Friday, Aug 12, 2011:
More engine progress notes:
"Parsed has been busy the last couple weeks integrating fixes we have received from the central rendering group. With the latest version of the Conan engine out and stable on the Conan live environment they have found a lot more time to work with us...which has been greatly appreciated. Work the last couple weeks has been on trying to sort out some minor differences in ground geometry between the old and new clients. The actual collision used is generated in the old version of the engine (all gameplay mechanics are still adjusted in the old systems) so if the ground isn't a good match it makes for floating buildings/people etc. Part of the problem was stemming from triangle diagonals in Conan ground being parallel while being alternating in old AO. (AO's system is better!) Shadows have also received a first go at optimization. Go Parsed, GO!"
Friday, August 19, 2011
And yet more engine progress notes:
This week Parsed managed to both get the tilemap textures to line up perfectly for the first time ever but also got the ground (heightmap/mesh) to be perfectly reproduced for the first time. Both of these were huge.
First off - the tile map:
There is texture blending in the Conan engine for all ground tiles making it impossible for us to do something you would consider commonplace...place a straight edge between two textures. As a sci-fi game we have many sidewalks, paths, metal plates etc on the ground that we don't want blending with their surroundings. In Age Of Conan and straight edge you see on the ground is in fact a mesh object that has been placed there. So very significant for maintaining the look and feel of AO to get that to work...and now it does.
Height Map/ Ground Shape:
This was incredibly important as the visual and physical aspects of the ground, while close, were not perfect. This would result in floating buildings, fences, and where the ground did not match previously generated collision floating players. There are some locations where you can see this in current AO where the collision of the mesh does not match the visual representations of it. Without this great work by Parsed this week this would have been a very significant problem going into the future. This was the source of by far the highest number of visual bugs in the new engine....and now it is gone. I'm very happy about this week's progress.
What about the stuff from this year? Edit: Nevermind
AO - JukeJoint "Nuhmudiira" Jezebel - 208/18 Trader - Omni
September 2nd, 2011
"New engine screenshots this week with the new resources made for Tir mostly in place. I have to keep shooing the rest of the team away from Parsed's desk...this is probably a good sign."
Note: the screenshots were apparently taken on the regular Testlive server, just as seen by the Devs thru the alpha new engine client.
September 30, 2011
More Tir screenshots, and a new headshot:
Last edited by Dagget; Oct 3rd, 2011 at 12:35:09.
Answers from FWM November 11th, 2011 Questions
Engine time estimate?
Full focus was put on creating a preview video in the middle of October and has been there since. Once the focus was put on producing something to be seen externally, with that kind of attention to detail that we know critical individuals will have, we began to see small but significant rendering issues we were not happy with. Fixing these issues has been far from trivial and has resulted in more than a few complete re-writes. In some cases we have run in to some limitations of the engine we were not particularly happy with and have been working with the engine team to resolve these issues. Many of these changes have also resulted in what would have been a complete rebuild of the entire client...which for the conversion processing of 400+ old AO playfields represents approximately 400+ hours of exporting/conversion. We all want a closed beta running as soon as possible but we are not really wanting to run that process before we are dead certain we will not have to run it again. It also goes without saying that we all want this out as soon as possible, but with years already invested in this undertaking we just aren't willing to deliver something we are not happy with. If there is a positive note in the delays associated with getting the code aspect of this change up and running it is that it has given the artists much more time to upgrade assets to be there for the day it goes live. Our goal is to have a closed Beta running this year. This is still an aggressive target at this point and it is hard to determine whether we will make it or not at this point as a great deal of this work is still "innovation" and "creative" in nature.
Can you give us a rough "roadmap" on the new engine. What has been done so far, and what is left to be done before a beta can be released.
The new engine is largely "feature complete" (most displays and runs as it should) with a few significant exceptions.
Incomplete Features Required for Beta:
- Completion of first run optimization.
- Proper "equipping" of the new heads with hair and new associated meshes.
- Equipping of new meshes associated with new attractors added to new character models. (ie: proper belt mesh to go with Belt-type items)
- Rebuilding of all in-game waterfalls and alteration of new engine technology to support them.
- Proper handling and display of transparent body texture types to allow for proper display of partial clothing. (ie: bump map of shoes to allow for correct display of bump map of feet.)
- Weather systems
- Integration of new effects system to be able to utilize new tools to produce new effects.
Areas that are still "buggy" that will likely have to be corrected before Beta:
- Shadow group size causing shadows to "flicker" from buildings.
- Grass (mesh and billboard) display distance being too close to character and changing colour incorrectly at certain times of day.
- Incorrect display of low level of detail version of certain meshes.
- Horizon display needs improvement.
Are you going to relaunch ao when the rebalance/gfx engi is done?
There is a plan to do some significant marketing once the new engine is up and stable.
Can i test the engine ?
When the closed external beta starts we'll have some kind of application process.
You mentioned long ago that all the trees in the game would need to be replanted, and you would not be using a program like Speedtree but manually placing them all. Where do things stand on this issue?
Replacing the trees has gone relatively smoothly with a few issues that pop up here and there. Replacement meshes have been built with similar height/size etc so that when the "final" export is finally done they should simply appear in the correct locations. I was a bit worried about this issue so it has been great to have this one work out quite well.
Since the new engine client will work alongside the old client, have you planned a future removal of the old client, much the way Eve Online ran for 18 months with duel clients before retiring the old one?
We'll be monitoring user stats after the launch and at some point will remove the old client. There will of course be lots of notice given as to when this might happen. If we see significant numbers of people continuing to use the old client going into the future it might remain quite viable to continue to run both...but we'll have to wait and see.
February 1st, 2012
3 Exclusive screenshots up at AO-Universe
City of Home
What utterly disappointing graphics to look forward to, I hope the 20 people still playing in 2015 will enjoy that patch.
Personally, I think Wartorn looks quite pretty.
But I'm definitely still firmly on the side of the naysayers
Colawren - Gimpest stripped adventurer on RK1.
The alts: Colawrenn, Iriotz, Empea
Proud member of Forsaken and Unity of the Rose.
This graphics engine, will look outdated before it's even released, sadly enough. I love the AO universe and story more than any other mmo, but even I gave up and moved on.
Arcticangel 220 adv
Arctica 220 soldier
Labbetuss 220 engi
Glacier 220 crat
Mullophys 213 MP
General of Ragnarok
March 27, 2012
Short video released demonstrating some of the new effects: