Discussion:
Snap-to Problem
(too old to reply)
c***@yahoo.com
2006-06-18 14:32:02 UTC
Permalink
I am having a problem with the Snap-To (entity) function. Whenever I
come up with a need for a circle, an want to Snap-To a tangent to
either a line or curve the Intellicad result at this point is wrong.
The Snap-to tangent function will instead snap to coordinates of
wherever the cursor was placed. Not to the nearest line.

In about 1 or ten times it will seemingly snap-to the tangent of the
line, but instead be about 1/1000th away from it. Thus not creating an
intesection, nor being tangent. I have played with the aperture opening
of the pick point, but this does not seem to alter the operation any.

Is there an adjustment for this snap-to tangent function? I cannot
locate such in the manual, only that the operation is supposed to work
like on other CAD systems.

It works without fail in Autocad, but not in Icad.

Wm.
jg
2006-06-18 20:30:54 UTC
Permalink
Post by c***@yahoo.com
I am having a problem with the Snap-To (entity) function. Whenever I
come up with a need for a circle, an want to Snap-To a tangent to
either a line or curve the Intellicad result at this point is wrong.
The Snap-to tangent function will instead snap to coordinates of
wherever the cursor was placed. Not to the nearest line.
In about 1 or ten times it will seemingly snap-to the tangent of the
line, but instead be about 1/1000th away from it. Thus not creating an
intesection, nor being tangent. I have played with the aperture opening
of the pick point, but this does not seem to alter the operation any.
Is there an adjustment for this snap-to tangent function? I cannot
locate such in the manual, only that the operation is supposed to work
like on other CAD systems.
There is no tolerance. How are you measuring 1/1000th? you can't rely on the
screen image to judge. There has been a bug in one recent version where snap
would stop working completely sometimes, but I have never noticed a small
inaccuracy.
c***@yahoo.com
2006-06-19 00:59:18 UTC
Permalink
Measuring the error of closure is quite easy, for I just construct a
line from the center of the over to a perpendicular of the line/curve.
Then zoom in tight and measure away. Also how I find out what is the
reason why there is no intersection at the locale.


Wm.
jg
2006-06-19 01:30:24 UTC
Permalink
Post by c***@yahoo.com
Measuring the error of closure is quite easy, for I just construct a
line from the center of the over to a perpendicular of the line/curve.
Then zoom in tight and measure away. Also how I find out what is the
reason why there is no intersection at the locale.
Hmm, perhaps it happens on mine & haven't noticed. I wouldn't be sure Icad
sees that as an intersection or would only look for an endpoint. It
sometimes has trouble trimming a circle into an arc between 2 parallel lines
or splines supposed to be 1 diameter apart. (to represent say a horizontal
pipe elbowing down).
smf
2006-06-19 09:05:58 UTC
Permalink
I just tried that, I think.
Circle > 2p > tan > select line > tan > select parallell line
And then it could be trimmed to an arc.

But the original problem I recognize.
Circle > tan ... and then it's hard to understand what happens. works
better with nearest.

Sven
c***@yahoo.com
2006-06-19 14:04:37 UTC
Permalink
Well, I use "Curve" 3P most often. About 10 times more often than 2P.

Pick one point, pick the second, and then Tangent to either a line or a
broad curve, become the standard operation here. It completes just fine
in Autocad, but not Icad.

The nearest option, will often make that last section of the curve
intercept at two points instead of one point. The one point is supposed
to be radial and perpendicular to curve.

Far as I can tell there must be something different in the calulations
or coding deep inside Icad to ignore the true intent.

The, not always, it happens when I want a straight line from some point
to contact the edge of a curve. It will go to some erroneous position
instead of true tangent. The tangent point is supposed to create 90
degrees (exactly) between the line and a radial line, but it don't.


Wm.
F. George McDuffee
2006-06-19 16:20:56 UTC
Permalink
Post by c***@yahoo.com
Well, I use "Curve" 3P most often. About 10 times more often than 2P.
Pick one point, pick the second, and then Tangent to either a line or a
broad curve, become the standard operation here. It completes just fine
in Autocad, but not Icad.
The nearest option, will often make that last section of the curve
intercept at two points instead of one point. The one point is supposed
to be radial and perpendicular to curve.
Far as I can tell there must be something different in the calulations
or coding deep inside Icad to ignore the true intent.
The, not always, it happens when I want a straight line from some point
to contact the edge of a curve. It will go to some erroneous position
instead of true tangent. The tangent point is supposed to create 90
degrees (exactly) between the line and a radial line, but it don't.
Wm.
Try using viewres to increase the accuracy of the curve
representation on the screen. At high magnifications the
location of the "real" curve and the vector approximations on the
screen can be very different.


Unka George
(George McDuffee)

There is something to be said for government by a great aristocracy
which has furnished leaders to the nation in peace and war for generations;
even a democrat like myself must admit this.

But there is absolutely nothing to be said for government by a plutocracy,
for government by men very powerful in certain lines and gifted with the "money touch,"
but with ideals which in their essence are merely those of so many glorified pawnbrokers.

Theodore Roosevelt (1858-1919), U.S. Republican (later Progressive) politician, president. Letter, 15 Nov. 1913.
jg
2006-06-19 23:37:41 UTC
Permalink
Post by c***@yahoo.com
Well, I use "Curve" 3P most often. About 10 times more often than 2P.
Pick one point, pick the second, and then Tangent to either a line or a
broad curve, become the standard operation here. It completes just fine
in Autocad, but not Icad.
The nearest option, will often make that last section of the curve
intercept at two points instead of one point. The one point is supposed
to be radial and perpendicular to curve.
Far as I can tell there must be something different in the calulations
or coding deep inside Icad to ignore the true intent.
The, not always, it happens when I want a straight line from some point
to contact the edge of a curve. It will go to some erroneous position
instead of true tangent. The tangent point is supposed to create 90
degrees (exactly) between the line and a radial line, but it don't.
Is it making a good effort at tangent, or ignoring the snap completely?
c***@yahoo.com
2006-06-20 04:22:46 UTC
Permalink
jg
Post by jg
Is it making a good effort at tangent, or ignoring the snap completely?
Not sure what you are trying to say here.

The 3P function always comes up with a circle, but most times for me it
comes up wrong. The pick-box will visably always overlap the line to be
tangent to. However, I seldom know "exactly" where a true Tangent would
happen to be at. I usually get it to within
"This..........................far" as being about the tangent spot.
The function then should do the calculations and figure out the true
tangent spot. And to be tangent per old high school mathematics, the
curve would then intercept at one unique point only. Not two points, or
miss alltogether.

Wm.
jg
2006-06-20 06:27:51 UTC
Permalink
Post by c***@yahoo.com
jg
Post by jg
Is it making a good effort at tangent, or ignoring the snap completely?
Not sure what you are trying to say here.
The 3P function always comes up with a circle, but most times for me it
comes up wrong. The pick-box will visably always overlap the line to be
tangent to. However, I seldom know "exactly" where a true Tangent would
happen to be at. I usually get it to within
"This..........................far" as being about the tangent spot.
The function then should do the calculations and figure out the true
tangent spot. And to be tangent per old high school mathematics, the
curve would then intercept at one unique point only. Not two points, or
miss alltogether.
Well, when snap dies with flyover snapping on, the aperture disappears.
Without flyover snapping the normal aperture is there but you can't tell if
it worked... unless the gap is big enough to see or you check as you did.
Trying to work out if your snap is inaccurate or just not working at all.
There was a snap bug where it stopped working & had to restart to get it
back, but it seems fixed now.
c***@yahoo.com
2006-06-24 02:10:32 UTC
Permalink
Post by jg
Well, when snap dies with flyover snapping on, the aperture disappears.
Without flyover snapping the normal aperture is there but you can't tell if
it worked... unless the gap is big enough to see or you check as you did.
Trying to work out if your snap is inaccurate or just not working at all.
There was a snap bug where it stopped working & had to restart to get it
back, but it seems fixed now.
I am not really a fan of the flyover option. It is one of those now
convenient commands that anticipates me a bit wong. I generally turn it
off.

Am already having troubles with Snap-to taking over when doing
dimensioning, as the text seems to be a bigger magnet than intersection
or end of line. It will look for and grab a point I cannot even see
within the viewport.

"Now, why'd it go way da hell over there? Hunn!"


Wm.
jg
2006-06-27 21:15:19 UTC
Permalink
Post by c***@yahoo.com
Post by jg
Well, when snap dies with flyover snapping on, the aperture disappears.
Without flyover snapping the normal aperture is there but you can't tell if
it worked... unless the gap is big enough to see or you check as you did.
Trying to work out if your snap is inaccurate or just not working at all.
There was a snap bug where it stopped working & had to restart to get it
back, but it seems fixed now.
I am not really a fan of the flyover option. It is one of those now
convenient commands that anticipates me a bit wong. I generally turn it
off.
Am already having troubles with Snap-to taking over when doing
dimensioning, as the text seems to be a bigger magnet than intersection
or end of line. It will look for and grab a point I cannot even see
within the viewport.
"Now, why'd it go way da hell over there? Hunn!"
Yes, it seems to have slightly different priorities than normal snapping but
you can tell if it worked and what you're going to get. Snapping is one
thing I think behaves better in turbocad.

jg
2006-06-19 23:38:45 UTC
Permalink
Post by smf
I just tried that, I think.
Circle > 2p > tan > select line > tan > select parallell line
And then it could be trimmed to an arc.
It works most times, but occasionally refuses.
Loading...