I'm giving this a quick tryout and so far I quite like what I see. On my 3y.o. iMac it renders modest models very rapidly - my whole-house model is taking quite a while though, unsurprisingly.

Two annoyances I noticed pretty quickly
a) once you've turned on 'render viewport' I can't see a way to turn it off. closing the model window seems to be the only option, followed by reopening. Not as friendly as it could be.
b) it seems to take an interesting view of what is visible and what is not. When rendering my house from a view supposedly only showing the layer for the timber frame the rendering was displaying many other layers but not all. Odd. If you would like to use it for testing I'll happily send you a copy of the model file.

Oh, and I just noticed that it leaves the sliders (used by opacity setting etc) lying around on the screen.

@slbaumgartner said:

I don't have an explanation for you problem, but it is extremely unlikely this is the cause. Computers don't take more time to do arithmetic based on the value of the numbers.

Sure they do. Just exceed the integer range of a typical CPU word-size and watch things take longer. Assuming of course you're using a sensible language that can handle arbitrary length integers. If not, it will take a really long time since you'll crash! And no, floating point is not a proper solution in many cases.

Lofting! That's the word I needed to remember... now what was I asking about?

Oh yeah; curviloft by Fredo. Does a lot of astonishing things and I might even be able to work out how to control which one it does for me
http://sketchucation.com/forums/viewtopic.php?p=248195#p248195

@dave r said:

Hi Tim,

Long time, no see. I hope you're doing well.

Hi Dave; yup, all ok over here. After the fun and excitement of building the house I got pulled into the Raspberry Pi world to develop and improve the Scratch program (anyone who has kids that are interested in tech things should at least look at http://www.raspberrypi.org) and pretty much haven't had time to do anything else for the last couple of years.

Now at last I have a little time to try to repair a large and beautiful model plane I crashed three years ago...

@dave r said:

So are you trying to take the skin (with thickness) off the wing structure? Are the ribs and spars already in there? Or is your goal to be able to create the shapes of the ribs and such? If it is the latter, what about slicing the wing shape--maybe TIG's Slicer plugin or manually placed cutting planes?

Good question.

What I have is an original dxf file of the design by Fred Guillfoyle that I imported and started working on. I noticed that when I 'assembled' the wing ribs they didn't seem to fit terribly well, as if the import had just been a bit vague.

So, I used a nice little importer (https://github.com/elliottt/sketchup-airfoil) to read in the standard coordinates of a the ClarkY airfoil it uses (no jokes about the choice please, not my decision) and compared the rib shapes to that. It took a while before I realised that I had to scale the airfoil to the full outside dimension of the wing and then offset inwards by the sheeting thickness, but the ribs are reasonably close once you do that.

Clearly, at this point I could just accept the data and move on but where would the fun be in doing that? So, I started wondering how to derive wing rib shapes in SU, given a root and tip profile. Note that the profiles may well be fairly different in many wings - we often use a lower camber tip profile to reduce the likelihood of tip stalling.

I can see two basic ways of doing the job (and an opportunity for an extension to automate!).
a) (probably only good when the root & tip profiles are the same even if scaled) extrude from the full root profile, select the tip face, scale it down. Use a tool that does something like jointpushpull to offset the surfaces inwards by the thickness of the desired skin - typically anything from 1/32 to 3/32. Then place 'sheets' of balsa where the ribs need to go, intersect and remove the unneeded stuff. Add slots for spars etc and off we go.
b) place the root and tip profiles where needed. Offset inwards on each one by the skin thickness. Remove the outer profile. Use a tool that will skin over them, hopefully coping with an shape changes due to different profiles. Then place the ribs, intersect etc as above.
I guess there is a third way where the outer profile is made solid, the ribs placed and cut and the offset done to each individual rib.

I'm sure I've seen one or more tools to do this kind of thing but simply can't remember what they were called even well enough to get any sense from searching! It wouldn't surprise me at all to find that some modeller had written an extension to do all this and lay out the result ready for cmc or laser cutting.

I've tried searching for tools to help with this problem but since I'm not finding much I suspect I'm simply missing relevant search terms. Feel free to suggest some.

What I need to do is go from a wing section (which I can import easily enough with a nice simple importer I found) to a tapered wing (easy enough with push/pull and scaling of the smaller end) and then I have to 'peel off' the skin to leave the shape the assorted ribs have to provide.

This is a shot of the smaller end showing an outer skin overlaying one attempt at the structure. You can see the thickness I need to 'peel off'.

Simply using offset on the larger end before extruding doesn't work because the scaling also affects the thickness of the 'skin'. I tried making the smaller section by scaling and then using offset to derive the 'inside' shape, which works ok. Then the problem is how to generate a solid with the large section at one end and the small section at the other.

Another option would be to make the outer shape and use something like jointpushpull to shrink it by the skin thickness; but so far I can't make it work on a wing shaped solid.

What am I missing? I bet there's a tool or simple technique that will do the job...

@d12dozr said:

...and then you have this guy smoking a Ferrari with his bicycle!

Hey! I rode the circuit Paul Ricard back in '78 during there Bol D'Or. Not at 207mph though...

Minor additional info - my 'old' LO file for the house model loads does whatever conversion is needed and appears to be fine. Updating the model - which seems to load the 'new' version of the house model as saved by SU14 yesterday - still looks good.

Very odd.

I'm having a possibly related issue from the looks of it. My house model ( which got started a scary five years ago and still needs some details fixing to reflect the actual building) opens is su14 ok but if sent to lo14 many layers are unseen. In some scenes it looks like most or even all are there, in others virtually nothing shows.
In a simpler model I noticed that only items on layer 0 (that is entire components - all my geometry is on layer 0 of course) were shown. After deleting all other layers to force everything to 0 the entire model displayed ok.

Haven't tried opening my 'old' LO file yet.

Hope it doesn't afflict everyone because I dread to imagine the pain it would cause people like nick sonder!

Any chance you could do it in LO? If you exploder the vectors and then add some lines to complete the shadow area, then use LO hatching...

@unknownuser said:

@solo said:

My eldest is busy doing the application process for West point and it seems he will have to say he is of some Christian denomination as the recruiter advised that his chances of entering will to slim to none if he says he is non religious regardless of his almost perfect SAT scores and other requirements.

That's ... weird.

Actually I'm fairly sure it's a violation of the US constitution, which makes it rather more serious than simply 'weird'. There has been a long running scandal at the airforce academy in Colorado along similar lines. Not a few of my friends and colleagues that are alumni have become very disturbed enough that several of the colonels and above have sent pointed notes.