I hope this is the right place for issues like this.

I've started using Houdini a few weeks ago and I'm running into features not working, and then after a few hours of trying and frustration suddenly some features start working as advertised.

Editing Capture weights for instance. After figuring how to get Capture Geometry to work finally, I try to edit the capture weights by using “Edit Capture Weights”

And no matter what I try I can't get it to work. I studied the manual, followed several video tutorials. It simply does not work.

I select some points, click “edit capture weights” nothing happens. I select no points, click “edit capture weights” Houdini tells me to “select some points” but I can't select anything.

Then I try “Paint Capture Layer”, doesn't work as well. But now suddenly “Edit Capture Weights” works flawless, just as advertised; I can select points, use the bone sliders to adjust the weight, brilliant.

I get this all the time, very frustrating.

Is this just the way Houdini works, could it be a hardware issue, should I try re-installing ?

Thanks,

R

yep viewer interaction can be pretty flaky… or just frustrating by design.

try to pin down a sequence of actions that lead to the problem, send to support.

Frustrating by design? That's bordering on conspiracy. Surely, nobody at SideFx gathers to discuss how to make some things more frustrating for some users.

The issue is that there seems to be no person or department designated to making the two main workflows - netview and viewer (naming which says a lot if I were to focus on semantics) - work seamlessly. On this particular issue, they seem to wing it, from one version to another, which falls on the devs' shoulders alone. They are some of the best in the industry, but they're not omniscient and can't anticipate or know what users of all stripes are wishing. It's clearly a management problem.

As cpb (and I on past occasions) has suggested, the best you can do right now is to present your problems to Support. If you feel like you need the community's support/feedback, beside the problem in question, post some solutions accompanied by images, scene files, etc, as well.
As a rant-queen myself, even I, most of the times, come up with suggestions to fix a problem followed by a RFE. Unfortunately, that's not a sure way to see your wishes come true, but it's all about increasing your chances, like with anything in life.

ok, thanks.

The thing is, as a newbie it's hard to know if this unpredictability is a bug or a user error. Until now it only becomes clear when something suddenly starts working even though I don't do anything different.

And at that time I can no longer reproduce the issue.

toonafish
I hope this is the right place for issues like this.

It's not.
There's a post stickied at the top of this forum that says READ THIS FIRST. It's got lots of handy info on how to report issues to us. Thanks!

chrism
There's a post stickied at the top of this forum that says READ THIS FIRST. It's got lots of handy info on how to report issues to us. Thanks!

Thanks, I read that. But like I said. I run into so much of this, I started to think it might be a hardware or user error.

But it almost sounds like it's just the exiting and magical world of Houdini, will it work or not ? ;-)

As a newbie myself stepping into intermediate, I can relate. But, until you're willing to put in the time to make a comprehensive post with your problem - with gifs, scene files, etc - you're going to keep being glanced over and ultimately end up frustrated and leave. People don't have time to try and understand your problems from your description alone, regardless of how willing they are to help and there are plenty of people around here willing to help.

Turning my eye to sidefx, I hope they're reading, the take home here is something I mentioned before - if you have users that are experienced in 3d, coming from other packages (and it's not like in a 3ds Max forums where you're dealing with questions from people that are just encountering basic 3d concepts for the first time) and they tell you things are not intuitive or working as if created for humans instead of AIs, then you should probably believe them and more importantly, take action.

From my experience heuristic evaluations would help solve this. It's a nice tool that guides development from a domain experts experience that results with the product being way better.

We are using then in our masters program and they rock for software design. Much more effective than endless forum posts or disparate bug reports.

It's a nice tool that guides development from a domain experts experience that results with the product being way better

links? Specifically before & after usage videos.

I have strong doubts about heuristic evaluations being any help when it comes to a software tool that is used for a vast broadness of workflows.
When I went to university, “heuristics” was considered “a scientific way of being almost completely wrong”.
While heuristic evaluation may be of some assistance where you can apply generic usability rules, they will surely fail when you are dealing with a highly “flexible” tool like Houdini (or Maya for that matter, if you consider Maya a user interface for a pipeline system). You just cannot induce a general flaw or, alternatively, “correct” behaviour from individual experiences - which “heuristic evaluation” basically is about.

As for “cbp”'s comment about “a video”: That was probably meant as a joke (“if there isn't a YT video about it, it does not exist”). Not funny, though, because many people today actually BELIEVE that only what has been documented on Youtube exists. A basic “common knowledge” of the world we live in OUTSIDE that “live-for-dummies” platform should not be neglected.

Marc

malbrecht
You just cannot induce a general flaw or, alternatively, “correct” behaviour from individual experiences - which “heuristic evaluation” basically is about.

That's an interesting viewpoint. From what I've seen is that most people use the tools in the same way, just creating different output. The tool should be functional, in a general sense.

@cpb sorry, you have to experience it yourself

> That's an interesting viewpoint.

My comment clearly was quite short for an in-depth discussion. I was mostly concerned about cpb's reaction that seemed completely out of place to me.

> The tool should be functional, in a general sense.

That *sounds* agreeable, yet, it might just prove my point: If you are talking about a tool that is built for a specific purpose, this claim holds true. If you are talking about a “tool set” (which I would consider Houdini to be), the claim is way too vague, because that “general sense” just isn't well defined.

In my experience, “heuristic evaluation” most often is used around web applications and/or single-task applications (not in the sense of “single-core”, but “specific for one task”). THERE I would say that - with the right set of general rules to check your experience against - heuristics can help if you are after QUICK AND DIRTY critics.
Again, with a “pipeline-oriented” toolset I personally would generally tend to reject any other person's “subjective experience” as being just that: Subjective.

Sure, Houdini suffers from fundamental design flaws (I am going out on a limb here to make a point): Using Qt for UI is one of the most basic mistakes one can make if you are after “good UXP” and one of the best proof of this claim is Houdini. Having dozens of different “must-use workflows” and virtually no consistency across toolsets *IS* a problem.
BUT … one has to remember that Houdini also tries to cater for old-time-users, carrying along a huge amount of outdated … let's call them “nodes”. I am not sure if making a brutal cut somewhere and “start from scratch” would even WORK for Houdini, even though this is precisely the feeling I get almost every time I work with it. I am trying to be fair here.

This is just to illustrate that I *think* that heuristics - that try to apply an arbitrary “general set of rules” - may not be the best approach to something that is by definition “not general”


Marc

Thanks Marc - you bring up some good points - the effectiveness of heuristics appears to be in the scope of the survey and the skills to find useful meaning in the data.

I guess the formal nature of a consistent questionnaire isn't something I've seen used in the DCC industry and thus development appears to be more ad hoc in nature. With all the strengths of Houdini, the UX is probably the area that could do with the largest overhaul.

In the end its not a silver bullet but it could be part of the UX repertoire. I'll certainly be trying to use it for my projects!

A toolset is simply a multitude of tools, usually that work together for achieving a goal. Maybe we're dealing with a small language barrier here, but I wouldn't hold it against us since most people communicating in their native language will also hit the semantics wall quite often. Except for, maybe Sanskrit speakers, of which there aren't many left. So let us not confuse a toolset with a platform that enables tool creation, Houdini being both.
Therefore, since thr heuristic approach can give you useful statistical info, if you take the time to xollect data, but cannot guarantee a precise picture nor an explicit method for approaching a certain problem, human input and interaction that goes both (or multiple) ways is crucial.