Welcome to Gaia! ::


It's an incredible undertaking, but it'd be an incredible feature. You have my vote!

Bashful Millionaire

7,250 Points
  • Mark Twain 100
  • Tipsy 100
  • Tycoon 200

9,050 Points
  • V-Day 2011 Event 100
  • Bunny Hunter 100
  • Friendly 100

Sparkly Lover

9,650 Points
  • Gaian 50
  • Invisibility 100
  • Friendly 100
YES. So the plain gloves can go UNDER the shirt sleeves!

Desirable Cutesmasher

heart emotion_bigheart heart

Friendly Explorer

9,000 Points
  • Unleash the Beast 100
  • Peoplewatcher 100
  • Statustician 100
All right. I have been thinking about this, programming-wise, for a bit, and here's the solution I came up with. Now, I do not know the full complexity of the current setup, but here's a way it could work:

I've seen a mockup in this thread for a new box on the right of the dress-up screen that lists the items equipped so that you can select their layer order. This can make the dress-up screen the source for the completed image, instead of having it re-rendered at other points in its use. Just have the dress-up screen capture the avi as designed, send that image to the cache servers, and have other processes go to those servers for the images. You could also have a flag of whether the avi has an animated image set, and when the flag is set, the dress-up app would render the additional frames of animation at the same time as the still, and other calling functions would read the flag and load the image as a single or multiple image as appropriate. "I Am" items already override regular items, so I see no reason why that would have to change. Just make the other items grayed out in the right hand z-level box for a visual indicator. Also, the "walk" animation in towns and zOMG is just a blurred leg swish, so that can be layered on top of anything. Finally, the back version of the avi can be set to the opposite of the front to begin with, but can be separately set as well. So, the final image set would be: front-right-facing, front-left-facing, back (flippable), f-r-kneeling, f-l-kneeling, back-kneeling, and when necessary, f-r-animation, f-l-animation, back-animation, f-r-kneel-animation, f-l-kneel-animation, back-kneel-animation, so either 6 or 12 images depending on animated item usage, and if you want a part of the site to not worry about the animation, just have it load the first 6 images.

As for layering within clothes, I'm sure there are some items that have more than one layer to accommodate backs to jackets that may be visible, and other such. What you could do is simply make the z-level of the item back layer a relative -1. You could further limit the regular item layers z-level, within the Dress-up interface, to a minimum of 1 or 2 to make sure that the relative -1 wouldn't break the code (and it doesn't need to be anything visible to the players, either).

So, in conclusion, make the dress-up program a little heavier, and you could actually lighten a bunch of the other processes, probably save on some server resource usage, and it would standardize the avi's look across the site as well!
Doram


I appreciate the thought you've put into this. biggrin Hopefully the suggestions will help persuade the devs not to get scared off the task.
User Image - Blocked by "Display Image" Settings. Click to show.

Questionable Cat

Doram
So, in conclusion, make the dress-up program a little heavier, and you could actually lighten a bunch of the other processes, probably save on some server resource usage, and it would standardize the avi's look across the site as well!
User Image


Awesome post! I hope the devs see this :3

Invisible Raider

Quote:
Note from Panagrammic: Moving this up to the main forum (and clearing votes for it), but you'll have to realize that this is such a huge implementation effort that it is very unlikely we would do it. The avatar composition code is very complex, and we have an entire suite of tools built around the way that it works now; changing that would require us to a) validate almost every item we have again, b) totally rewrite the avatar composition (which includes such things as filmstrip building, which are used in the flash games), c) rework the way we do animated items, d) change the way that we persist avatar data, e) come up with an interface for controlling the layering, f) figure out migration for old avatars, so on and so forth. Users are unhappy with the number of "green goo" problems we have now; can you imagine the number of problems that would be raised if layering was suddenly completely up to the user?


Got my new cookie in.

I'm sure that if the guys from TekTek were able to do it, you should be able do get it done as well.

Heck, you could hire those guys or pay them to license their code!!

Sparkling Treasure

13,950 Points
  • Alchemy Level 3 100
  • Fluff Rehab 200
  • Wall Street 200
I'm all for this. Although if this is too difficult for the site makers, I can understand that. What if it was simplified for to the idea of being able to wear more items at the same time. Like being able to wear two tops and they will layer together. Some items will allow this, and some won't now. Maybe this would be easier? Maybe? whee

Aged Gaian

Doram
All right. I have been thinking about this, programming-wise, for a bit, and here's the solution I came up with. Now, I do not know the full complexity of the current setup, but here's a way it could work:

I've seen a mockup in this thread for a new box on the right of the dress-up screen that lists the items equipped so that you can select their layer order. This can make the dress-up screen the source for the completed image, instead of having it re-rendered at other points in its use. Just have the dress-up screen capture the avi as designed, send that image to the cache servers, and have other processes go to those servers for the images. You could also have a flag of whether the avi has an animated image set, and when the flag is set, the dress-up app would render the additional frames of animation at the same time as the still, and other calling functions would read the flag and load the image as a single or multiple image as appropriate. "I Am" items already override regular items, so I see no reason why that would have to change. Just make the other items grayed out in the right hand z-level box for a visual indicator. Also, the "walk" animation in towns and zOMG is just a blurred leg swish, so that can be layered on top of anything. Finally, the back version of the avi can be set to the opposite of the front to begin with, but can be separately set as well. So, the final image set would be: front-right-facing, front-left-facing, back (flippable), f-r-kneeling, f-l-kneeling, back-kneeling, and when necessary, f-r-animation, f-l-animation, back-animation, f-r-kneel-animation, f-l-kneel-animation, back-kneel-animation, so either 6 or 12 images depending on animated item usage, and if you want a part of the site to not worry about the animation, just have it load the first 6 images.

As for layering within clothes, I'm sure there are some items that have more than one layer to accommodate backs to jackets that may be visible, and other such. What you could do is simply make the z-level of the item back layer a relative -1. You could further limit the regular item layers z-level, within the Dress-up interface, to a minimum of 1 or 2 to make sure that the relative -1 wouldn't break the code (and it doesn't need to be anything visible to the players, either).

So, in conclusion, make the dress-up program a little heavier, and you could actually lighten a bunch of the other processes, probably save on some server resource usage, and it would standardize the avi's look across the site as well!


This deserves to be on the front page for Panagramic to see!!!!!

OG Smoker

Doram

User Image







What he said........ :3




Mega Hoarder

Quick Reply

Submit
Manage Your Items
Other Stuff
Get GCash
Offers
Get Items
More Items
Where Everyone Hangs Out
Other Community Areas
Virtual Spaces
Fun Stuff
Gaia's Games
Mini-Games
Play with GCash
Play with Platinum