www.si-community.com

Definitely relative I just mean that we'll start building some different things in the tutorials now that we're past installation and UI basics. There's a section in the forum where people are (hopefully) sharing canvas graphs they've built, so there'll be some more advanced stuff in there (along with the presets we're building each week as well).

Thanks for that, Paul! And I also appreciate the pace of these too, making things crystal Clear. The Erics are doing a great job with these.

Great to hear Tim! Trying to keep a good pace in the videos so it's easily digestible, while not putting the viewers to sleep.

Let me know what topics you're interested in having us cover - we have a list we're working through

FabricPaul wrote:Let me know what topics you're interested in having us cover - we have a list we're working through

Ok forget me if I missed some, I hope I'll get more familiar with FE during this Christmas time.

1: equivalents to ICE, or no equivalents if there's something structurally different. Let's say, geometry queries first. How to get ICE Get Closest Location, especially ID to location. I know there is some Shrink Wrap Canvas example, but I think it wouldn't hurt to go further - again, especially with something like ID to location - how to get some value from polygon back to point, or so.
ICE context, how Fabric deals with it - I've read somewhere that equivalent of ICE context is more flexibel in Fabric, that even character could be 'context' - I could be wrong here, but, still, I think this should be interesting.

2: a bit more complex samples, let's say two animated bones and one mesh cylinder for start, now how to get matrices, build some simple envelope from scratch, apply matrices to all that at the end.
Or, equivalent of Helge's ÍCE ''static instance'' - so, how to copy a polygon island several times, find center of each copy, apply some SRT or deformation - could be a nice example of several features.

3: multi dimensional arrays, some example of advantages.

It's my understanding that unless you want to hardcode it into a node with KL using PEX funtions, then all the "per point" stuff Has no canvas nodal equivalent just yet.

However, I'm so used to the execute ports and the get/set data workflow. It would be great to have a few example based tutorials comparing how you'd do something in ICE.

personal, i think Thiago Costa have a good way to teach. His tutorial is more friendly and artistic feeling ...

Thiagos ICE training was superb, because it explained the concepts of why as well as how.
For a newcomer to the idea of making your own tools, as I was, he did a fantastic job of explaining say, vectors in a way which caused me to no longer think of them as opaque mathematical concepts, but useful building blocks that could be used to do anything you could imagine; All the time illustrating this with increasingly exciting examples.

His training inspired me to try and teach what I had learnt to others.

These fabric tutorials are much appreciated and useful, but it would also be great to have some following the thought processes and logic behind the creation of actual tools, however simple they are.

Hey guys,

I know these are quite basic right now but we are covering ground that users inexperienced with ICE will need to know before we get to the more in depth videos. We hear you and understand where you're coming from. All I have to say is we're nearing the end of the basics topics we had established. I really want to show the workflows on Geometry Locations especially as I know very well how much they are used in production. Luckily Jerome has done an amazing job with the location system (just like he did on ICE).

Hang in there. Won't be long. The requested topics on this thread are going directly on our list of things to cover so keep posting ideas and questions!

Latest video on Subgraphs, Backdrops, and Comments:

This video showing off how to create, export and load presets. Also how to add custom preset locations via the Environment FABRIC_DFG_PATH variable.

Thanks very much. I'm absorbing these with interest and looking forward to the locations one.

It would be great to cover how to visualise stuff too. I find that invaluable in ICE.

By the way, is it possible to have a 'top level' nodal setup in FE Canvas, without 'ropes' ( connections to input and output), just like we have in ICE. Maybe I missed something, but at least in Canvas standalone, when I 'go up', it's only a window with parameters. However I think I sow the 'top level without ropes' in some Modo - Canvas example.

I'd like to understand how the equivalent of 'Execute' works in Fabric.

In ICE, I'm thinking almost in terms of lines of Code with an exectute at the end, where I Do this, Do that, then Set the variables, then go on to the next bit.

Canvas however appears to be one long stream.

Mathaeus wrote:By the way, is it possible to have a 'top level' nodal setup in FE Canvas, without 'ropes' ( connections to input and output), just like we have in ICE. Maybe I missed something, but at least in Canvas standalone, when I 'go up', it's only a window with parameters. However I think I sow the 'top level without ropes' in some Modo - Canvas example.

Hi Mathaeus,

Are you speaking about Fabric inside Softimage? If so, no you don't have a top view of the Canvas node as Softimage works differently than other DCC's like Maya and Modo (not speaking about ICE btw). In Maya you open the Node Editor and can see the top level Canvas node. It may or may not have connections in and out of it. I'm not sure about Modo as I haven't used it but I'm suspecting the same / similar.

In Softimage, Canvas / Splice is implemented as Operators so this is why the interface is different.

Pooby wrote:I'd like to understand how the equivalent of 'Execute' works in Fabric.

In ICE, I'm thinking almost in terms of lines of Code with an exectute at the end, where I Do this, Do that, then Set the variables, then go on to the next bit.

Canvas however appears to be one long stream.

I'm going to be writing up a blog post on the similarities and differences between Canvas and ICE next week. Stay tuned for it as it can explain many things people are asking about.

Canvas can work in many different ways and isn't structured like ICE was. You can have one long chain of nodes doing many many things interacting with data set on a previous node. You can also have many "streams" that are getting passed into an output port and the next stream picking up data via variables that were set in previous streams. It really depends on how you structure your graphs and it can be a bit more open of a workflow than ICE was.