Forum Replies Created
-
Posts
-
Signs of all the array cells for labels in four quadrants:
Upper right [ 1 ][ 1 ] 1 () [ 1 ][ 2 ] 1 () [ 2 ][ 1 ] -1 () [ 2 ][ 2 ] 0 () [ 3 ][ 1 ] -1 () [ 3 ][ 2 ] -1 () Lower right [ 1 ][ 1 ] 1 () [ 1 ][ 2 ] 1 () [ 2 ][ 1 ] -1 () [ 2 ][ 2 ] 0 () [ 3 ][ 1 ] -1 () [ 3 ][ 2 ] 1 () Upper left [ 1 ][ 1 ] 1 () [ 1 ][ 2 ] 1 () [ 2 ][ 1 ] -1 () [ 2 ][ 2 ] 1 () [ 3 ][ 1 ] -1 () [ 3 ][ 2 ] 1 () Lower left [ 1 ][ 1 ] 1 () [ 1 ][ 2 ] 1 () [ 2 ][ 1 ] -1 () [ 2 ][ 2 ] 1 () [ 3 ][ 1 ] -1 () [ 3 ][ 2 ] -1 ()James M
Hi Csilla, thanks for the reply, that is the sort of thing I am looking for. But I’m not seeing that result from looking at [2][1]. It seems to always be negative. The only consistent sign change I can see between right and left sides is at [2][2] – it is always zero for labels on the right, and always non-zero for the left. But unfortunately, the non-zero values are extremely tiny (e-16 or 17!) and it seems strange to evaluate for such a difference with zero. Or, should I go ahead and do
IF ABS(connectionSide-0)<0.000000000000000001 THEN...Thanks again.James M
Yes, I have lately seen this regularly though not always, including AC18. I haven’t observed a pattern, but I have seen the behavior in a fresh object consisting of not much more than a single editing hotspot. Geoff’s attachment is a good, simple example of what seems to be a bug.
James M
+1 to TUBE renovation. Working around the twisting is very cumbersome, and it’s surprising that the angle argument doesn’t help it. When you renovate TUBE, take a look at how a cross section with sharp corners renders in OpenGL. It seems to assume that TUBEs are generally curvy. To get around this, I have resorted to a chain of mitered PRISM_s. The trouble with that is the contour of the curve becomes invisible. I’d like to see something like PRISMTUBE, where we have PRISM_ type status codes, but it can follow a path.
James M
Thanks for the reply, I did discover the 2-array solution on my own. I still think a GLOB_MODPAR_NAME{2} for individual items would be nice.
James M
I have an array of XYZ points, where either X or Y must be zero. So if the user puts a value in X, it would be nice to have the Y automatically zero out, and vice versa. For non array parameters:
IF GLOB_MODPAR_NAME='xParam' & ABS(xParam-0)>eps THEN PARAMETERS yParam=0 IF GLOB_MODPAR_NAME='yParam' & ABS(yParam-0)>eps THEN PARAMETERS xParam=0James M
Strongly seconded on all points. I would add:
GDL does not even support everything that is possible in Archicad input. For example, no three point arc. I never dreamed my job would involve so much circle geometry.
The environment does not even allow you to see large parts of GDL functionality: Cutting, group operations, graphical editing, 3D hotspots, and I’m sure there are others.
My hope is that this forum is the beginning of a serious recognition of GDL’s importance. The next step is to make the tools and environment less miserable.
James M
Can you tell me if that issue, according to the database, also affects markers? For example, in a section marker, the hotspots fail at one end but not the other.
James M
Here’s one from working on curved windows:
!otherwise you get div0 IF GLOB_CONTEXT=1 OR GLOB_CONTEXT=5 THEN WOD=-4' ELSE WOD=WIDO_ORIG_DIST ENDIF hA=A/2 !half of included angle hNg=ATN(hA/ABS(WOD))James M
Initialize that val_TEST variable.
val_TEST ="" re_parameter = REQUEST (“ASSOCLP_PARVALUE”, “TestPAR”, nameind_TEST, type_TEST, flag_TEST, dim1_TEST, dim2_TEST, val_TEST) labelText = val_TESTIf you don’t initialize you will always get a warning. Sometimes it doesn’t matter, but with Requests it often does. If variables aren’t initialized, they are assumed to be numbers, and that variable needs to be a string.
James M
A value list with output options is hooked to a length parameter, also with two values. You can edit the length with hotspots, so you can graphically switch between the output choices.
Here’s a simplified drawing symbol object which shows how it should work. Graphical editing works mirrored and unmirrored. (Though I am getting a strange error in the parameter script, it says value list values are duplicated ‘0’, but they aren’t. Also the parameter script doesn’t fail. See comments.)
Also, here is a label version of the same thing, all I did was change the subtype. The graphical editing works when unmirrored, but fails when mirrored.
Thanks for your help. In the label it is ‘nice to have’, but if it doesn’t work I won’t pursue it.
James M
Sorry, I’m not following. That’s a Model Element subtype, right? It seems to read the surface of the inside/ref line side of the default wall settings. But I don’t see what it has to do with direction. Reading the globals description more closely, it says the sides can vary depending on the opening direction, and I’m not sure how to apply that to a label. Which is what I’m working on, and I didn’t make that clear.
James M
