Atlas in game Editor?
by Dark Tengu · in Torque Game Engine Advanced · 09/24/2006 (10:54 am) · 59 replies
I thought I remember seeing that TSE was going to include in game terrain editing. When I look at the TSE development page I can't see it listed anymore. Has this been dropped? I so home not. :(
#42
Atlas is awesome, and i know how hard it would be to make that fully editable in game. But What i think everyone would probably be happy with, is if you guys were able to implement an enhanced legacy terrain. EI higher res textures, or even better mutiple detail textures. I'm aware how hard that might be. but i would figure the people the coded it originally would be able to figure something out.
09/27/2006 (8:13 pm)
@Stephen ZeppAtlas is awesome, and i know how hard it would be to make that fully editable in game. But What i think everyone would probably be happy with, is if you guys were able to implement an enhanced legacy terrain. EI higher res textures, or even better mutiple detail textures. I'm aware how hard that might be. but i would figure the people the coded it originally would be able to figure something out.
#43
Stephan, that is what you said that got people confused. I took it as: "Making a good in game editor would take a lot of time, so we dropped the idea and you will have to export the terrain to edit it". As you have since clarified, you will indeed be adding terrain editing tools. Just thought you should know where some of the confusion came in, at least on my end.
09/27/2006 (8:37 pm)
Quote:IMO, much better than any in game editor we could write without a year+ of additional development.
Stephan, that is what you said that got people confused. I took it as: "Making a good in game editor would take a lot of time, so we dropped the idea and you will have to export the terrain to edit it". As you have since clarified, you will indeed be adding terrain editing tools. Just thought you should know where some of the confusion came in, at least on my end.
#44
09/27/2006 (10:54 pm)
Quote: (of myself)
There are still plans on having some form of ingame adjusting of Atlas terrains--I didn't mean to give that impression. Providing a full free poly editor system in game isn't in scope of TSE 1.0 however, so keep in mind that detailed editing (creation, etc) if needed would probably go through the exporter to 3rd party app chain.
#45
Just wanted to weigh in here for a second.
It's pretty cool to see my reasoning on Atlas architecture & tools being basically recapitulated here. Makes me feel happy that I've chosen a direction that will work well for people.
Right now the Atlas runtime is pretty solid. But the toolset is minimal, which I totally agree sucks and very sharply limits the usefulness of the tech.
The longterm goals are to have:
Easy external editing w/ semi-realtime preview. Export the raw geometry to a common format that lots of things can load (.3ds is the first one - even milkshape and blender can deal with it happily), let them edit it, and load it back into the engine w/o a restart. This is closer than the others. Something similar for texture data, as well, so you can edit in photoshop or gimp or what have you. There's no reason for Torque to try to compete in the image editing department. :P
Useful but limited internal editing. I don't think there's any win by making TSE into 3ds max. Even integrating Constructor or something would be a TON of work. But there should be a very immediate, easy way to tweak things. At first this will be things like (and I'm going off my design notes here, not from your guy's excellent suggestions - just so you know that I'm on the same page w/ you :) "subdivide face", "remove face", "nudge vert." Eventually it should be a way to paint deformations onto the atlas geometry. Similar, possible to paint materials for the blended terrain texturing schema, too.
Small, tight, file formats and blazing performance. The blended texturing stuff has brought a big atlas terrain into the 10mb range - with aggressive optimization of the file format it could go into the <5mb range (or much less if you only need terrain quality comparable to the legacy terrain).
The main issue is prioritizing this work vs. finishing the rest of TSE vs. XNA vs. community and doc work vs. future tech R&D vs. TGB vs. Constructor vs. TDN vs. everything else we do.
I've also been listening to the artists using Atlas (and if I haven't heard you, please get in contact with me!), including talking to the MaxGaming guys (hi Adrian!) and talking to others. So you guys are definitely being heard as regards Atlas design.
09/28/2006 (7:49 pm)
Hey guys,Just wanted to weigh in here for a second.
It's pretty cool to see my reasoning on Atlas architecture & tools being basically recapitulated here. Makes me feel happy that I've chosen a direction that will work well for people.
Right now the Atlas runtime is pretty solid. But the toolset is minimal, which I totally agree sucks and very sharply limits the usefulness of the tech.
The longterm goals are to have:
Easy external editing w/ semi-realtime preview. Export the raw geometry to a common format that lots of things can load (.3ds is the first one - even milkshape and blender can deal with it happily), let them edit it, and load it back into the engine w/o a restart. This is closer than the others. Something similar for texture data, as well, so you can edit in photoshop or gimp or what have you. There's no reason for Torque to try to compete in the image editing department. :P
Useful but limited internal editing. I don't think there's any win by making TSE into 3ds max. Even integrating Constructor or something would be a TON of work. But there should be a very immediate, easy way to tweak things. At first this will be things like (and I'm going off my design notes here, not from your guy's excellent suggestions - just so you know that I'm on the same page w/ you :) "subdivide face", "remove face", "nudge vert." Eventually it should be a way to paint deformations onto the atlas geometry. Similar, possible to paint materials for the blended terrain texturing schema, too.
Small, tight, file formats and blazing performance. The blended texturing stuff has brought a big atlas terrain into the 10mb range - with aggressive optimization of the file format it could go into the <5mb range (or much less if you only need terrain quality comparable to the legacy terrain).
The main issue is prioritizing this work vs. finishing the rest of TSE vs. XNA vs. community and doc work vs. future tech R&D vs. TGB vs. Constructor vs. TDN vs. everything else we do.
I've also been listening to the artists using Atlas (and if I haven't heard you, please get in contact with me!), including talking to the MaxGaming guys (hi Adrian!) and talking to others. So you guys are definitely being heard as regards Atlas design.
#46
I think TSE (TAT) is a great engine. I get pretty good performance on relatively low spec computers. However, I'm very worried about how long I will have to wait for a real asset/art/tools pipeline. GG has so many projects going. I don't want to shell out more money to go with some other tech. But, time is money. It might be better for me to purchase $300 for a weekend warrior license from Lawmaker and upgrade to commercial later when I am ready.
Is there any kind of time frame on the tools side?
10/04/2006 (2:22 pm)
I guess my worries are summed up here by quoting Ben:Quote:The main issue is prioritizing this work vs. finishing the rest of TSE vs. XNA vs. community and doc work vs. future tech R&D vs. TGB vs. Constructor vs. TDN vs. everything else we do.
I think TSE (TAT) is a great engine. I get pretty good performance on relatively low spec computers. However, I'm very worried about how long I will have to wait for a real asset/art/tools pipeline. GG has so many projects going. I don't want to shell out more money to go with some other tech. But, time is money. It might be better for me to purchase $300 for a weekend warrior license from Lawmaker and upgrade to commercial later when I am ready.
Is there any kind of time frame on the tools side?
#47
10/07/2006 (1:03 pm)
Does anyone know if we will atleast be able to edit terrain textures in-game in TSE? It will be a huge pain not to be able to paint paths and other stuff. 
#48
10/07/2006 (1:27 pm)
...Quote:
Similar, possible to paint materials for the blended terrain texturing schema, too.
#50
And then Ben wrote:
(I understand that Stephen was writing about a solution integrated with Torque, but the tone seemed to imply disapproval of efforts toward making art modeling tools, in general.) We've got artists (generally unhappy with current DIF modeling solutions) to whom we have explained that Constructor is almost ready. The feeling I get from the above quotes is that Constructor is on the "to do" list, but isn't necessarily a priority and could very well be cancelled due to it being an art modeling tool (and thus not a core compentency of Garage Games).
Could someone from GG please state authoritatively that Constructor either will or will not be released? If it definitely will be released, can someone offer any sort of timeframe for that release? (If additional assistance testing would help, I'm sure we could commit resources to that.)
Tom
10/07/2006 (3:28 pm)
A bit off-topic from Atlas, but a couple comments in this thread have me a bit concerned. Stephen wrote:Quote:
Hopefully once you stopped drooling (I admit is does sound like an awesome idea initially) you'd realize that it would be a completely wrong decision--we don't make art modelling tools, we make game technology.
And then Ben wrote:
Quote:
The main issue is prioritizing this work vs. finishing the rest of TSE vs. XNA vs. community and doc work vs. future tech R&D vs. TGB vs. Constructor vs. TDN vs. everything else we do.
(I understand that Stephen was writing about a solution integrated with Torque, but the tone seemed to imply disapproval of efforts toward making art modeling tools, in general.) We've got artists (generally unhappy with current DIF modeling solutions) to whom we have explained that Constructor is almost ready. The feeling I get from the above quotes is that Constructor is on the "to do" list, but isn't necessarily a priority and could very well be cancelled due to it being an art modeling tool (and thus not a core compentency of Garage Games).
Could someone from GG please state authoritatively that Constructor either will or will not be released? If it definitely will be released, can someone offer any sort of timeframe for that release? (If additional assistance testing would help, I'm sure we could commit resources to that.)
Tom
#51
10/07/2006 (3:36 pm)
I think you are looking into those comments too much, that's a big twist.
#52
Joking aside, If you actually LOOK said list you quoted, most stuff on there is obviously high priority(if we ignore constructor and xna since we dont know about it, your still looking at TSE and TGB, as well as TDN and what not) and the fact constructor is included within is actually a good indication that they're...well...working on it.
I really cant recall any time GG has cancelled something they where working on, especially anything like Constructor, so i really dont see where your acute paranoia is coming from.
In regards to where atlas' pipeline is going, this is good to hear. I really never liked messing with TSE for TTR too much just yet because atlas seemed like an overly painful system to get running. Good to know HOW it's being fixed, as opposed to 'if' :p
10/07/2006 (8:16 pm)
Never Thomas. Clearly GG is out to steal all your shineys. They've already gotten into your bank account and are stealing your credit cards too D:!Joking aside, If you actually LOOK said list you quoted, most stuff on there is obviously high priority(if we ignore constructor and xna since we dont know about it, your still looking at TSE and TGB, as well as TDN and what not) and the fact constructor is included within is actually a good indication that they're...well...working on it.
I really cant recall any time GG has cancelled something they where working on, especially anything like Constructor, so i really dont see where your acute paranoia is coming from.
In regards to where atlas' pipeline is going, this is good to hear. I really never liked messing with TSE for TTR too much just yet because atlas seemed like an overly painful system to get running. Good to know HOW it's being fixed, as opposed to 'if' :p
#53
10/07/2006 (9:41 pm)
My concern is that they are spreading themselves too thin.
#54
Sorry, I didn't mean to sound paranoid; this is a business concern. I just need to help manage the expectations of our art team.
Tom
10/08/2006 (11:52 am)
@JeffSorry, I didn't mean to sound paranoid; this is a business concern. I just need to help manage the expectations of our art team.
Tom
#55
In regards to Constructor's release, Tom stated that it's really far along, and his personal guess on release is 3-4 months out at best.
10/08/2006 (12:04 pm)
I understand, figured i'd toss alittle poking fun in to get the point across. Good to see you caught it as a joke :)In regards to Constructor's release, Tom stated that it's really far along, and his personal guess on release is 3-4 months out at best.
#56
10/08/2006 (12:34 pm)
Is there a format specification anywhere for atlas?
#57
No; right now the format is what is written out. Are you trying to do an exporter/importer/parser? It would probably be best to use the core Atlas SDK, as it's designed to make it easy to stuff data into/get data out of the format in a threadsafe way.
10/09/2006 (5:12 pm)
Yes, we're still doing constructor. No, we don't comment on release dates (although sometimes we slip and give something out ;).No; right now the format is what is written out. Are you trying to do an exporter/importer/parser? It would probably be best to use the core Atlas SDK, as it's designed to make it easy to stuff data into/get data out of the format in a threadsafe way.
#59
I've always disliked the "we'll get it done when we get it done" philosophy. Sure it gives the impression of "we dont want to release something until it works well for you guys", but it turns into "our deadlines keep slipping and though we know you guys want something, we're going to do it when it's comfortable to us."
/rant off
That being said, its good to hear that GG is planning on Useful but limited internal editing. The points made above about needing to make changes to fit buildings et cetera are right on. There is a huge potential to drop the ball here by making Torque less user/newbie-friendly. And we're all watching.
11/06/2006 (4:45 am)
I agree that they sound too spread out at least from what I can gather from the forums. This is the least accurate way to gain information on such a subject but on the other hand the impression you give your customers should be noted.I've always disliked the "we'll get it done when we get it done" philosophy. Sure it gives the impression of "we dont want to release something until it works well for you guys", but it turns into "our deadlines keep slipping and though we know you guys want something, we're going to do it when it's comfortable to us."
/rant off
That being said, its good to hear that GG is planning on Useful but limited internal editing. The points made above about needing to make changes to fit buildings et cetera are right on. There is a huge potential to drop the ball here by making Torque less user/newbie-friendly. And we're all watching.
Torque Owner Hewster
Hewsoft
Making holes in Atlas terrain like this would be a BIG plus IMHO.