How to combine entities to a new outline? [SOLVED]

[CVH]

Offset off multi selections migt be troublesome.

Clipping or not may differ a lot.
There are 4 methods:

• Poly offset OQ
  Selection offset OF, OH
  Misc/Draw/Poly Alternative

It might that an other method has better outcomes

Like I said, do this a lot.
Posted on it a lot.
A few problems I could let resolve already.
Complaining is easy, isolating the problem is not.
eg. See:
https://www.ribbonsoft.com/bugtracker/i ... sk_id=1948
and related forum topics

Nobody else seems to have troubles with it!!!..???
Or its called expected behaviour.
Proper offset are a base for good CAM outputs offsetting!!?

Divide your problem up.
Offsets
Combine Offsets at your discretion.

Cheers,
CVH

[Husky]

3 polygons are selected but only 2 want to offset.

Only 2? Is that the preview or the final result of the use of the Offset tool? Please post the 3 polygon example. Thanks.

[CVH]

Final Result, you see the selected and the resulting in both screenshots.
I don't need these screenshots to know they could be off.
Remember:
viewtopic.php?f=32&t=6632&p=25047&hilit=offset#p25051

Cheers,
CVH

[plaatjesprater]

I'm sorry to replay so late, I was away for a few days. Today I copied the file to make it ready to post here. The original drwawing is altered, and now the problem disppeared. (?!?) So I can't reproduce it. Next time I will post the file immidiately after a problem occurs, this is strange and frustrating, the screenshots were made of results and not of preview. I tested it several times before making my post. But anyway I can't reproduce it...

Thanks for your reactions!

[plaatjesprater]

Last post the unexpected results were not easy to reproduce. Now I got a new file with the same kind of results as the first.

In both cases my design was made with a lot of elipses. The advice of CVH was to convert the ellipses into polylines. The collection polylines was selected to make an offset as a workaround for boolean functions. The idea was first to make a offset and then reduce it with the same value so a contour will be made.

CVH made allready a bug report so this file can be used as illustration for the bug. A further description is made within the file. Personally I see this surely as unwanted behaviour. No contour is made, there is no usable result at this moment.
The offset method used was OF.

This is a hobby-project and I finished the contour in Inkscape as before, but of course this takes much more time and it would be nice to do this in Qcad.

[CVH]

Hi,
We know it fails, I posted a lot on this.

I am used to do these:
First divided your problem up in pieces!

With a adapted strategie for ellipse arcs...what are not closed forms on their own...
Its a few clicks away from halfway.

strange results for offset2_rev.dxf

(471.96 KiB) Downloaded 399 times

org=on
set a on
set b on
clear org
taw c
set c on
set d on; freeze c
set mirrors on
set org on

Regards,
CVH

[CVH]

In the end you 'll need to trim bulging polys to bulging polys.
Be aware you have to indicate the segments that cross not the polys 'somewhere'.
Also Reported...

CVH

[plaatjesprater]

I really appriciate your suggestions but I don't know what you mean. I understand the idea of dividing the problem, but I don't see what you mean with the description you made. Also the file doesn't clear this. Can you please be more specific? What command are your notes for?

I have viewed the file, but there is still no outline as I was trying to construct. I see you centered the file and deleted some lines and made it ready to start again.

[plaatjesprater]

Now I see these are the layers you made, that's the first clue. Can you give me a little more detail about the steps you made with the layers? I don't see how to achieve these results.

[Husky]

Just curious - why is this topic marked as [solved]?

[CVH]

Everyone went to sleep ... had some time to spare....

We Know:
A: Ellipse are no simple entities. Darwing them is one, doing math on them is harder as you might think.
B: Offset is not always flawless in Qcad. -> Countless reports and discussions.
C: Offset of a multi selection sometimes end premature -> bugreport.
D: The Booleans you mention are no easy task to code.
They may look simple in one case but universally it is something else.

That is what we have to work with untill things get better.
Every bugreport can be a step forward.


As in your former design I see some symmetry what also means that we don't have to stretch it to the limits.
So I divided your problem up in pieces.
Pitty I did not have the ellipses. They were exploded in 32 arcs.
The accuracy here is size relative and in this size the 32 is rather coarse.

Second problem is that some of them aren't closed forms.
Qcad will try to offset closed forms with a closed form, it will not necessarily do that for an open form.

So I copied the entities of your second challenge and put them in a new document (Layer _Org)
There is no reason why or why not dead center.
I defined a cropper as guideline (Layer Cropper)

Duplicated Layer '_Org' (YY) to a new layer 'b Restant ...'
Hint: Always try to keep your original entities aside... there is no way back from exploding splines or ellipses.
Selected all on layer 'b Restant...' (Y+)
Activated Crop To Rectangle (CLR + clip to Area) and pointed at the corners of the 'Cropper'.

Aside from cropping them the polys get exploded too.
Recombined with Polyline from selection (OC).

Added an offset layer 'c Offsets' (YA) what becomes active on creation.
Tried some offsets (OQ + Whole, Round, Clip) followed by undo just to see what was comming....
Hint: The larger te offset value, the less details that will remain, 5 is huge.
Rule off thumb is at least 3 times smaller then the smallest detail needed.

You can find the two indication points that worked best in a sublayer.
Left one first and then its obvious to select a bulge that wasn't offsetted.
Isolated the bad offsets in an other sublayer.
Seeing it is believing it: Bad offsets.

Selected the outer strings, duplicate (DP), added layer 'd Isolate...' and changed the 'General propeties' of the selection to that.
(That's just my way to do this in one go)
Recombined the outer offset strings with Polyline from selection (OC).

Mirroring this and I am halfway in creating the offset outward.
The bottom half can be done in the same way.
After some trimming and recombining to one outer offset the last instruction will be a single offest inwards.
(OQ + Whole, Miter, Clip)

Like I said:

It's a long way around.

And because of the poor approximation:

Very bad result too

But in the end I will get there.

And if all fails there is still 'Break out segment' (D2) and lots and lots of clicking.
Use it near the outer boundary and In a 100 clicks or so ...

Regards,
CVH

[plaatjesprater]

This is great! I'm really happy with this, because this is a repeating task for me. At this moment I haven't got the time to dive into it, hopefully I can do that in the weekend. Thanks a lot to CVH!

@Husky: I tought it was more or less finished then I found out it wasn't. I changed the title on 19/1 but the next answer it got again the [solved]-quote. I don't know why this occured. Logically it would again be: [partly solved with bug report].

Anyway the post of CVH is almost a tutorial, maybe we can make one for these tasks. In fact I do prefer a few guides for these complex tasks instead of a over-complete software with so much functions you always need the manual.