Hi Allen, I agree on many things you say. A public roadmap is quite important. Perhaps everyone in the community should create a wish list so we can create together a roadmap and hopefully more people get involved implementing it. I think we have many example programs but you only get them if you compile OpenSG yourself. We could copy the important test programs into the tutorial or example section so they are automatically included in the binary dists. On linux it is really easy to compile OpenSG, on windows it is too complicated. For the 1.8 I try to improve this and for OpenSG 2.0 you are already working on it :-)
Ok some ideas how to improve it: - add flex, bison, ... (win32) executables to the cvs and include the search path to the visual studio project files. scons related stuff. - get rid of the scons dependency. There exists a special local scons distribution just adding this to OpenSG cvs removes the scons dependency ;-) - get rid of the cygwin dependency. Right now we use some tools like sed, flex, bison just add the executables to the cvs. On linux there is no dependency left just type "scons" and that's it. On windows there is one dependency left but installing python is really no big deal. Yes the webpage is a bit confusing. I already created the wiki page to collect some OpenSG 2.0 ideas. Perhaps we should extend this and make a real webpage out of it. The most important point is we need more people getting involved in the further development. Andreas > Hello all, > > What I am about to say may be flamebait, I don't know, but it has > started to bother me enough that I couldn't hold back any longer. I just > feel I need to speak up. I know many of the things in this e-mail may > sounds harsh, but as someone deeply involved in OpenSG and who is > betting their company's future on it I must call it like I see it. So > here goes, let's stir the pot up. :) > > The problem > ----------- > I have been using OpenSG since the beginning and my company is currently > using it for all of our development. We do this because IMHO OpenSG is > the most technically advanced open source scene graph out there. But > recently I have had to start re-examining whether it makes good business > sense to keep using OpenSG. Don't get me wrong, I love OpenSG, I think > the developers are great, but OpenSG is losing out and is falling > further and further behind other efforts, most notably OpenSceneGraph > but also relative to other projects out there. > > For an example of this just check out the following links: > > - http://www.openscenegraph.org/ > - Community activity: http://tinyurl.com/m3sxe > - Siggraph 2006 presentation at: http://tinyurl.com/levs5 > - Contributor list: http://tinyurl.com/rl7xc > > OpenSceneGraph is growing by leaps and bounds. They have nearly 200 > contributors to the codebase and have more plugins on the way. I highly > suggest that any OpenSG developer take a detailed look at the OSG > website to get an idea of what they are up to. As they say, know your > enemy. I know we don't often think of other open source projects as the > "enemy", but remember that every user and contributor on OpenSceneGraph > is one less for OpenSG. Unfortunately it is a zero-sum game. > > But back to OpenSG. I have a simple question: What is the goal of > OpenSG and the OpenSG developers? > > Is the goal to have a codebase to do research? To work on fun projects > on the side? Or is the goal something large like to create the most > widely used open source graphics scene graph? > > The state of things > ------------------- > As it stands now I don't know what the goal is. It really seems like > there is no central direction or push to make real effort on the project > towards a goal. There is no roadmap and it often seems like there is no > rhyme or reason to the development direction. I know this sounds harsh, > but I have heard multiple people describe it this way. > > For example here are a short list of some of the symptoms I see. > - No roadmap describing what needs to be done and who is working on > it and no urgency to make progress. > - Relatively few contributors. There are several core contributors > but very few community members contributing small bits of code. > - Relatively few people in the graphics community know about OpenSG. > - Still too few loaders compared to other scene graphs. No one has > an incentive to do this grunt work but it still needs done. > - Many things never get completely done. (FLT loader, experimental code) > - OpenSG 2.0 development is languishing. There doesn't seem to be a > big push to move over to this codebase. When GV originally started on > it many of the ideas were more advanced then any other scenegraph, now > others are catching up and OpenSG 2.0 is still not moving forward very > quickly. > - Relatively difficult to build and install. > - There are multiple build systems. I don't even know which one > is the right one to use. > - Need more pre-packaged builds for common platforms. > - Need to get OpenSG into Fedora Extras and Debian (like > OpenSceneGraph) > - No promotion of new features > - OpenSG has cluster-based rendering. It can render over 1 > billion polys using a cluster but no one knows this. > - Cluster rendering on tiled displays. > - Occlusion culling was added a while back and there wasn't even a > user announcement. > - The shadowing code still has no public announcement. > - Staged rendering in OpenSG 2.0. > - These are all killer feature that nobody else has and OpenSG has > had it for quite a while but there has been no demo, no announcements, > nothing. If we don't tell people about OpenSG how can we expect them to > use it? > - The webpage is confusing and hides the important information. It > is just hard to navigate. > - No where does it say, "This is why you should use OpenSG" and > "This is how it compares to scene graph XXX" > - And more.... > > Sorry about all these, but I think it would be hard to say any of it is > not true. :( > > As a side note, I think that the reason for the current state of things > (and correspondingly the reason that OpenSceneGraph is succeeding) is > that as far as I know there is no one whose livelihood directly depends > upon OpenSG. There is no one whose full-time job is OpenSG development > and consulting. OpenSceneGraph benefits greatly in that Robert and Don > have to make OpenSceneGraph succeed and work or they don't have an > income to support their family. That is a great motivator for success. > > > Call to action > -------------- > Now to what can be done. First, what do you all think? Are these > things issues in your opinion and more importantly do we really want to > solve them? If the goal of OpenSG is just to be a place to work on > research code, then these issues might not matter. If the goal is to > create the dominant and most featureful scene graph, then I think these > issues all matter. > > I think many of these things can be corrected but it will require a > commitment from the entire OpenSG community to take action and grow the > community further. The two areas I see as being most important in the > short term are active promotion and a shared roadmap. > > Active promotion is important because we need to get the word out. > People need to know about OpenSG and why they should use it. If there > is a new feature added to OpenSG, there needs to be a small demo in the > source tree that is compiled with every distribution. There needs to be > a page on the website with screenshots and descriptions and we need to > promote this demo on the web (opengl.org and all the others). We need > to make the website more community oriented and better organized to put > the important information right there in the user's face. > > I know marketing is not something developers worry about much, but it is > important. If the graphics community doesn't hear about what we are > doing then all our effort is wasted. It reminds me of something I heard > about research one time. "It is not research until it is published". > Well, correspondingly it is not open source development until someone > knows about it and uses it. > > A shared roadmap is important for two reasons. Users need to see where > the project is heading and developers need to agree on where the project > is heading so they know what to work on. I think this is the key to > keeping the project moving forward and maintaining inertia. It is also > a great way to get new contributors because it can help them be > interested in what is going on. > > What do you think? Are these the most important current needs? > > > What we can do > -------------- > So what can we do? First, I think anyone that cares about this issue > should reply to this e-mail and say so. Correspondingly, if you think I > am full of it, then reply and say that to. > > After that discussion I hope we can start really documenting the goals > and roadmap for the project. > > For my part, I am willing to put my money where my mouth is so to > speak. I will do anything I can to help out the project and I will put > the resources of my company at the community's disposal (within reason :) > > For now what I can do is: > - I will provide server space and network access for any tools, > websites, build farms, whatever is needed. > - I am working with Dirk on a single integrated build for OpenSG 2.0 > so we can finally have one build system > - I am working on making pyOpenSG work with OpenSG 2.0. After this > is done I hope to create a small viewer application that could serve as > a good demo of some of the advanced capabilities of OpenSG 2.0. > - If there is interest I will consider giving free non-commercial > access to our max exporter for OpenSG. It automatically generates > shaders to photo-realistically render the scene as closely as possible > to the max software renderer. (exports normal maps, bump maps, etc) > - I can devote some time for our graphic designer to work on things > like website graphics, models for demos, etc. Let me know what you need > and I will see how we can do it. > - I will help out with marketing. If you write the code and a demo, > I will help make the webpage about it and post it everywhere on the web. > > > In closing I would like to say again, I love OpenSG. It is technically > superior to everything else out there. I just want to make sure it > stays that way and grows to its full potential. > > -Allen > > PS. At first I considered just sending this e-mail to the core > developers that I know but I decided that since OpenSG is Open Source we > should probably have all discussions in the open. :) > > > ------------------------------------------------------------------------- > Using Tomcat but need to do more? Need to support web services, security? > Get stuff done quickly with pre-integrated technology to make your job easier > Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 > _______________________________________________ > Opensg-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/opensg-users > > ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Opensg-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/opensg-users
