Hi,
Can FlexPainter still be used with QCAD 3.29.4.0? I briefly played with it probably over a year ago. It is now greyed out under the
Misc/Script Examples/Mathematics/… menu.
I am looking for an easy method of creating revision clouds around changes in my drawings.
Most likely yes, it runs just fine under QCAD 3.29.4.0 on Windows.
And probably yes for MacOs using QT6 there I see a commit to the code by Andrew regarding that.
Remark that tools that require a selecting are grayed out without an existent selection.
FlexPainter mimics a roller stamp and requires a base entity as a trajectory to follow.
One can delete the base entity(-ies) in one go with an non-persistent option in the dialog.
I just thought it would be a valuable tool…
Without … Meaning: ‘NOT availible’, ‘NOT functional’, ‘Not the intended result’ or avoiding by preference?
Draw a polyline shape (PL) around the intended region with several segments.
Each node/vertex intended as a point where 2 so called ‘arcs’ meet.
(A bulging segment is only closely related to an Arc shape )
Use the ‘Change Segment Type’ (OX) method to alter each straight segment so that it bulges outwards or inwards.
OR
By changing each bulge factor to something else than zero in the Property Editor when the Polyline is selected.
To 1.0 for example to resemble a 180°Arc, 0.668 (~135°), 0.4142 (~90°), …
Positive for left, negative for right or respectively outwards/inwards when CW or the inverse when CWW.
Wow, that’s awesome. I’m now using exactly that method for revision clouds. It’s fast, and I can make the cloud puffy at exactly the right places without needing to mess with a scale, as with flexpainter. Thank you!
Reply by mail (All access has ceased as of early Juli 2, 2026) As all on one line then … @ iOne_iTwo: I would not call this approach fast with many segments. I further never understood your major problem with a so said ‘Scale’. The value to specify is the ‘Width’ in drawing units, as for the traverse size of a roller stamp with a pattern. The programmed size is automatically adapted, scaled uniformly or non-uniform when not proportional. It also displays the length and stepsize for the chosen pattern and specified width. Depending the cloud type about 10 to 20 units long for a width of 5 units. All this is very similar as specifying the text height in units instead of specifying the 7-based scale for the glyph size. The painter sets are also 7-based … Although not limited to 7 high like some TTF fonts. With the difference that how long a text like ‘ABC’ would be then fully depends on the font type. You can only guess how long that would be. You probably don’t use a text size of 5 mm on a housing project in real size. To be readable, to make sense, the text height (or traverse width) is most likely adapted to the overall size . Then don’t use a 5mm thick painter pattern just the same. A real example then: Given an ellipse of 200 by 100 mm, painter ‘RevisionCloud_L’, Width 10mm. The ellipse is about 484.5 mm long and about 21.8 bulges spanning about 22.2222.. mm may fit the entire shape. For a closed shape it dynamically scales the painter by a factor (0.94776..) to fit 23 bulges. Most obviously it would fit 219 bulges for a Width of 1mm (484.4224.. / 2.2222.. / 0.9953645..) There is nothing wrong with that only when that was the intention. You don’t have to mess with a scale, just specify how ‘Thick’ or ‘Wide’ you want it. As last, to my knowledge, FlexPainter is the only QCAD tool that adapts its options for a unit change. Regards, CVH
Oh well, why not.
The ellipse example is somewhat incorrect.
On QCAD the length of a full ellipse is based on the Padé 3/3 approximation.
Already a major improvement implemented on my behalf compared with the former used Ramanujan approximation.
For a limited ellipse, QCAD uses the Simpson rule over a fixed number of 20 intervals, whereas this should be an adaptive number.
Then the sum of four quarter ellipses is an even worse approximation.
The major flaw sits in the length of a half ellipse, again over not more than 20 intervals for twice the length.
To be utterly correct one should implement elliptic integrals like I did for personal use.
Then the length of a 200 by 100 ellipse is 4aE(k²) = 484.42241102738376 in full floating point.
As correct as possible for every absolute size and ellipse ratio.
)