State/feedback/rant -> Ogre 2.1

A place for Ogre users to discuss non-Ogre subjects with friends from the community.
User avatar
devxkh
Halfling
Posts: 84
Joined: Tue Aug 02, 2016 6:07 pm
Location: Germany
x 12

State/feedback/rant -> Ogre 2.1

Post by devxkh »

State/feedback/rant -> Ogre 2.1

Use Case/About me:
Hobby, Sparetime game programmer.

Pros:
- open source
- metal, directx, openl renderer
- pbs rendering
- skeletal animation (pbs only)

Cons (as someone said: an opinion is like an asshole, everyone has one):
- missing a lot of basic features
* Morphing
* pbs texture animation
* texture channel selection
* skeletal animation for unlit
* fog
- Instant Radiosity is unusable slow (4x slower)
- missing a new GPU particle system
- no gltf loader
- no tools! ok spookyboo's awesome hlms editor. It's impossible to write a shader material editor.
- Its hard to mix compositors from the samples
- the overall easy usability is not there
- pull requests are just ignored or discussed/declined to death
- the cmake system is a mess
- code isn't stable -> nearly each update i pulled broke some shit
- i like one megashader where i can turn features on and off. In Ogre 2.1 you need to implement each feature in each hlms!
- the code is bloated with old trash and this v1 -> v2 dirt
- 2 materials systems!?
- Ogre 2.1 is still using bitbucket while everything else is using github.
- after 3 years its still a science to make/add a gui?
- community is dead! (well, i think this has to do a lot, with the 'there's only my way' attitude)
- the main dev mathias took a break to write a book for the last 3-4 users?

My Opinion:
ogre 2.1 just lacks a lot of features to make a game as non graphics programmer (what was able with ogre before 8+ years).
The hlms is very flexible, if you can code c++ and can make shaders and can make a rendering engine by yourself!
It's just sad that ogre didn't pull off like Godot and is dying/dead.

So who is still using ogre 2.1 and why?
Who Switched to alternatives and why?
Who considers using ogre 2.1
My little OGRE engine -> FrankE WIP
xrgo
OGRE Expert User
OGRE Expert User
Posts: 1148
Joined: Sat Jul 06, 2013 10:59 pm
Location: Chile
x 169

Re: State/feedback/rant -> Ogre 2.1

Post by xrgo »

devxkh wrote: Tue Feb 06, 2018 2:36 pm So who is still using ogre 2.1 and why?
Who considers using ogre 2.1
Hello! I am happy 2.1 user =), I am no super mega genius programmer, but I have a few years of experience so its usually not that hard for me deal with ogre. I agree with most of the points you make, and disagree with some, I'll comment some =)...

For me 2.1 was a very good update from 1.9 because of performance and pbr, I do mostly VR and performance is critical, still need more performance but its waaaay better than 1.9.

The thing is... people want stuffs that works out of the box with all the features, and Ogre is far from that yet (nor 1.10 nor 2.1) besides Ogre its just a rendering library that actually does a lot more than just basic rendering stuffs (scene manager, math, many advanced rendering features already integrated like azdo, clustered forward, etc. ) But its definitely not a game engine, and most people want to make games! so there's many options for that (godot included) that satisfies that need out of the box. But for people like me (I do simulators) which like to have (almost)full control of the engine and tailor it to our needs Ogre its perfect, I can't imagine writing my own renderer in opengl, I think it will take me 5 years and will never have good performance lol.

The main problem is that there are very little people writing their own engines, and many less people wanting to contribute to a rendering engine, so there's Matias doing all the work alone, so unless we can clone him :D, chances are Ogre will still advance at this "slow" rate. My personal dream is that paroj jumps to the 2.2 wagon and with those 2 brilliant minds focused in one version things will be quicker.

Its our job (the community) to extend Ogre from its basis... gui, tools, etc. and let the big minds focus on making the best rendering engine around =D!
devxkh wrote: Tue Feb 06, 2018 2:36 pm - missing a lot of basic features
I think... for a rendering engine it has a lot of features, most of the missing ones can be implemented by the users utilizing core ogre features/code
devxkh wrote: Tue Feb 06, 2018 2:36 pm - the overall easy usability is not there
I half agree: for just showing stuffs using pbs its fairly simple, with just a few lines of code you can have a window rendering an object with materials and stuffs.... but when creating your engine I agree Ogre its not the most easy to use library, but its a problem difficult to solve since Ogre its suppose to give you the rendering tools for you to implement an engine as you like, so there's many, many things to learn. I think this is solvable with more samples and docs, but time/people :P
devxkh wrote: Tue Feb 06, 2018 2:36 pm - pull requests are just ignored or discussed/declined to death
Agree... but time/people :P
devxkh wrote: Tue Feb 06, 2018 2:36 pm - code isn't stable -> nearly each update i pulled broke some shit
Strange, it only happened to me a few times, usually matias make a post anouncing the api break, and takes me 5 minutes to fixe it. I think its very stable
devxkh wrote: Tue Feb 06, 2018 2:36 pm - i like one megashader where i can turn features on and off. In Ogre 2.1 you need to implement each feature in each hlms!
this is what the pbs implementation do... automatically enables and disables features as needed/configured
I do think that unlit and pbs should be mixed in together :P
devxkh wrote: Tue Feb 06, 2018 2:36 pm the code is bloated with old trash and this v1 -> v2 dirt
yeap... but a big piece of that old trash is gone in 2.2, since its texture related and its a key outdated part of ogre that matias wanted to change since the begining
devxkh wrote: Tue Feb 06, 2018 2:36 pm - 2 materials systems!?
that should be easily fixed but I think its ok to maintain certain compatibility for people transitioning
devxkh wrote: Tue Feb 06, 2018 2:36 pm - community is dead! (well, i think this has to do a lot, with the 'there's only my way' attitude)
sadly that's the main problem

lets keep cheering Matias so he can continue his great work! and maybe hope someone else joins the team =)
Saludos!
frostbyte
Orc Shaman
Posts: 737
Joined: Fri May 31, 2013 2:28 am
x 65

Re: State/feedback/rant -> Ogre 2.1

Post by frostbyte »

xrgo wrote: Tue Feb 06, 2018 11:50 pm The main problem is that there are very little people writing their own engines, and many less people wanting to contribute to a rendering engine, so there's Matias doing all the work alone, so unless we can clone him :D, chances are Ogre will still advance at this "slow" rate. My personal dream is that paroj jumps to the 2.2 wagon and with those 2 brilliant minds focused in one version things will be quicker.

Its our job (the community) to extend Ogre from its basis... gui, tools, etc. and let the big minds focus on making the best rendering engine around =D!
Saludos!
exactly...but there is an egg-chicken problem here
problem is that ogre 1.10 still have far larger user base( ROS and other large projects that rely on ogre )
it also provides far better documentation/knowledge base and have steady support for mobile/web

so ogre is stuck with two live versions which are actually two very different projects( heck...two repositories )
thus requiring at least two maintainers....i'm not sure anyone has a solution for this akward situation

speaking for myself porting my huge 1.10 code base to 2.1 is a lot of hassle( 1-2 months ), and for the future i imagine for my project, without proper mobile/web support it would probably be better for me to look into other rendering solutions
which i really really would like to avoid...
devxkh wrote:It's just sad that ogre didn't pull off like Godot and is dying/dead.
sure today all kids mammble "godot" in their sleep, but this is not a popularity contest https://github.com/godotengine/godot/issues
also don't be fooled by the stars on github, most of them were there long before the newly released godot-3.0 others were rewarded by existing user-base
godot2.x was much the contender of libGDX mainly used for 2d/2.5d
so comparing popularity of a 2d game engine to a 3d "i'm only a renderer" thing is not fair
but playing along with this comparison, i hope you can show me one 3d game made and released with godot...goodluck...

btw...kudos for your engine, looks very well designed...i might use it if/when i port to 2.1 :wink:
the woods are lovely dark and deep
but i have promises to keep
and miles to code before i sleep
and miles to code before i sleep..

coolest videos link( two minutes paper )...
https://www.youtube.com/user/keeroyz/videos
Hrenli
Halfling
Posts: 73
Joined: Tue Jun 14, 2016 12:26 pm
x 19

Re: State/feedback/rant -> Ogre 2.1

Post by Hrenli »

I want to comment it from slightly different angle. I want an open-source 3D-renderer. I don't want a fully bloated game engine because I'll have to strip it down from a lot of unnecessary crap because I am not doing a typical game with premade scenes/levels and stuff (if you read Project Sightseer's developer's blog you'll get an example of how far some people have to go to make a game in Unity and ending up with almost no Unity underneath).

So, what are my options? Not much to start with. And between those very few which are available Ogre 2.1 looks to me as the best solution. Yes, you have a point in the rant department (although from my experience it's quite stable, if it breaks it usually means I am trying to use new code with old hlms templates yet again, that's it). But there is no better alternative. Or is it? :)
al2950
OGRE Expert User
OGRE Expert User
Posts: 1227
Joined: Thu Dec 11, 2008 7:56 pm
Location: Bristol, UK
x 157

Re: State/feedback/rant -> Ogre 2.1

Post by al2950 »

I feel like I have to comment on this topic, so here goes. I have used and still use most versions of Ogre. I actively develop with 2.* but have systems which I support which still use Ogre 1.7 - 1.10.

Everything is harder than it used to be, but then again everything used to be fixed function. Graphics in the modern world is just simply hard! And there is not much we can do about it. Ogre does not force (this is not entirely true) specific graphical implementation, but tries to provide tools/snippets that allow you to create your own rendering/games engine. This point is quite important as I don't really see Ogre as a out of the box ready to go rendering engine.

Before I say the next bits, I would like to say I owe dark_sylinc a lot! Ogre was pretty much dead when he started the 2.x work and if he had not done all that awesome work I may have resorted to Unreal, which probably would not have turned out well! However he is a one man band, and his roadmap seems to be based on what he is excited on doing, which admittedly you can't blame him for. I would like to have a more organised roadmap, and for dark_sylinc to be more comfortable dishing out tasks to the community, or maybe trying to get some more Ogre team members on board (current or new team members). If dark_sylinc decides to pack it in tomorrow, it would leave a lot of us up shit creek with no paddle... well maybe half a paddle!

Bitbucket vs github.... well Git (especially on windows) pisses me off! but I like some of the feature on github! either way Ogre uses both :D

Particle systems really needs to be high up the list. I have an implementation to test some concepts but needs a good week of development to produce something everyone can use.

There is plenty more I could say, Godot has does some impressive stuff recently, but its not without its issues. Ogre should be easier to build and should have more features ready out of the box, I am looking at you reflection systems.... ogre is carrying some baggage from 16 years of actively development. BUT over the years I have found Ogre fairly easy to use, well documented, and although the community has shrunk there is always someone who will help if then can; furthermore I am now able to produce systems that look better than competitors 100x my size, not to mention perform better. Most my systems now use 1 PC to drive 4-6 screens, competitors have entire server rooms!
hedphelym
Gremlin
Posts: 180
Joined: Tue Nov 25, 2008 10:58 am
Location: Kristiansand, Norway
x 23

Re: State/feedback/rant -> Ogre 2.1

Post by hedphelym »

I'm very happy with Ogre, our team (at my workplace) have been using ogre for many years, and I like it because you start from the ground up,
you get to choose the exact components you want to use together with it, physics, network, audio, GUI and everything else.
This is why Ogre shines if you ask me. There are not a lot of libraries like it.
We use it for high-end simulators, so we are extremely picky about what libraries we use, and how we use them.
With other engines you are more dependent on what they choose to do, and often it might not be changes that are good for you.
And Ogre 2.1 is amazing, it's a huge leap forward, and puts Ogre up there with the other 'big' game engines.
We're currently porting all our 1.9 stuff directly up to 2.1, and I'm very happy about it.

The downside I personally think is the number of Ogre developers compared with other well known engines,
this is always something that comes up when people ask us about what we use (I understand why they ask this).
This also affects recruiting some what, because there are a lot of developers using for example unreal then Ogre.
But personally I feel it does not matter much - if you know c++ you can learn both, just takes more time to get into to begin with.

Also there are some small things that puts newcomers off, and I think one of them is just getting their assets into Ogre and on to the screen.
It's been mentioned before that for example GlTF support would be great, that lets you quickly get stuff out from Blender and into Ogre.
Community members are working on that, but it would be great to have it officially supported in the future as well.
Either that or get a new blender addon made, that writes to binary .mesh files directly, without having to use the command line tool to convert the xmls.
I might pick that up soon and do that (I've wanted to make that for a while) - unless a format blender can export gets added to ogre itself.

I think it's important to discuss these things, so I'm glad the original posted wrote the post, gets us all to reflect on things and share what we think about it.
frostbyte
Orc Shaman
Posts: 737
Joined: Fri May 31, 2013 2:28 am
x 65

Re: State/feedback/rant -> Ogre 2.1

Post by frostbyte »

devxkh wrote: the main dev mathias took a break to write a book for the last 3-4 users
looking at latest commits and forum answers i don't think that is true
but anyhow, if he does then i do hope he will write a "customize your HLMS" chapter( hint... hint... :wink: )
hedphelym wrote: Fri Feb 09, 2018 6:50 am you get to choose the exact components you want to use together with it, physics, network, audio, GUI and everything else.
This is why Ogre shines if you ask me. There are not a lot of libraries like it.
this is true, but still i think ogre can do better on the AddOns department
as their are eventually a limited number of AddOns i guess it would be better to tie them up in one cmake-script for downloading and installing ogre addOns
ideally the script would be provided with ogre, and ogres team will merge pull-request for it( made by plugins develeoper )
this is what ogitor's build system does with its plugins and i find it to be a very simple( just few lines of cmake code ) and effective way for getting ogre plugins( just hit cmake configure..generate..build..... and binaries ares ready :) )
hedphelym wrote: Fri Feb 09, 2018 6:50 am Either that or get a new blender addon made, that writes to binary .mesh files directly, without having to use the command line tool to convert the xmls. I might pick that up soon and do that (I've wanted to make that for a while) - unless a format blender can export gets added to ogre itself.
are you aware of xrgo's blend-loader? https://github.com/yoyTeam/BlendLoader
if not then this quote just emphasize the need of having a centeral place/ build system for ogre related addONS/plugins

like xrgo said, right now the best thing that can happen to ogre is that paroj decide to jumps the 2.x wagon, i think he has good in-sight about making ogre more wholly, accessible and friendly both to simple users and core-developers
but like i said before 1.x still has a big user base, so i'm not sure if ditching 1.x is a fair or good idea...
eventually all i want is 2.x level features/performance but with 1.10 level support, i'm sure it will happen with time,i just hope it would'nt take too much time...
the woods are lovely dark and deep
but i have promises to keep
and miles to code before i sleep
and miles to code before i sleep..

coolest videos link( two minutes paper )...
https://www.youtube.com/user/keeroyz/videos
User avatar
EricB
Bronze Sponsor
Bronze Sponsor
Posts: 360
Joined: Fri Apr 09, 2010 5:28 am
Location: Florida
x 213

Re: State/feedback/rant -> Ogre 2.1

Post by EricB »

I guess I'll throw my two bits in.

I view 2.1 as absolutely necessary in order to save Ogre. But at the same time I also consider it beta software that shouldn't be called "2.1" yet. Hell it shouldn't even be called 2.0 yet. Why? Because as devxkh mentions, it's missing a lot of features. I think the premature naming scheme has hurt more than it helps. I mean, we've already got rumblings of 2.2, when 2.1 API isn't even stable yet. Most folks who come by on IRC (know thy audience, foss programming community generally doesn't use fad crap like "glitter" or whatever it is) expect 2.1 to be stable, have documentation and tutorials, etc. Just because of the name. They're sorely disappointed when we tell them it doesn't.


I think as a whole Ogre died 6 years ago for 3 reasons.
1) Sinbad left. Large foss projects need either a strong man or a strong counsel. Post Sinbad leadership was neither, and the majority of them slowly drifted away to different projects and communities. The work they did do was fairly disjointed, leaving many half baked ideas, projects, and schemes. (As an aside, I think Dark is filling the role of a strongman well.)

2) The rise of free engines. Unity and Unreal has stole the limelight. Hell even Sinbad is making a game in Unity now. These engines became popular. Why? Tools. Standards. Ease of use. Most folks don't like reinventing the wheel. This is why things like Godot is becoming popular.

3) The insistent attitude that "Ogre Is Not A Game Engine". Yes, I agree Ogre is just a rendering engine. But this is what ultimately killed Ogre's momentum. It was that hell bent stubborn attitude that killed Ogre. While it is fine to keep Ogre as a stand alone renderer, I don't see why we couldn't have throw resources into maintaining official addons and tools when Ogre was at it's peak. Instead we have a hodge podge of half implemented, hackish band-aid, and completely abandoned addons which are essential to fast paced game development that is required these days. Not to mention, core things like raycasting to the polygon level is a wiki code snippet with 30 different variations when it should be a core feature... I spent 6 hours trying to go from a raycast, to triange, to uv cords, to RGBA value AND FAILED a couple nights ago. It's 3 lines of code in other engines.

Let's put it to you this way. There are 3 major 3d modeling programs. I remember a time when we had no free exporters for 2 of them, and the Blender one was hackish at best and didn't work with the latest versions. All because "Tools" are not a part of a rendering engine.





GearCity is 7.5 years old now (the engine about 8 years). It's not heavy in 3d, but there is some. It's heavy in GUI, and I was burned quite badly by choosing QuickGUI (which was abandoned about 2 years into my project). All the other plugins I use have also been abandoned, berkelium, OIS, OgreOggSound. it's a wonder they all work on modern OSs. (But not for long on OSX.). And I've gone as far as having to rewrite parts of some of them my self. The Ogre addon community is in a state of bit rot, all because the community threw resources behind abandoned renderers and not picking winners and losers to devote resources to. As such, usability of the rendering engine has degraded significantly for most users.

To put it into prospective, my main competitor ditched their homemade engine for UE4, it took them about 12 months, but now all their artwork looks insane. They click and drag shaders, and it just looks pretty with very little fuss. I highly doubt they do much C++ programming for graphics or UI anymore. Me, I fight making panels and buttons in QuickGUI because Ogre had 8 GUI libraries, none of which had a layout editor in 2010. And the Ogre team didn't say "This is the one you should use".


Which brings be back to the future. As a 1.x user, I applaud Paroj's efforts in maintaining 1.x. However, I believe we should only be doing bug fixing and keeping the code working on modern OS's. I read that 1.11 will have 100+ breaking API changes... Why? I believe pushing such work is pointless and trivial. And it kind of defeats the purpose of keeping 1.x around (our old code still works). If time is going to be spent on new development, it should be put into 2.x. Of course it's Paroj's choice on how he spends his time, but I think as a whole it doesn't help Ogre.

Which brings me to Git and Bitbucket. They both suck, but having two different repos is annoying. Pick one.
Data rot. We almost lost everything in the Addon forums. Wiki has been slated to be deleted/replaced? by github pages. I can't find comments for 1.7.4, 1.8, 1.8.1, etc anymore... The Data Rot has got to stop. It's piss poor management. I can grab a FreeBSD comment as far back as 1993. It's pretty pitiful that text based content is being lost in this community, when hosting is essentially free.



To answer the OP's final questions. I do not use 2.1. I have no direct plans on using it. But I see it as necessary development if Ogre doesn't want to disappear. I don't see Ogre ever becoming a popular game development engine again. And I don't believe a robust tools/addon community will ever appear again either. My option is that Ogre will maintain the status quo. A few small indie games here. Someone who wants to take advantage of it being FOSS there. Industrial uses abound. But the majority of video game developers (and thus the bulk of pre-1.8 users) are not coming back.

I personally use 1.7.1 on Windows/Linux and 1.8.1 on OSX. I will be bouncing everything up to 1.10.11 in the end if it is more stable (no point in chasing technology if shit just works as it is).

I may have one additional spin off game using my Ogre3d Engine, after which I am moving to QT3D. (If the renderer doesn't have all I need, I will consider plugging 2.1 in IF it does.) It'll be around 11 years with Ogre. It's about time I retire to easier systems. :)
Image
User avatar
spookyboo
Silver Sponsor
Silver Sponsor
Posts: 1141
Joined: Tue Jul 06, 2004 5:57 am
x 151

Re: State/feedback/rant -> Ogre 2.1

Post by spookyboo »

I agree that all forces should be combined in the 2.x versions, but that's the choice of paroj (and eugene).
I think Ogre should not even try to compete with Unreal, Unity and Godot, because it can never win this battle. Ogre fills in a niche market (visualization) and that must be emphasized even more. Put on the homepage that you should not build a game with Ogre, unless you have a very specific reason to do so. Add some examples of non-game projects. This is where Ogre really shines and that is the future of Ogre in my opinion.
This also decreases the scope of tooling very much. A full blown particle system or a scene editor may not be needed. That still leaves the question what is needed (eg. native import of Blender or Max models)...
User avatar
Ybalrid
Halfling
Posts: 89
Joined: Thu Jul 10, 2014 6:52 pm
Location: France
x 31

Re: State/feedback/rant -> Ogre 2.1

Post by Ybalrid »

Here's my grain of salt:

I think I agree with most of what has been said here. Ogre is still an awesome free and open source rendering library, and Matias is doing an incredible job on the 2.x branches essentially all by himself. 2.x is still missing important features, but I think we gained a lot in term of raw performance, graphics capability, and flexibility from 1.9 back then.

Contrary to what spookyboo's saying, I don't think Ogre should be confined to be just a "rendering library for visualization". It can still be a building block for game engines.

But I really agree on this : I don't think the on going support for 1.x and the divergent development that is occurring right now is a good thing. The 1.x branch is supporting/adding features to RTSS after having back-ported HLMS? Why? 1.x should be on the "we fix bugs, but new shiny stuff is developed in Ogre 2" stage now.

Also, the addon situation is obviously a bit problematic. I ran into it recently while trying to use ImGui with Ogre 2.1 (I bodged something together, I ran Ogre in an window and OpenGL context I create with GLFW (or SDL2, but I picked GLFW at random and I'm happy with it currently) so I get something that works a little bit by putting the ImGui final rendering in OpenGL only in a RenderTargetListener...

For some context on what I do, I'm just a student in engineering school, I don't have much real experience in graphics programming, I wanted (mostly for fun, and to learn how the headset worked) to build a game engine by hand for using the Oculus Rift, and Ogre was the perfect open source building block for that.

The switch from 1.9 to 2.1 was a bit painful, the lack of clear and easy to follow documentation was a big problem. Some obscure things I only solved by asking Matias on his twitter or on this forum. I gained a easy way to get awesome looking material (with the stock implementation of the PBS HLMS) and also lost the ability to easily export models with materials from something like blender. Granted, Spookyboo's HLMS editor is a great project and is really plugging that particular hole, but well, you see where I'm going with this.

I really enjoy Ogre, I think it's a great project. I don't know enough about graphics programming to really help with it's internals, but there's definitively features that I would like to see (particules and morph target comes to mind, there have been open PRs about them on BitBucket for ages, they are really big code changes and I understand Matias 100% for putting them aside, But It would be neat to see them merged-in a some point if possible...)

Basically, the main problems Ogre faces are:
  • There's virtually 3 persons actively maintaining Ogre
  • Development is splited between Matias doing 2.x and paroj/eugene maitaining 1.x
  • The community has shrunk, probably because now full fledged commercial game engines are free to use (unity, unreal) and some Open Source projects moved away from Ogre (OpenMW switched to OpenSceneGraph for example)...
Somehow, I don't think we are in a sad place. We have a great code-base under a great license that has some problem due to it's age, and not enough people maintaining it and using it in new projects...
Ogre_glTF Ogre v2-1 GLTF2 loader : topic link github repo
BtOgre21 Fork of btOgre, for Ogre v2-1 : topic link github repo
OIS Current maintainer : Official repository
Annwvyn VR focused game engine using Ogre : https://github.com/Ybalrid/Annwvyn https://annwvyn.org/
paroj
OGRE Team Member
OGRE Team Member
Posts: 2106
Joined: Sun Mar 30, 2014 2:51 pm
x 1132

Re: State/feedback/rant -> Ogre 2.1

Post by paroj »

Ybalrid wrote: Thu Mar 08, 2018 7:37 pm I don't think the on going support for 1.x and the divergent development that is occurring right now is a good thing. 1.x should be on the "we fix bugs, but new shiny stuff is developed in Ogre 2" stage now.
There are actually many projects for which 2.x is a too big leap to take. For instance the loss of all existing addons as you noticed. Sure, the 1.x architecture imposes performance bottlenecks and headaches for the maintainer, but there is still enough room to boost performance and add new shiny stuff.
So with all the legacy code in mind, which would immediately benefit - why should we not?
Ybalrid wrote: Thu Mar 08, 2018 7:37 pm The 1.x branch is supporting/adding features to RTSS after having back-ported HLMS? Why?
this basically
https://github.com/OGRECave/evolution/b ... th_RTSS.md
hyyou
Gremlin
Posts: 173
Joined: Wed Feb 03, 2016 2:24 am
x 17

Re: State/feedback/rant -> Ogre 2.1

Post by hyyou »

I use Ogre because it is free and open-source.
For me, the outstanding aspect of Ogre was the very great tutorials (http://wiki.ogre3d.org/Tutorials).
I guess the reason that people may switch to others are currently
  • hard to install (esp. SDL), lacks ton of good tutorial, not so many games use Ogre (recursive problem)
Ogre is still attractive for people who love free library and flexibility, or suffer from performance cost of heavy-weight game engine (Unity & Unreal). I heard that they are not fast as expected.

------ Below is off-topic ------
paroj wrote: Thu Mar 08, 2018 8:31 pm So with all the legacy code in mind, which would immediately benefit - why should we not?
Because a person who is very capable, talent and generous enough to dedicate time in developing Ogre is so much valuable.
It would be nice if such precious resource is used to create things that will not be obsolete for 5-10 years in the future.

With my respect, this is just a newbie opinion of mine.
If only I can convince you ... XD
User avatar
Herb
Orc
Posts: 412
Joined: Thu Jun 04, 2009 3:21 am
Location: Kalamazoo,MI
x 38

Re: State/feedback/rant -> Ogre 2.1

Post by Herb »

As someone who has used Ogre since 08, and just getting back to Ogre (been close to 3 years since I did much with it), there's some real value to what @paroj is saying... I picked up 2.1 as everything seemed to point to that and the 2.x stuff had been under development for a long time.

The trouble was lack of examples, code snippets, tutorials, and add-on libraries. Other than a great library, the biggest value (in my opinion) of Ogre is the community. The community around Ogre that answers forums questions and provides a lot on the wiki (for tutorials, snippets, etc). I just don't see that with 2.1. I feel like the development on 2.x needs to slow to get more documentation and tutorials going. Also, to @paroj's comment on add-ons....I know Ogre is just a rendering engine, but when making a game/simulation/whatever, it's difficult to re-invent the wheel when 1.10 will get you there with a lot of examples and addons right out of the gate.

In the end though, there is strange feeling of Ogre being "two" libraries going forward...
User avatar
syedhs
Silver Sponsor
Silver Sponsor
Posts: 2703
Joined: Mon Aug 29, 2005 3:24 pm
Location: Kuala Lumpur, Malaysia
x 51

Re: State/feedback/rant -> Ogre 2.1

Post by syedhs »

Ogre can actually slide into its own niche - which is professional-hobbyist rendering (or games engine) provided that
1) it has nice tool-chain, from model export down to material editing and world builder (yes world builder).
2) documentation (but this is second reallly)..
3) some sort of scripting which is officially supported - yes this fall into I am looking at you and you are looking at me.. :mrgreen:
4) and nice dynamic lighting support.. maybe far-fetched but we can always use NVIDIA VXGI.. for an example.

UE4 is nice, but it has its own problems. It is huge and huge as its own drawbacks. Ogre is like a backyard-tool but can be powerful provided the tool-chain is there.

And yes, I agree that the notion 'Ogre is only rendering engine' was what really killing it because back then, when Ogre3d was very popular (and no freeish UE4 and lowcost Unity around), I have often met with resistant when I asked many times why there is no tool-chain project in GSOC.. Well no need to dwell in the past I know, just saying.
A willow deeply scarred, somebody's broken heart
And a washed-out dream
They follow the pattern of the wind, ya' see
Cause they got no place to be
That's why I'm starting with me