Hi! Sorry for late reply.

At the moment there is only crosshatching with edges. ver 1.6.

The pictures you have seen is probably screen-grabs from developpment.

I just want to assure you I have not abandoned the project. In the future those features will be available, if all goes well.

Thanks.

HOLY macaroni!

Been away from Sketchucation and all that has todo with a computer for a while.
And now I see this

Got some reading to do, 50 + pages.

I wonder how this will affect Ruby API etc
I'm worried..

@unknownuser said:

I thought you were saying we could try your plugin if we wanted to risk it...

If you are referring to Hatchfaces? I think I said I might release a lighter version without some of the experimental features to speed up the release-date. If it sounded like something else, I apologize. It will come, it will come

The point I was trying to make regarding this topic is that "Fill a shaped surface with a modeled object" is fully doable. But it will be very compute-intense, in fact so slow that it might not be usable unless dealing with sparse geometry.

I have been playing with that function. But it's shaky stuff

I think this issue has been discussed before by TIG and others.

When you say fit onto a shape, do you mean surface or single face?

If it's a single face you are working with, then I have been experimenting with a tiling-feature in "Hatchfaces" that will allow you to fit 3d components to a single face(like the picture of your roof tiles). It works "ok", and I do think of implementing it into the plugin whenever I can get the time to work on the damned thing (current house renovation takes all my time )

If you don't feel like waiting, anyone trying to turn this into a plugin better be aware that it may not be a very practical feature. If the tiling geometry is dense the computation will be VERY slow. Also doing booleans it's a bit tricky to get the face normals pointing in the right direction sometimes.
But If geometry is kept simple it can be functionable. And still better than modeling it by hand.

I think the problems regarding a plugin like this has been discussed before by TIG and others...

Ah, see what you mean, Z-fighting.
I plan to have background colors/textures available in combination with edges(vector lines as you call it).

So, in my code I use the glue-to face-function. Appears to be working pretty ok for me.
Have you had issues with glue-behavior?

Can't give you any feedback on Open GL programming. Way over my head, sorry

Could you elaborate a little bit more, please? Function .

I have a hunch what you are asking for but I don't think it's doable in Ruby with current API for what I know. I could be wrong.

BTW. Sorry ALL, for the delay in developping this plugin and releasing updates. Have been very busy with little or no time left for programming. But will be back on track soon.

@unknownuser said:

Just 'patterns,' which are in fact component instances which can then all be changed by editing one of them

Brookefox, I'm afraid that will not be possible if you mean altering them after the hatch-components are fitted to the face.
What will happend(in the code) is that all tiled instances will be exploded before doing an intersection with the face. And in the end become 1 large component.

I think Smidths plugins was an request to solve a particular problem, spacing out plants and stuff, including triangular array. In those cases one may not wish to delete geometry outside the faces. Works very well for what it was designed for, I think.

Hatchfaces plugin obviously will be more "pattern-fit a face" sort of thing.

Thanks for answering.

Thanks for the input and ideas Brookefox

Gridded components you say? Do you mean patterns or using some sort of workplane?

You will be able to load premade patterns(components) from folders. Where you also can make your own patterns and stuff them in there.

Anyway, The way the plugin is working now, is that a face-edges determine the boundarys. That way the user don't have to worry about which plane one is working on.
Everything outside the face will be trimmed, including holes. Just like it's working in the current downloadable version.

The plugin as is now, is not beta ready . I actually might do a "light" version for user testing at first. Cause there are so many things added it might be difficult for me to keep track of things when complaints starts to hail in

Hello everyone!

I AM working on the plugin. Just have been very busy lately with house renovation(I do this SU-programming on spare time).

Brookefox. The downloadable version you have, only does linehatches at the moment.
Although I've earlier posted pictures of current development features (here and in LO-forums)in order to get feedback and suggestions.

Must admit, for a while I got the impression most people where content with using textures for hatching(At least at the LO forums). So I got a little side-tracked
Started to work on other features, like 3d hatching and such..

Thanks for the feedback guys Really apreciate it.

Will be back on track soon.

/ Joel.

Hey! There you go

Did not notice that plugin cause you have it on your blogg, Sdmitch. Sorry

Indeed seams to be able to meet the requirements in this request.

Hi!

I understand your request. However to select each face and run a specialized plugin (in your case 1000 times) will be quite slow as well

Hatchfaces will run on only 1 face, but you will be able to do some tiling.

Doing a plugin for your request would only need a pushpull(in face-normal direction) and some scaling with a simple dialog. That way multiple faces could be altered in 1 go.
Not to difficult to do, I think?

I'm swamped with things to do so I can't help you right now. But maybe later if nobody else is interested.

TIG. My guess what Deaneau want's to do is removing "inside faces"(red faces on pic) on non-solid geometry. Am I Wrong?

Which can be a difficult feat..

Hi!

I will have some paneling features in "Hatchfaces"-plugin, soonish. Been busy with other stuff.

Back to subject.. Does sounds like you only want to pushpull each face and scale it down a little to get an beveled edge? Not really paneling, then?

Anyway, did some experimenting with pushpulls. And there could be issues of getting hold of the original face when pushpulled?(if I remembered correctly). To fix that I assigned a temp material to the face (so I could get a hold of it later on).
That material can be discarded later on.

Just some observations for anyone involved doing this plugins.

Thank you Richard!

TIG is right. It's not really beneficial for a fast workflow.

In "Hatchfaces" I'm working on 3d tiling and fitting on face.(thats why the plugin is delayed)
Enthusiastic as I was I got this feature working quite fast, but pretty soon realized there
plenty of problems to deal with. Solid subtract is PRO only, and awfully slow? Reason for that is probably cause it does a good job at getting the face normals in correct order when intersecting. Which I'm struggling with doing standard intersection with a dummy group.. Nuff details..

This kind of feature should be taken outside Ruby really, to be usefull.
But thats a different ballgame in which I can't play yet.

This simple roof on the pic took about 14 sec(original on the right). So you can imagine if the geometry is more complex and roof is larger.....

Nevertheless this feature will hopefully be included if I can make it work a little better.
But if you pay TIG to do a plugin it would probably be more efficient and proffesional. Maybe make a C-extension of it
/Joel

Heh Yes that's the one I meant.

Thank you Richard.

@unknownuser said:

By CD do you mean Construction Document? Look at the thread at the top of the Layout section.

Yes, I was meaning Construction Document. I see your point, this thread might be heading a bit towards the topic you mention. But I think you Sonder and pbacot has shown some nice workflows.

pbacot wrote

@unknownuser said:

Let us say the model helps speed by generating all at once--but when you get to details, my fear is that I might have too much more to deal with than in CAD, time being a factor. The patterns must read in grayscale. You don't have traditional graphic symbols to help if a texture is not explicit. Darker images can cause conflict with text, leaders, and line work. You must draw and texture a 3d object instead of (say, for rebar) a dashed line or a dot.

Is this something that would work? My view of a theoretical workflow would be do sections with TIG's sectioncut. Then use my(upcoming) plugin to do the hatching/details. For ex the insulation you can name desired number of rows in width. And scale it in the other direction.
When done swith over to LO for additional detailing.
Woulden't that be a decent workflow eliminating the need for CAD(of choice)?
The lines renders great when printing from LO. And can be exported or hidden as well..

PS: I'm not trying to promote my plugin, just want some ideas for whats needed in conjuction with LO to satisfy timnors and others request.

pbcaot

@unknownuser said:

Although I tried to keep the resolution up, eventually it looks weak

Yes, that is what I noted as well.

I think we are discussing 2 different matters here. I think Sonder is describing more architectual/design style hatches, whereas there is also CAD-style line based patterns.
The latest mainly grayscale if any color.

However I don't think one can deliver for ex a CD with such vibrant colors?
Correct me if I'm wrong.

Interesting topic. Please keep discussing.

Well, Sonders approach is the way to go if you like to work with textures. I've done it like that, before as well.

In my personal taste I found it to give me not acceptable results when printing. That's why I'm developping the plugin. Maybe it's not such a good idea if people prefers to work with textures anyway?

Anyway the main pitfall I've found with textures is that you have to have perfect scaling of them to get good results. And I prefer vector to raster mode in LO.
On the other side, textures will be ligther on the file size, than using hatched edges. A complicated hatchpattern can contain quite a large amount of edges.

I think that Australian guy Richard has come up with a clipping mask technique for LO as well.

You can create your own edge patterns and do an intersection manually already, with native tools in SU. But not with colors. So it's easy to do some tests for printing, Timnor.

I'm open for discussion about the subject.

FYI I'm working on a plugin that in the future will do just what you are asking for(hopefully ).

http://forums.sketchucation.com/viewtopic.php?t=38637

(Current version only has linehatching.)

Can't say when it's ready. But I've been working very hard on the plugin last months, and things are progressing.

Be sure to check the link, I think I've posted some screenshoots from the plugin.(some old pics, got some new stuff not shown..)

Feedback + wishes is always welcome. Especially from someone doing civil/ structural work.

BTW it is and will remain a free plugin...