Trim surface problems continues...

Hmm, I'm confused then because if you helped with the extend, that has nothing to do with trimmed surfaces? But anyhow, yes, the surface colors are difficult to decipher, especially when they are insided of the surfaces and identifing what is selected/included. The visual clues suck. The new light brown intersection lines help but not enough. So, yes, the isoparms or trimmed mesh boundary would help knowing what faces are trimmed.

I don't think it surpasses Pro/e, I would say that is subjective and the new optons address more in how someone may deal with adding/removing trims. More options are nicer but the visualization and managment of the trims are by far more important. As for the selection/change and delete, that has been around for a while.

..

Mark M. - regarding your comments on Mark B's thoughts on surfaces, I had interpreted this differently: Like my car, I want it to get me from A to B, but I really do not care to much about how it does it. But it has to do it right! Likewise for CAD, I want to achieve a certain form while behind the scene it is accurate.

The issues you point out about poor quality models going into production and being changed in tooling is close to my heart. This is one reason, as an Industrial Designer, I use SW. The last thing I want is the design my client signs off on being changed by someone who has no idea why is was designed this way in the first place! Wrong assumptions get made even with the best intentions.

However, from my experience it is too easy to make errors using surface tools like StudioTools, Solidthinking, or Rhino. I see this all the time when I am transferring surfaces in and out of ST and SW. However, I know many industrial designers who would prefer to use SW if only it had better surface creation tools (i.e. Nurbs curve tools) or as Mark B. suggests another paradigm for working with surfaces / forms that is fast, intuitive, but yields precise solids at the end.

I am disappointed 2004 does not have some of this, but at the same time, I am very happy about many of the other features that have been added. And some of the new surface tools are far more powerful than are available in StudioTools (some saved my butt on a current project). I just keep emailing SW with my wish list until I get my way. maybe they will add cool surface and curve tools before 2005 :-)

Cheers, Daniel

Mark Mossberg wrote on 16.9.2003 6:52

Mark, without going back to that thread, my memory of it was that I was trying to make the point that in consumer product design, the surface, in most cases, is inconsequential as long as the boundaries of that surface/s (design intent) is held accurately and the the surface is "behaving" in a normal manner. Currently, the problem that we face with SWX in Lofts, Sweeps and fills is NOT having the surface behave in a logical and consistent manner. Obviously, the other requirement is that the surface fit functionally around the packaged components.

Mark, most of the work I do in SWX and ProE is released tooling data, and we have 5 Fadals 50 yards from my desk, so I know your concern from my comments, but again, I was trying to make a point that unless you're in the aerospace or automotive industry, advanced modeling of consumer type products: surfaces are highly subjective and are and should be driven by the Industrial Designer.

Mark Biasotti wrote on 16.9.2003 19:15

When I first read that, I read "infadels"... And I thought you were slandering the industrial designers.... LMAO when I realized I read it wrong!!! :-)

Have you tried customizing the default colors in SWX for temporary shaded, hightlight etc.?

The mesh in Pro is nice, but I've gotten lost doing trim (merges) in ProE also, so I don't know if just putting a mesh on there is the answer, and option: yes!

Paul, if you have done so already, send an enhancement request on this and to make the point really drive home, attach a screen shot to show just how confusing it is.

Mark

Sometimes adding a piece of geometry, like a sphere, at an obvious location temporarily can help focus the mind IF I have the right problem in mind and you can do it ....

Mark,

The colors are part of the problem and yes I have played with different colors. The problem is the SW gui in managing the display during the trim or mutual trim, it fails to show the colors properly or show the colors at all.

Yes, the mesh would be welcome and it is a step above what is now. And yes it is not perfect and I agree, in Pro/e you can get confused at times but you can easiliy add/remove faces clarify or help with the visual boundaries, it is not a problem.

I know they have Pro/e in house so why should we have to show them? I can only imagine they, the developers and inhouse users, are NOT users? If the are SERIOUS about user satisfaction and about creating a gui which is user friendly, and with users requesting this interface to be enhanced and easy to use as well as referencing Pro/e (as I have) why would it have take this long for them figure it out??????

Why should I or you show them what is so obvious??

You see, they are getting paid for this, correct??? We the users are paying them to do this, correct??

The math is simple, they make a hell of a lot of money!

IDEO is a company which consults on gui, and I can only guess they have had contracts with SW Corp on user interface for SW??

SW Corp has plenty of cash to solve this problem so what does that say????

Think about it???

..

surface colors are difficult to decipher,

Paul, actually they do not have a single copy of Proe. My understanding, and I could be wrong about this, is because of legal issues started back with SWX95, they have a legally binding agreement with PTC to not have a copy of it on site and I think they have honored that. In the many times I've been there over the last 6 years, I have never seen it. (but that doesn't mean that many of them are not familiar with it and it's latest features).

Paul, again, I must respectfully disagree. A good number of them could run circles around you and I when it comes to modeling. Don't mistake there silence as if they don't know or don't care. They are aware of

99.9% of the bugs and issues you and I and others have raised. They are bound, on some of these issues, by D-cubed and Parasolid, and some they're not, but they do want to fix them and I believe they will.

The best thing we can do as users is to send them well documented parts etc. I know one thing that works for them is to put annonations inside the part pointing to the issue/s so it clear what to try. Also, screen shots in you html email body are helpful.

I could say a lot more, but I don't want to jeopardize my relationship with them (the manager of product definition is probably looking at this thread right now and cringing). So this is the last I'm going to say on this subject.

I want you and others to know that for the last 7 years I have been a "thorn in the flesh" in the side of SWX development pushing on them in the area of advanced modeling features. But at the same time, I want to see the product improved because I have invested my time and resources into it like many of you. Could I walk away from it tomorrow, and just use WildFire? You bet I could! Especially since WF is a better hybrid surface solidmodeler. The reason that I have not is because our useage of SWX in IDEO, in the last 2 years, as outpaced Pro and what this really means is that more and more of our clients are using it. In the overall picture, there are other issues with PTC that I don't have time to elaborate on, but you're probably well aware of them.

Back to the issue of visualization with trim, please send some screen shots to them either via support or the enhancement page. Its not going to get fix unless they're aware of the problem. One thought I have about the problem; would having a dynamic clipping plane for getting to interior surfaces to trim solve the problem? Another thought would be to use more than 2 colors. Say for instance, that the trim surfaces were assigned different colors and then there face names in the trim surface property manager had little color blocks next to the surfaces. That way you would have both the screen and the PM to select and deselect and would be more implicit about the selection.

Ah, I don't think they or PTC would ever disclose what software they have.

I respectfully disagree as well because if you want to put it into different words.. Users which can model fast or who have more knowledge of the inner workings which can model faster? Or users which can apply design and model fast and create product?

Who are they??

Oh, I believe they will fix things, with some free help, over time.

Yeah, for free.

Ahm, well ya know, you have not said anything really. Except, you do represent IDEO, and since this is most likely on IDEO's time, you more at risk with IDEO, no? Is you contract per yourself and SW Corp, or as a IDEO employee?

I'm sorry but I can not believe for a moment that the time you have invested has not been reinbursed or covered by SW Corp or by IDEO. This seperates most of use from you in helping. BTW, WildFire is NOT a hybrid modeler. And, unless I am totally wroing, you do not have it WF or SW, IDEO licenses them for you to use.

Well, SW clients are cheaper = cheaper SW users = SW work > than Pro/e work in this economy. Yeah, well, PTC is still screwing up and it's sad. (the root of it is because of $ and arrogance but still the BETTER tool!)

Sorry man, I actually feel my money relative to the 250K users involved has not been well spent...

Anyhow, here's the big trim hope thrown to the wind....

Simple is best so, if the boundaries can be clearly recognized, that is the priority. Editing and redisplaying the newly formed boundary via a preview is also a priority. NO graphic degradation or loss of visual clues will be tolerated!

An idea, (doesn't take too much to think about it) since they have texture mapping, why not apply that to the trimmed surfaces as well as transparency and enhanced the boundary edges? Z clipping may help but I think what is more important is having some preview which clearly shows the boundaries and if that preview can toggle with a Z clip, sure. Your last suggestion is interesting and reminds me of the other Pro/e trim arrows for trim sides/direction!? So, yeh, if the visual clues help to identify and enhance boundary and preview as well as editing the boundary with no graphically inconsistences, hell yea!

The concern I would have with SW graphics now is they are not consistent and degrade so would these wants clutter the interface or the visual clues as well as over burdened the already non-robust SW interface and the graphics fail the way the do now!? Seriously tho, SW graphics are not robust enough now with trims alone, the graphics faulter when making changes to the trim select list as it is... So, I can only imagine this getting worse when adding more visual clues and of course I hope like anyone it gets better but over the years it clearly is NOT consistent.

..

Mark,

In that context, I agree. If the surface can be "defined" in a looser free form way, it can still accurately represent what it actually is. It would just be easier to make it smooth and flowing. You still have to be careful though, about loosening internal tolerances to facillitate solutions. These "tolerant" solutions constitute most of toolpathing problems I see now, especially with some fillet cases.

To drift a little off subject, and since I have your ear (and you seem to have theirs), why is it, really, that SW keeps getting slower with each release? Is it that they don't see it as a problem ? Since SW2000, if I wanted to maintain the same relative performance, I would have had to upgrade my hardware for each release. This seems very unreasonable, and is completely contrary to the concept of "development". Do they ever talk about this ? Have you asked them and gotten a face to face response ? Sorry, just had to ask.

Regards

Mark

To me it looks like what appears to be an inordinate reliance on D-Cubed is the major stumbling block for SW2004 and one of the major reasons why surfacing is not better.

I believe this is the kind of functionality that is very badly needed in *3D DCM* not just 2D DCM.

think3 does not have what I label an inordinate reliance on D-Cubed. Instead think3 has someone who IMO SolidWorks does not have anyone close to:

I don't believe that many of these kind of people exist or can be easily hired.

Pro/E Wildfire has nothing even close to the functionality one gets with thinkshape Global Shape Modeling. It's a good start though.

jon

I don't know the answer to this other than to speculate that as the program inherits more functionality, it is going to proportionally have more "overhead" to manage. Its important to specify where the program is slower. To build a cube in 2004 verses SWX95 is it slower? I think not, but you're probably referring to larger assemblies and non-analytical modeling (surfaces) right? "Feature Statistics" should be a pretty good indicator of what's slowing down the rebuild. I am not liking the slow down anymore than you. But if SWX were to ask me where its slower, I'm not sure I'd be able to tell them. So, I take it that this discussion is a general level for debate. You got me thinking thou and over the following weeks I'm going to keep and eye on it.

What makes it hard is that every 2 years I get a faster machine and the overlap of successive revs of SWX make the issue a blurr.