Osmode
For some time I have worked with Osmode normally set to include 7 osnaps (osmode 127).
Lately I have noticed if I ask for a Distance from say an Endpoint to a Perpendicular (from the popup snap menu), as likely as not the distance will be incorrect (if the lines are not the same length).
I just realised that's because Per is not among my set Osnaps and is ignored unless I either include it (osmode 255) or set osmode to 0 and use only the popup snap menu.
Is that as intended? Is it another kooky Autocad WFT?
Comments
-
While using osmode 127, choosing PER (from the popup snap menu, the toolbar or by typing it) seems to work as expected. But I don't know what you mean by: "if the lines are not the same length".
0 -
I do a lot of these types of snaps and have not noticed any errors. Is there a chance you have elevated entities John. If the entities are not all in the X-Y plane the snap point and perpendicular distance will not be either.
0 -
If I offset a line, Dist will probably give the correct answer when it goes for say an End as it ignores a Per.
If the lines are different lengths and it does the same, going for an End or Mid... or something, it sometimes gives incorrect Dists.
When it does this it gives 2 or 3 different (reproduceable) answers to the same Dist depending on where I pick and the orderI pick the lines.
I have checked all lines are Flat and work in 2D only.
It took a long time to realise what's happening, it's not consistent or confined to particular files.
It often works correctly and I have only identified Per as a sometimes problem so far, so it's hard to mount a credible SR.
Do you guys use running snaps? I wonder if it's safer to avoid them.
0 -
Drafting without running osnaps would be terrible!
If you choose PER from the popup snap menu do you get the correct information in the command line and the correct snap markers?
From your explanation I conclude that you are testing this using lines that you consider parallel. You should check if they really are. If they are not then dist+PER will, logically, produce different values depending on the order of the points you pick.
Points are stored as floating point values and therefore have a limited accuracy. When you say "incorrect" what is the inaccuracy?
0 -
John, in addition to Roy's questions, what type of entities are you using when you see problems?
0 -
Roy, when I got a couple of wrong answers I tried Per from the popup and got the same. I can draw a line using the exact same method and it will always List correctly. It's Dist which introduces the errror whether in a lisp or from the command line or toolbar.
The inaccuracy in the last case was 650mm perpendicular opposed to about 2000mm, the offset dist vs the dist from my endpt pick to the endp of the line I hoped to go Prep to. And I did offset that line so it was certainly paralell, then I Extended it to some other line.
Greg, I most often Dist between 2 lines, but my latest error happens when I measure end to end across the diagonals of the square duct riser under a roof fan and it (sometimes) finds the dist from the frst endpt to an intersect between, get this, an extension of one leg of the of the square polygon (the duct) and intersect of that line and a radius representing the curved corner of the fan body in plan, all within a block of the roof mounted fan.
0 -
John, it sounds like maybe the snap tracking is biting you... I always have it disabled so I was not thinking of it.
Do you have snap tracking on? It is indicated by the STRACK on the statusbar of the main window.
0 -
It does sound like snaptrack at times, but that didn't last long before I turned it off.
Don't like that line dimensioning while you draw thing either, clever as it seems.
I reckon this has been happening since Bcad (apparently) followed Acad's behaviour of overriding any snap within a lisp with osnap settings, but it's happening from the command line too.
It has taken a long time to be sure it is happening, you often don't know what dist to expect or check it. When I do it often gives the same answer. Sometimes the discrepancy is very small.
I can't be the only one.
0 -
I get the Osmode overriding picked or calculated points within lisp routines thing (pain that is is), but when I run one of those with Osmode set to other than 0 and there is no existing point within range, it goes hunting for one or ignores an obvious point in favour of a less obvious choice. It seems to me this happens with Dist and Osmode.
0 -
There has been a changing feeling to the relative weighting of running snaps. I have always kept perpendicular snap out of the running snaps because I found it interferes more than helps. I have PER (and several others) assigned to a function key to make it convenient when using. I do not use snaps in my Lisp (at least not commonly) unless under the user control.
I am not able to cause an unexpected result while working interactively. I am likely not doing the same operations as you but I do a lot of END to PER snapping. I believe you mentioned you an repeatably generate an unexpected result, can you include a small recorded script and command window output?
0 -
John,
we have been following this thread with interest from the very start. Till now we failed to reproduce the problem as we think you experience. The moment you have something reproducible, a scenario, a script, a video screen capture, please let us know.
In fact, it would be helpful to see a screen capture, or get a drawing with some directions on what we should do to have a fair chance to trigger the problem.
0 -
I don't running Per much either.
I don't think I know how to show it here, but I will try and do some screen captures and submit with a file.
0