Dimensions randomly changing when using "save as"

AutoCAD LT 2015. It doesn't happen often that I have drawings that are similiar enough to have me "save as" an existing drawing to start a new one but I have 4 parts that are very similar and created their drawings this way. I drew one up and then save as that one to create the next and so on. All dimensions are in paperspace.

All dims originally were to the 3rd place decimal, angles were to no decimal place. I did these drawings several months ago and everything was fine, a few weeks later someone else looking at them brought to my attention some wonky dims. I couldn't figure out why some randomly changed at that time, I deleted the messed up dims and put new ones on in their place. Back to the drawings again a few months later and several of the dims are messed up again.

I attached a picture showing the issue, the first drawing is the one I started with and then saved as to create the next and so on. The issue being all the dims that are to the fourth place decimal and the 134.4444g dimension, WTF?

I am the only person who opens these files in AutoCAD LT, everyone else can only view them in Trueview.

I have had this happen on a few other random drawings and the only consistant factor is that they were done with save as and this instance is the only example I have. Even a few of the dims on the first drawing are changed but in the past I think only the copy drawing had issues, not the original.

Any input is appreciated.

Zach

bag die.pdf

The usual cause of this problem with paperspace dimensions is with the ones that are not quite associated with what one may think they are. If an associative dimension node misses a modelspace snap point by even a tiny bit, it is not associated to any object at all, but to a random point in space. These unassociated dimensions can and will sort of search around for a point when the program tries to write them back to the display screen. Any differences between the Units precision, and the dim style precision seem to make it worse. A lot of the time, they can show paperspace measurements instead of modelspace measurements, if they are not associated with any modelspace geometry.

I have seen paperspace dimensions go berserk just when zooming.

The only fix is accuracy with the cursor as far as I know.

I use annotative dims instead. They actually work.

Very well explained Dana. I have done dimensions in paperspace for 8 years with little to no drama, as they were neither associative, nor annotative.

Now that we are using ProSteel, the dimensions are mostly done by the software, but can be difficult to sort out manually.

Paperspace dimensioning MUST be associative or it just won't be trustworthy. It also helps if you lock the viewport display otherwise it is very easy to pop in to model space, do some editing, then pop back out to paperspace and in doing so change the scale without realizing it.

A PDF is useless to us. Please attach a copy of one of the actual drawings that you are experiencing a problem with and someone here will take a look at it.

We have always placed our dimensions, and text, in our layouts and not in model space for years with no problems what-so-ever. And just last week I posted a link to an article, from the AutoDesk Knowledge Base, that supported the use of paper space dimensioning. Why would such an article exist if the method was unreliable?

Is the PDF so useless no one has bothered to even look at it?? The dimension itself hasn't changed, the PRECISION of the dimension has changed. All the dims to the fourth place decimal were to the 3rd place decimal when saved. All on the same dimension style. I don't see how that would at all be affected by the relationship between model and paper space.

Zach

Do you want some help or not? Provide an actual .DWG file we can open and take a good look at. Attach the drawing to your next post.

I believe the precision of a dimension is governed by the system variable DIMDEC and this value is saved with the drawing. Is it possible someone used a custom lisp routine that resets this system value? I don't see how or why AutoCAD would arbitrarily change the precision of a drawing all on its own.

Is the PDF so useless no one has bothered to even look at it?? The dimension itself hasn't changed, the PRECISION of the dimension has changed. All the dims to the fourth place decimal were to the 3rd place decimal when saved. All on the same dimension style. I don't see how that would at all be affected by the relationship between model and paper space.

 

Zach

You want to split hairs? How is this? A dimension is not a distance, it is a written record of a distance, a symbolic representation. The words dimension and distance are not synonyms. If the precision level of the symbolic representation of a distance changed then the drawing object CALLED a dimension has indeed changed. NOBODY said the physical distances or the measurements changed. Stop using the words interchangeably. Your own thread title says the dimensions changed.

A pdf is useless in diagnosing a problem with a dwg file because the pdf is merely a picture of the result, while a dwg file has a database FULL of evidence in it. The next time you want your car fixed take a picture of it to your auto mechanic.

I had a whole long explanation about how problematic paperspace dimensions are (no matter how long they may go without problems, and no matter how many people don't seem to have problems with them) typed in the following 300 or so words but erased it.

Thanks for not publishing that dissertation Dana.

Whether one wants to dimension in a layout or in model space makes no difference to me. Each method works and it is clearly up to individual users, or companies, to decide what method is in their best interest to use. I am not saying that any one method is better than the other.

BTW, I too was thinking along the same lines re: the car analogy. Dana decided to actually verbalize it.

Thanks for not publishing that dissertation Dana.

 

Whether one wants to dimension in a layout or in model space makes no difference to me. Each method works and it is clearly up to individual users, or companies, to decide what method is in their best interest to use. I am not saying that any one method is better than the other.

 

BTW, I too was thinking along the same lines re: the car analogy. Dana decided to actually verbalize it.

If one REALLY wants to see how 'lost associativity' affects paperspace dimensions, activate a viewport that has paperspace dimensions, unlock it, go back to paperspace, and zoom out to reveal a screen full of gray, then erase the viewport. I hope you're wearing safety glasses.

Zach:

This is what I would do if I were to acquire a copy of the drawing.

I would open the drawing and do a zoom extents. I would run Overkill on it, -purge it (twice), then Audit. Satisfied that I have a clean file I would then do a "saveas" and name it version 1. Then I'd open version 1 and do a "saveas" and name it version 2. I'd continue doing that until I got to version 10. At that point I'd stop and compare version 10 to the original specifically looking for changes to the precision of the dimensions. If the precision is really changing then it should well evident that many versions removed from the original. And that's why I asked you to supply us with an actual .DWG file as this little experiment could not be run using a PDF. Got it? Good. Now let's get to the root of the matter shall we? Thank you.

Sometimes a picture is as good as you can do if you cannot get the car to the mechanic. I am not sharing my companies files on the internet. When I cruise the net for answers to a problem I am looking for the problem having affected someone else and what their fix was.

Dana, your opinion about model space and paper space dimensions was off topic (you may not see it that way but I do since it didn't even come close to helping me and only padded your post count), I didn't take the time to post up my question looking for fodder.

Zach

Zach571

We don't need the entire drawing. Strip out the titleblock and border and any "top secret" stuff but leave enough so we can still troubleshoot the problem. The alternative is we call it quits and we all go back to our own sandboxes. They don't pay us enough to waste any more time on the matter. Matter of fact...we all volunteer our time and effort.

Sometimes you need to help others help you. We cannot determine what is different, if we don't have access to the information.

See this big >>>>>>>>>shrug

Zach: Would you be willing to email me the drawing? We don't work in the same industry (I work for a specialty chemical manufacturer) and you have my word that I will not share the drawing with anyone else and that I'll delete it after I've run my test. That is the best I can offer at the moment. If that does not work for you then I'm afraid we are done here. Personally, I'd rather get to the root of the problem if we can.

Sometimes a picture is as good as you can do if you cannot get the car to the mechanic. I am not sharing my companies files on the internet. When I cruise the net for answers to a problem I am looking for the problem having affected someone else and what their fix was.

 

Dana, your opinion about model space and paper space dimensions was off topic (you may not see it that way but I do since it didn't even come close to helping me and only padded your post count), I didn't take the time to post up my question looking for fodder.

 

Zach

 

Zach571

Hey Zach, get a handle.

If you aren't looking for help, which is all any of us hope to do, pardon, us if we occasionally display human traits like humor.

Are you sure that you aren't just looking to pad your post count, by posting your 16th post in 5 years? Each one you post makes a statistically significant change in your board history.

You don't need to include the drawing sheet, which might have some proprietary information.

A handful of dimensions and one or two small objects to which they are meant to refer would be sufficient.

You are making this WAY harder than it needs to be. Help us to help you. Please lighten up.

The root of the issue is associative paperspace dimensions that are not "stuck" to anything.

If that is the case it would mean the person who did the drawing did not use osnaps and basically eyeballed it.

The root of the issue is associative paperspace dimensions that are not "stuck" to anything.

I'm fairly certain it's got nothing to do with dimensions in paperspace. My bet is it is a dimstyle issue.

I tried my best to replicate the problem but after creating a new drawing then doing a "save as" ten times in a row (each time adding one new dimension) and comparing each saved version to the next not one dimension (to the fourth decimal place) changed.

I think that Zach's theory that "save as" is the source of the problem has been busted.