Hi Eric, I will go back up as our inline inline is getting too hard to follow. What we have been doing the last few years is bringing on very focused dev talent like the NEX guys or unfold 3D guys and signing them to multi-year contracts so we do not get the plugin pasted in and it dies. This gives us 2-3 releases to avoid the rot syndrome that we have seen in other packages.
The concerns over Bifrost are totally valid. I told you that at our meeting. Again revamping the UI with QT, the viewport, and the data model has taken many years and lots of hard lessons learned in the heart of production. Bifrost will get put immediately into the hands of hundreds of thousands of students, professionals, pipelines and pirates so we will know if the base engine will run on the shittiest hardware we track all the way up to the machines of tomorrow we run in the labs of places like HP and Intel. Again the first important thing we wanted to get out of Bifrost was scale and performance. As we build out the TD UIs (node based and python) and then developer platform there will be challenges at each step but with Maya in the most grueling pipelines in the world we get the hard feedback very quick. I understand why everyone here is skeptical about our ability to innovate or trust what is being written. We have to show progress on these topics in short order and keep it up or you will find other solutions eventually. cv/ ________________________________________ From: Eric Thivierge [ethivie...@hybride.com] Sent: Wednesday, March 19, 2014 12:35 PM To: softimage@listproc.autodesk.com Cc: Chris Vienneau Subject: Re: "Top List of ICE Nodes That Cover 80% of What You Do With The Toolset" Replying Inline as well. :) > A few questions: > 1. How do you respond to the people who have been long time die hard > Softimage users who have also been exposed to other DCC's, maya specifically, > who have little to no faith in AD being innovative or responsive to their > user base as history has shown. I can give you a specific example. Skin > painting. How many years has it been that it has been in its current form, > and your user base asking for it's interaction model and tool set cleaned up > and extended. Yet here we are just prior to the 2015 release and it's gotten > no attention. It's been years people have been asking for this. Yet nothing > from AD. Same with the blend shape tools. No attention. Take a look at the > various threads on this list and 3D Pro where Maya veterans say that the use > of 3rd Party tools is a must! We need a developer that actually listens and > turns around results quickly. Not taking 5+ years to not even address it. > > - As we have been saying the last few weeks the main drive for the last few > years was to modernize Maya's core so that we could start to modernize the > workflows on top of it. Having a good workflow for fluids or particles and > not be able to generate the quality and scale required to generate the fx > that are the norm was not acceptable. The first two areas we focused on was > FX and modeling and yes we started with the NEX toolkit but we have kept the > developers that built the plugin on to integrate all the technology correctly > into Maya. I think if you look at the work done in uv editing, retopology, > base modeling operations, and workflow the work is detailed and broad/ We are > looking at making big progress in animation and lighting/rendering very > shortly. Maya LT is releasing every three months and we are looking at > building back up the pace with Maya. Thanks. As per the NEX devs, is it intended only to keep them on until integration is done or are you guys thinking that their strategies can be useful in the future and plan to have them work on other areas? In the same vain as the core modernization, will there be other things that may come up that prevent you from working on the various tool sets much like the core modernization has? > > 3. So the first release of a tool built with your node graph will be released > in Bifrost. How long do we have to wait until the node graph is accessible? > (Granted I know you can't tell us, it just has to be asked). It's known that > with Maya releases and new features that the first version is never > production ready. You could say that for most new features in all software, > but when we think about it, if we hypothetically say the node graph is > another year off, that first release won't be usable and so that puts us to > the next year's release, 2017. At that time most studios will need to have > had to transitioned off of Softimage and onto another platform, such as Maya. > So at that point we have more or less zero time to get acquainted with the > new system and integrate it into the pipeline and build tools around it. > That's all with the wishful thinking everything goes to plan. > > -You know I can't tell you that but since you have seen the graph with your > own eyes you at least know it is there. Anyone under NDA can get all of the > information here. I know you can't tell me, but do you see where there are valid concerns? We could still be 2 years away right now from being able to use and integrate that new toolset into our pipeline. The whole initial plan was to have everyone transitioning off of Softimage within the next 2 years, then AD backpedaled on that. What would have happened if you guys stuck to your original plan? There would have been a huge gap left. > > 4. Let's not kid ourselves. AD is a company who for the majority of their > major new features lately, acquires technology and integrates it. NEX is the > most recent to come to mind. In a larger scale sort of way that is exactly > what you did with Softimage. Bought it for the devs and are now trying to > integrate the technology. How is this supposed to bring confidence to users > who need to use Maya? It's just a bunch of plug-ins that were bought and > slapped together. There doesn't seem to be a unified workflow thought out of > how these all need to play together and thus gives you a very fragmented > workflow. Not to mention, what happens when there is a year where you don't > acquire a software? Does Maya not get a new feature that release? > > Isn't your pipeline a collection of applications glued together with a > unified workflow or intent? We are clear that Maya needs a UI/workflow > overhaul to bring coherency to many of the workflows and there are tools in > Softimage like Gator which show the power of such an approach. In this > release we had several features that were developed in house like Geodesic > Voxel Binding and the whole viewport effort was built on technology entirely > developed by Autodesk. So Eric by that logic are you saying that putting > effort into Ptex support, UV tiling, Alembic, Open EXR 2.0, python, etc... > are not good things? Yes it is, even more reason why I don't need one of those application's glued together within itself. Pipelines are complex enough let alone having to fight within one application to have it do what I want and work productively at a good pace. There is a difference in integrating Open Source projects than just buying tech, integrating it, and letting it rot. The OpenSource projects continue to get dev from outside of AD and evolve. > 5. Lastly, who are these other key people who remained at AD that worked on > ICE? It may give us reassurance to know what good hands we've been left in. > (Not really expecting an answer here because it'd be dangerous for AD to list > their employees, but it's more the point that we don't know who these other > key people are and thus, have no reason to be confident in them.) > > -I will ask those employees if they want to have their names listed but I can > understand if they don't want to do so. ICE was derived from a very novel > programming methodology that had fallen out of favor and which is why Fabric > can start up with no harm of IP infringement on ICE. I have many names over > the years mentioned as key contributors of ICE and wanted to make sure that > the fact that many people who worked at Soft still work at Autodesk and that > key people from the ICE team work at Autodesk and on Bifrost was understood > as well. Not sure why Fabric is getting dragged into this... again. But what novel programming methodology are you speaking of? Surely not node based programming. In the end, all I'm trying to illustrate here is why the vast majority of Softimage users have little confidence in AD. If you want to move forward and convince us you really need to be honest about the history and put a tremendous effort into changing the way things have been done at AD in the past on all fronts. Not just updating bits and pieces and hitting bullet points. Personally, if Maya didn't have the market share and be the only product on the market that had the amount of decent rigging and animation tools it currently has, I would be jumping off the AD ship ASAP. I don't like the way AD works overall. However, since you have the market cornered in these respects, I have no other option. If the AD mentality doesn't make a big shift, and another solution comes along, you're going to find yourselves with a lot of people jumping ship. I wouldn't take it for granted. Eric T.
<<attachment: winmail.dat>>
