Reverse bay

[Silverine]

Except for rendering, there is only the one case of a glitch horse popping up that I recall (at least that was noted), and that was my mare that displayed Lp pattern when she was not. The mare that I sold to you, Silverine, during the earlier stages of Lp extension through the neck. She was the only case I know of and though that was a bit before To, it does appear that after the To release, they skyrocketed.

Oh yeah, I remember her!

I wonder what it is about the To that messes with the rendering. Being a programmer, I'd love to have a look at it.

[Baranduin Brewster]

Here is a Glitched Palomino Turk

Horses Name

Wow, that is a glitch I'd like to have for real in-game; I love the pure white mane with the solid/base black body <3 does she per chance have cream and double black/aa as base and just can't decide on the red/black factor?

Possible, I'm not real sure as I haven't bred him to any chestnuts

[BlackOak2]

Wow, that is a glitch I'd like to have for real in-game; I love the pure white mane with the solid/base black body <3 does she per chance have cream and double black/aa as base and just can't decide on the red/black factor?

Possible, I'm not real sure as I haven't bred him to any chestnuts

There's two possibilities in my book for this glitch. One is this horse carries silver and has glitched the dappling/gray on the body, however, turks aren't supposed to have silver (along with a cream gene).
Or we're seeing a hidden cream and flaxen (again flaxen shouldn't be in turks, but who knows?), which is my first opinion (flaxen isn't my first opinion though).
What I personally am sure of, this horse is a single cream carrier and has an E/e base. But I've been wrong before!

[Nazarach]

Possible, I'm not real sure as I haven't bred him to any chestnuts

There's two possibilities in my book for this glitch. One is this horse carries silver and has glitched the dappling/gray on the body, however, turks aren't supposed to have silver (along with a cream gene).
Or we're seeing a hidden cream and flaxen (again flaxen shouldn't be in turks, but who knows?), which is my first opinion (flaxen isn't my first opinion though).
What I personally am sure of, this horse is a single cream carrier and has an E/e base. But I've been wrong before!

I've though, (but am not sure as I have only very limited experience with programming), there is a finite number of gene combinations, right? Is it possible that somewhere in the code they forgot to extend this limit with +1/+2 (as in To/to spaces) and that surpressed the first entie (the E/e factor) and because of this there are these problems with black/chestnut bases?

[Baranduin Brewster]

Possible, I'm not real sure as I haven't bred him to any chestnuts

There's two possibilities in my book for this glitch. One is this horse carries silver and has glitched the dappling/gray on the body, however, turks aren't supposed to have silver (along with a cream gene).
Or we're seeing a hidden cream and flaxen (again flaxen shouldn't be in turks, but who knows?), which is my first opinion (flaxen isn't my first opinion though).
What I personally am sure of, this horse is a single cream carrier and has an E/e base. But I've been wrong before!

He is definitely a palomino that I do for certain, so single cream is correct, his gallery does show that. I think there is some question about the Flaxen in Turks at the moment. I will eventually breed him to a few more mares, once i get this round of Quarters done.

[AkumaKashi Holding]

So I picked this guy up in the adoption center during the tutorial. He's some kind of a funky sucker. I just made this account as an extra to age broodmares rapidly. I got this guy because I thought he was really strange looking and I wanted a black Belgian with white on him. He has 2 red stats and a gold, but I'm wondering if it would be worth it to breed to a couple of high gp mares to try and put that color on more valuable horses. The best way I can describe his color is as a reverse bay. I have seen nothing like him in game or real life.
Sunburst Styles

Interesting! I have a similar Belgian on my holding acc.

Awesome! May as well be the same horse.

[AkumaKashi Holding]

So I picked this guy up in the adoption center during the tutorial. He's some kind of a funky sucker. I just made this account as an extra to age broodmares rapidly. I got this guy because I thought he was really strange looking and I wanted a black Belgian with white on him. He has 2 red stats and a gold, but I'm wondering if it would be worth it to breed to a couple of high gp mares to try and put that color on more valuable horses. The best way I can describe his color is as a reverse bay. I have seen nothing like him in game or real life.
Sunburst Styles

Here is a Glitched Palomino Turk

Horses Name

That thing is really beautiful. I'd like to have a real horse that looks like that. I have the Palomino part and a light gray with a black name and tail, some I'm sort of there.

[BlackOak2]

I've though, (but am not sure as I have only very limited experience with programming), there is a finite number of gene combinations, right? Is it possible that somewhere in the code they forgot to extend this limit with +1/+2 (as in To/to spaces) and that surpressed the first entie (the E/e factor) and because of this there are these problems with black/chestnut bases?

I would assume there is a finite number of gene combinations. After all, we only have so many genes, so that does limit how many combinations we can have. I want to say it's strictly a rendering issue, but that denotes that each gene is rendered in layers on top of one another and it's only the very top layer that we see, which doesn't make too much sense, since some of those layers must be semi-clear to show through the other layers. Therefore certain color glitches shouldn't be able to happen (though I can't come up with one specifically at the moment).
So if it's not a rendering issue, it must fall back to some sort of coding error that either is being overlooked by the program, or something else. After all, technology is far from perfect.

But that is as far as my knowledge and understanding extends.

[Silverine]

So this post is probably going to be long. I apologize in advance.

There's three main ways that HWO could possibly store horses' genetic information. (There's actually vastly more ways than that, but the others make things way more complicated than these three.) The first is what you have suggested - having a pre-programmed list of combinations and each horse 'points' to whichever combination they are. The second is for every horse to have its own record of what genes it carries, much like horses in real life having their own chromosomes. The third is a mix of these two. Each method has pros and cons.

First Method
-Pros: it saves space (but only if you have more horses than you have gene combinations), quick lookups for entire genome.
-Cons: there are a ridiculous number of combinations that must be pre-programmed, combining horses becomes a lot more complicated.

Second Method
-Pros: much easier combination of horses, quick lookups for individual genes, programmer doesn't have to pre-load millions (potentially billions or more*) of gene combinations.
-Cons: can potentially take up ridiculous amounts of storage space

Third Method
-Pros: saves space, quick lookups for full genome and specific genes (though not as quick for either as the above methods are for their specialty), easier combination of horses
-Cons: still have to pre-program a ridiculous amount of combinations, database queries can become super complicated

*After some quick calculations that only included the genes involved with grey, extension, agouti, champagne, cream/pearl, flaxen, pangare, silver, tiger eye, sooty, and the lp/to switch genes (the 'on/off' ones, not PATN or white-coding genes) I got 50,331,648 potential combinations. That more than 30 times the number of active horses on HWO right now, and that's not even including all of the genes!

(Also, as a side note, I realized something. Does anyone know how HWO handles dun? Last I heard I thought it was a simple D/d handle, but I realized, especially with NAB 'barely-there' dun that is entirely possible HWO uses the recently confirmed D/nd1/nd2 model. D is regular dun, nd1 is dominant over nd2 and codes for very faint dun markings ('barely-there' dun), and nd2 is no dun at all.)


Considering the huge number of genes and the potential millions of combinations that HWO uses, I find it very unlikely that it uses either the first or third method. While the second has potential to take up vast amounts of storage space HWO does take an effort to reduce the amount of data it must store - limits on how many living horses a player can have at one time; rehomed horses are removed completely so there is nothing to store for them; and (I will admit this is an assumption) deceased/retired horses also dump their genome (they have no need of it since they can't breed and only store their gallery photo rather than rendered images). The ability for players to purchase extra horse spots also helps take care of the problem - the few dollars that players put into those extra spots can pay for the needed storage space. I can also guarantee you that we do not have more horses than genetic combinations (see above).

According to one of the Patreon posts, HWO displays horses' images by rendering them at five different ages then doing some combining, layering, and clipping to show them at whatever age they currently are. (This is a simplification. I've never done image rendering so I'm not able to translate what was said very well.)

So far I have only encountered this rendering issue, or seen others encounter it, with AC horses. (Please let me know if you have seen it elsewhere, I would love to know! ) My guess is that there is an error in the way the AC generates horses. Whether that issue is with the horse's genome or with the image renderer itself I can not say. If it is with the genome, my guess is that something about the generation of the To genes (most likely the white area coding rather than the switch itself, though it's hard to say for sure) causes an error in other genes. This could as simple as 'pushing' them one space to the left or right. For example, genes that should be stored as "ggAaEe" get stored as "gAaEeg". This would lead to corrupted data being sent to the renderer, affecting the horse's image.


So yeah. Sorry for the long post. I'm just super curious as to what is going on behind the scenes.

[BlackOak2]

...

Very insightful (at least for me), and I understand it well.
As far as dun, I haven't found an actual dun release topic. As far as I know, it's only a straight dun, no dun and the actual appearance is handled the same way all of our other colors are, some are darker, some are lighter before being affected by sooty.
However, since we're all aware of how our admins have been programming the colors and genes for the others so far, I suppose it depends on when the recent discovery for the dun gene was published. If it was published after admin introduced dun, then I don't think it was changed after being added, however... if the game was introduced to dun after the dun discovery was published... then I suppose I will have to update my personal notes!
I haven't really noticed a difference between 'types' of duns... If it's possible, I'll have to do some of my own research to see what the publication says.

I wonder if admin is willing to comment on which dun coding they may have used?

I'm curious to the behind the scenes myself... but I'm sure I'll get lost pretty quick! Hah!