'Tindalos' (Ogre v2.0) SceneManager redesign
- sinbad
- OGRE Retired Team Member
- Posts: 19269
- Joined: Sun Oct 06, 2002 11:19 pm
- Location: Guernsey, Channel Islands
- x 66
- Contact:
Thing is, we can already vary the threading behaviour - it's just done via the preprocessor instead of templates. Both require the same sort of compile-time change to alter so I'm not seeing the difference. Except for people who want to re-use our internal classes for their own structures, but that's really not what we're here for.
It's less to do with the time and more to do with the fact that it clutters our code more than it needs to be.
It's less to do with the time and more to do with the fact that it clutters our code more than it needs to be.
- raicuandi
- Gargoyle
- Posts: 1092
- Joined: Wed Nov 09, 2005 12:56 pm
- Location: Adelaide, Australia
- Contact:
Sorry to go a bit off-topic... but:
I've noticed that there is some kind of resource management in Ogre where SharedPtr<Resource> is used, and the ResourceManager class always keeps the 'last' pointer. So if it would want to know which ones it could unload/whatever, it would have to check a list of its child resources, and see if the use count is 1, then act on it.
For a project of mine, I've done my own 'shared pointer' class, with the modification that it has a Listener class, with 2 methods called when:
1) there is only one last 'use' (usecount == 1)
2) the last 'use' has been removed (usecount == 0)
I figured that this will help me get rid of going trough that list like above, and checking each one, which is, if I'm not wrong, O(n) in performance...
I've noticed that there is some kind of resource management in Ogre where SharedPtr<Resource> is used, and the ResourceManager class always keeps the 'last' pointer. So if it would want to know which ones it could unload/whatever, it would have to check a list of its child resources, and see if the use count is 1, then act on it.
For a project of mine, I've done my own 'shared pointer' class, with the modification that it has a Listener class, with 2 methods called when:
1) there is only one last 'use' (usecount == 1)
2) the last 'use' has been removed (usecount == 0)
I figured that this will help me get rid of going trough that list like above, and checking each one, which is, if I'm not wrong, O(n) in performance...
- Falagard
- OGRE Retired Moderator
- Posts: 2060
- Joined: Thu Feb 26, 2004 12:11 am
- Location: Toronto, Canada
- x 3
- Contact:
- KungFooMasta
- OGRE Contributor
- Posts: 2087
- Joined: Thu Mar 03, 2005 7:11 am
- Location: WA, USA
- x 16
- Contact:
- raicuandi
- Gargoyle
- Posts: 1092
- Joined: Wed Nov 09, 2005 12:56 pm
- Location: Adelaide, Australia
- Contact:
I'm sorry, I just don't get it: so there's an unique and irreplaceable (?) WorldManager, and any amount of implementations or instances of SceneManager and other goodies (Locations?) right?
You said that that scene manager enumerator will be out, so how will one create a registered scene manager now? By its name? (/edit: gonna have some kind of factory?)
You said that that scene manager enumerator will be out, so how will one create a registered scene manager now? By its name? (/edit: gonna have some kind of factory?)
- sinbad
- OGRE Retired Team Member
- Posts: 19269
- Joined: Sun Oct 06, 2002 11:19 pm
- Location: Guernsey, Channel Islands
- x 66
- Contact:
I'll post more details when I have them.
I did say not to hold your breath, this was just planning & vision work. v1.6 / summer of code comes first, and right now I'm diverting a lot of effort to another project of my own I've had on the back burner for ages. v2.0 will be worked on as and when I have the time. Reality has set in and I can't work on everything at once
I did say not to hold your breath, this was just planning & vision work. v1.6 / summer of code comes first, and right now I'm diverting a lot of effort to another project of my own I've had on the back burner for ages. v2.0 will be worked on as and when I have the time. Reality has set in and I can't work on everything at once
- SunSailor
- Gnoll
- Posts: 699
- Joined: Sun Jan 02, 2005 5:45 pm
- Location: Velbert, Germany
- x 2
- Contact:
- KungFooMasta
- OGRE Contributor
- Posts: 2087
- Joined: Thu Mar 03, 2005 7:11 am
- Location: WA, USA
- x 16
- Contact:
- betajaen
- OGRE Moderator
- Posts: 3447
- Joined: Mon Jul 18, 2005 4:15 pm
- Location: Wales, UK
- x 58
- Contact:
- aeyr
- Halfling
- Posts: 67
- Joined: Thu Sep 21, 2006 9:51 pm
- Location: Seattle, WA (USA)
Reading over your world update document, Sinbad, it's not entirely clear where bounding boxes go in the new TransformSource setup. Does the transform source hold a BB now as well? Does it / can it hold the children's (a "cumulative" BB for lack of a better term)? How would a "cumulative" BB's be propagated/calculated up a hierarchy if the hierarchy is completely abstracted/external (or does this become the responsibility of a TransformSource implementation)? If a TS does hold the BB, how does it gather a change in extents from a MovableObject (e.g. ManualObject or unbounded/flexibly-bounded particle system)?
Just throwing it out there. The new architecture thoughts do look promising!
Just throwing it out there. The new architecture thoughts do look promising!
- sinbad
- OGRE Retired Team Member
- Posts: 19269
- Joined: Sun Oct 06, 2002 11:19 pm
- Location: Guernsey, Channel Islands
- x 66
- Contact:
Short answer is it won't. Forcing an AABB to be propagated upa transform hierarchy is too limiting, and in fact that premise doesn't exist at the moment anyway. Only the basic scene manager forms nested AABBs based on SceneNodes, others like Octree do not because it is wasteful in update time. AABBs will only be required at leaves.
- oddrose
- Orc
- Posts: 470
- Joined: Thu Feb 15, 2007 2:08 pm
- Location: Gothenburg, Sweden
- Contact:
It's a bit scary waking this monster of a thread, but I am so eager to here any news on this topic that I can't keep myself from it. More than a year has passed since the last post so I thought it was appropriate to discuss what has happened during this time. Even if no implementations have been made, I want to know if there are any new thoughts that have come up on the topic. I must say that I was and still am excited about the system that is sketched out in this thread so I hope any new plans don't entirely wreck the old ones.
Glad to hear any input or comments!
Fredrik
Glad to hear any input or comments!
Fredrik
- Praetor
- OGRE Retired Team Member
- Posts: 3335
- Joined: Tue Jun 21, 2005 8:26 pm
- Location: Rochester, New York, US
- x 3
- Contact:
Not much, except that there will definitely be a 1.8 before we get to 2.0. We still have a lot of work to do before we are ready to make this change. Work by other members in this forum have helped us understand a lot of the issues we'll have to face when more threading is introduced. So far, the work has focused on the resource system and that has been painful.
That being said, I am still excited as well and this will be a necessary change. I imagine we'll need a lot of input from scene manager writers and maintainers.
That being said, I am still excited as well and this will be a necessary change. I imagine we'll need a lot of input from scene manager writers and maintainers.
- sinbad
- OGRE Retired Team Member
- Posts: 19269
- Joined: Sun Oct 06, 2002 11:19 pm
- Location: Guernsey, Channel Islands
- x 66
- Contact:
-
- Gnoblar
- Posts: 9
- Joined: Mon Apr 13, 2009 3:19 am
Re: 'Tindalos' (Ogre v2.0) SceneManager redesign
Does anyone have more current information on the progress made on this? I've been checking back every now and then for about 18 months now.
-
- Kobold
- Posts: 26
- Joined: Mon Aug 02, 2010 5:36 pm
Re: 'Tindalos' (Ogre v2.0) SceneManager redesign
wysardry wrote:Does anyone have more current information on the progress made on this? I've been checking back every now and then for about 18 months now.
I'm hopeful that work can get started on this quite soon as well, but I think people are going to have to step into the breach. I'm happy to contribute where possible, but I'm not exactly sure how the development process in these parts works and as a relative newbie (albeit long-time lurker), I'm not sure how much my opinions would be appreciated! Perhaps one of the mods/devs can enlighten me on that one?
Fred
- Klaim
- Old One
- Posts: 2565
- Joined: Sun Sep 11, 2005 1:04 am
- Location: Paris, France
- x 56
- Contact:
Re: 'Tindalos' (Ogre v2.0) SceneManager redesign
You should try to contact the team (here or by mp) to have more insights about what is the current state, who is working on it, what's the current plan regarding this feature etc.
Or you could first give here your own analysis about what you think you could do, then wait for critics from the team before starting something.
Or you could first give here your own analysis about what you think you could do, then wait for critics from the team before starting something.
- sparkprime
- Ogre Magi
- Posts: 1137
- Joined: Mon May 07, 2007 3:43 am
- Location: Ossining, New York
- x 13
- Contact:
Re: 'Tindalos' (Ogre v2.0) SceneManager redesign
A big change like this needs to be broken down into several small changes, between which the whole thing should be operational and testable by all, with a cooling off period so that it can be stable before moving onto the next chunk of changes.
- jacmoe
- OGRE Retired Moderator
- Posts: 20570
- Joined: Thu Jan 22, 2004 10:13 am
- Location: Denmark
- x 179
- Contact:
Re: 'Tindalos' (Ogre v2.0) SceneManager redesign
A bit late, but here goes:Fritznorm wrote:I'm hopeful that work can get started on this quite soon as well, but I think people are going to have to step into the breach. I'm happy to contribute where possible, but I'm not exactly sure how the development process in these parts works and as a relative newbie (albeit long-time lurker), I'm not sure how much my opinions would be appreciated! Perhaps one of the mods/devs can enlighten me on that one?
I don't think any of us know exactly how the development process works, after the retirement of Sinbad as project lead, and as a developer.
If this project is going to survive - which I believe it will - we need to find new ways.
As you probably know, Ogre's source repository is hosted at BitBucket, and that opens up a whole lot of new opportunities.
Maybe it's time that we take the next step and start utilising the 'D' in 'DVCS' ?
For example, you clone the Ogre project.
Then you start working on your thing, periodically keeping up to date with changes in the default branch (by pulling and merging it into your clone/branch).
As progress occurs, discussion can take place in the forum and the code/direction can be reviewed.
I am not sure if Bitbucket has code review functionality..
When you are feeling confident, issue a merge request, and let some Ogre developers review it, and (hopefully) merge it into the main default branch.
I think that it would be good to issue merge requests several times during the life time of the clone/project.
Maybe you are even given push priviliges ?
I don't know - let's hear from the team of Ogre developers - but I believe we need to open up development more in order to move forward.
/* Less noise. More signal. */
Ogitor Scenebuilder - powered by Ogre, presented by Qt, fueled by Passion.
OgreAddons - the Ogre code suppository.
Ogitor Scenebuilder - powered by Ogre, presented by Qt, fueled by Passion.
OgreAddons - the Ogre code suppository.
- Eugene
- OGRE Team Member
- Posts: 185
- Joined: Mon Mar 24, 2008 4:54 pm
- Location: Kraków, Poland
- x 41
Re: 'Tindalos' (Ogre v2.0) SceneManager redesign
It just stopped. Usually bugfixes and new features were added to tracker, and two last months patches are not reviewed, nor accepted.jacmoe wrote: I don't think any of us know exactly how the development process works, after the retirement of Sinbad as project lead, and as a developer.
http://sourceforge.net/tracker/?func=br ... tid=302997
-
- Ogre Magi
- Posts: 1120
- Joined: Wed Nov 15, 2006 7:41 pm
- Location: Finland
- x 5
Re: 'Tindalos' (Ogre v2.0) SceneManager redesign
Well, I became concerned when I noticed that Sinbad retired without saying who is the new project manager.
I hope this situation gets resolved. This project can't continue without a manager.
Sinbad has done a super-good job with this project and asking him to do more is really hard. But perhaps he could do one more thing - to clear it up who, if anyone, is the new project manager. If there is no project manager at all in sight then it would be good to hear about plans of how to get one for this project. Maybe the community can help to find a new project manager.
Sorry but no I am not volunteering - I can't do such a job - I rather choose to do smaller jobs such as the small patch I've submitted (yes, that patch has gone without a response too).
But I really hope this situation of not having a manager make things move gets resolved. This project is very very good and it would be a big shame to see it stop making good progress.
I hope this situation gets resolved. This project can't continue without a manager.
Sinbad has done a super-good job with this project and asking him to do more is really hard. But perhaps he could do one more thing - to clear it up who, if anyone, is the new project manager. If there is no project manager at all in sight then it would be good to hear about plans of how to get one for this project. Maybe the community can help to find a new project manager.
Sorry but no I am not volunteering - I can't do such a job - I rather choose to do smaller jobs such as the small patch I've submitted (yes, that patch has gone without a response too).
But I really hope this situation of not having a manager make things move gets resolved. This project is very very good and it would be a big shame to see it stop making good progress.
-
- OGRE Retired Team Member
- Posts: 2903
- Joined: Thu Jan 18, 2007 2:48 pm
- x 58
- Contact:
Re: 'Tindalos' (Ogre v2.0) SceneManager redesign
The internal decision was not to appoint a new project lead, and I still think that can work.
However, you are right that we have neglected the trackers. So obviously we do need someone to regularly go over those and fix or assign to the appropriate people. I'll see what we can do, though if there are any community volunteers, do speak up
However, you are right that we have neglected the trackers. So obviously we do need someone to regularly go over those and fix or assign to the appropriate people. I'll see what we can do, though if there are any community volunteers, do speak up
- Praetor
- OGRE Retired Team Member
- Posts: 3335
- Joined: Tue Jun 21, 2005 8:26 pm
- Location: Rochester, New York, US
- x 3
- Contact:
Re: 'Tindalos' (Ogre v2.0) SceneManager redesign
I've been pretty absent lately because a client project has been monopolizing my time. That's temporary. I think some ebb and flow is to be expected. There's been some slow-down in development in some areas. But at the same time, look at the great work going into increasing Ogre's support for instancing.
I'm suspecting that in a lot of ways we may not get a grand 2.0 redesign. We never really knew what it was anyway. I don't think grand revolutions are the best strategy for open-source projects. We will march along with what appear to be small changes, with a few lurches ahead every now and then. Eventually we'll push on the DX11 render systems. We'll make it possible to parallelize sections of the engine. At some point we'll look back and see that we went pretty far from 1.7. But it will be a jog, not a sprint.
I'm suspecting that in a lot of ways we may not get a grand 2.0 redesign. We never really knew what it was anyway. I don't think grand revolutions are the best strategy for open-source projects. We will march along with what appear to be small changes, with a few lurches ahead every now and then. Eventually we'll push on the DX11 render systems. We'll make it possible to parallelize sections of the engine. At some point we'll look back and see that we went pretty far from 1.7. But it will be a jog, not a sprint.
Game Development, Engine Development, Porting
http://www.darkwindmedia.com
http://www.darkwindmedia.com
- sinbad
- OGRE Retired Team Member
- Posts: 19269
- Joined: Sun Oct 06, 2002 11:19 pm
- Location: Guernsey, Channel Islands
- x 66
- Contact:
Re: 'Tindalos' (Ogre v2.0) SceneManager redesign
Sorry I haven't been around much lately - I've embarked on a total gear change and have to concentrate my reduced PC time on a new commercial endeavour before I run out of money
My hope was that Mercurial would help distribute the load a bit more. It's far easier for people to collaborate among themselves, and for community members to check other user's work, rather than it just being up to someone at the centre. That's why as I began to realise my role was getting too much for me, I sought out a DVCS.
If you're concerned that patches aren't being reviewed, you can help by reviewing them yourself. Basically, if 3 people in the community have tested a patch and said it works for them (even better if it's on different platforms), it makes the process of central acceptance considerably easier. And if there are problems, helping resolve them by making a fork on BitBucket, applying the patch, then refining it, even giving others in the community chance to help so that the final version is acceptable is a smooth fashion, all help these things go quicker.
In the past I was this nucleus that everyone sent everything to in the first instance, but sadly I can't be that guy anymore. There's a big community here, and a little collaboration outside of the centre can go a long way. It wasn't really very practical with SVN, exchanging patch files was a pain, but the tools are there now for everyone to work on refining code modifications outside the core so that the re-integration process is slicker and not dependent on one person putting lots of hours in to check & verify things. Try it!
My hope was that Mercurial would help distribute the load a bit more. It's far easier for people to collaborate among themselves, and for community members to check other user's work, rather than it just being up to someone at the centre. That's why as I began to realise my role was getting too much for me, I sought out a DVCS.
If you're concerned that patches aren't being reviewed, you can help by reviewing them yourself. Basically, if 3 people in the community have tested a patch and said it works for them (even better if it's on different platforms), it makes the process of central acceptance considerably easier. And if there are problems, helping resolve them by making a fork on BitBucket, applying the patch, then refining it, even giving others in the community chance to help so that the final version is acceptable is a smooth fashion, all help these things go quicker.
In the past I was this nucleus that everyone sent everything to in the first instance, but sadly I can't be that guy anymore. There's a big community here, and a little collaboration outside of the centre can go a long way. It wasn't really very practical with SVN, exchanging patch files was a pain, but the tools are there now for everyone to work on refining code modifications outside the core so that the re-integration process is slicker and not dependent on one person putting lots of hours in to check & verify things. Try it!
- Wolfmanfx
- OGRE Team Member
- Posts: 1525
- Joined: Fri Feb 03, 2006 10:37 pm
- Location: Austria - Leoben
- x 99
- Contact:
Re: 'Tindalos' (Ogre v2.0) SceneManager redesign
Maybe some of us could volunteer to help out with patch testing. But there is still a need for a guy which assigns this guys patches to test (because its worthless when i test a patch and say he its working here but nobody is interested in merging it into to trunk). I would volunteer on win/mac for patch testing. I think there should be a patch master from the ogre core team which assign the test jobs (also the patch master should change every month).