.