Which is better

and keeps our bound drawings pretty light....

If you are exporting (a bound set) a series of drawings from one file with multiple tabs looking at multiple xrefs, do you not find that the resulting bound drawings are much 'heavier' <<

We never "bind", NEVER. Not only does it remove intelligence from the file it does make the file size huge. Not only does it unnecessarily bloat the file size, but it creates an enormous overhead of layers, linetypes, blocks, dimstyles, textstyles, etc. that further bloat the drawing. With ETRANSMIT options available there is no reason to bind.

 

Having alot of info on paperspace fills me with worry... what are the advantages ?<<

First why worry about the location of the data?

The first major advantage we have to annotation in PS is an uncluttered model, nothing in the model but the model. No annotation layers to freeze in XREFs, nothing to turn off for plotting model views, no clutter.

The second major advantage is reduction in annotation layers. In model space layers would be required to separate different scale factors and different views of the model so that those layers could then be frozen in individual layout viewports to display the proper annotation. We did it that way in R14 for quite some time, and it's a royal PITB.

There are others, but those two alone were enough to convince us to annotate in PS.

 

We always have our 'model' xref'ed into the sheet file. So the actual model file is clean of clutter and annotation. The annotation is in modelspace on the 'sheet' file.

This gives the same advantage as your first point.... except in the situation where you would use multiple layout tabs, in this case, as I mentioned before, I think working in MS falls apart again ...

Just out of interest, what discipline do you work in?

I'm working in architecture, and there are some engineers here too...

Are you working without xrefs in your sheetfiles.... ie the main information in a drawing is drawn into model space, and the sheet is placed in paperspace.... ? So if I had a floor plan in model space.... then paperspace would contain sheet-tabs for the floorplan, the rcp, lighting, electrical, etc.... so no xrefs for the basic floorplan? ...

If you are an engineer, do you consider your techniques transferable to other professions ?

I know that there are alot of pros and cons to each approach to the problem, but it is very interesting to hear from everyone's point of view, especially from different disciplines.

 

also....

in relation to working in teams of 2 or more people,....

how many PS sheets do you typically have for one drawing...? do you find it a nuisance not to be able to have more than one person working on a series of sheet files ? We find that we are constantly needing to ask other poeple to shut their drawing down to gain access to it, and that is with seperate files for every drawing.... I can only imagine that this problem is made worse when you work with multiple sheet tabs.... Is your solution reliant on careful management of who is working on what package of drawings?

 

We always have our 'model' xref'ed into the sheet file. So the actual model file is clean of clutter and annotation. The annotation is in modelspace on the 'sheet' file. <<

Tried that, it requires maintenance of many more files and hinders fluidity of design. Changes must first be made to the model then the XREF re-loaded in the sheet file before continuing changes there. Uses the layout space in the model file allows for rapid editting when changes occur and greatly reduces the instability of (R2002) DIMASSOC=2 trans-spatially.

I am Cad Manager for the Civil/Structural/Architectural Dept. Most of our work is industrial petro-chemical with some commercial building, pulp/paper, power, distribution, manufacturing and a smattering of highway and rail thrown in the mix. We are a full service engineering firm that can take a project from conception to completetion and through operation.

Each model has it's own sheets in Layout tabs along with any required XREF's of background data (i.e. foundation model is XREF'd into the underground piping model as a background).

We don't do "plans", we build 3D models. The models are divided along discipline lines, then sub-divided along process/function lines within the discipline. Large projects maybe be split up by area as well.

In your "floor-plan" example, the electrical designer would use the structural, mechanical and Arch. models as xref's onto which he would place his required electrical appurtenances. In layout tabs, he would then place and annotate the views required to impart the information to construction.

Using XREF's for background models (i.e. structural model XREF'd into Arch facade model) we can control the plot display of background data by plotting those elements with a very fine or screened line, thus allowing focus to remain on discipline specific elements.

These techniques have worked well for everything from a modular refinery in Indonesia or a commercial high-rise in Houston to a residential plat in Florida or a storm water run-off study in California or a power plant in Panama.

 

Thanks for the description of your work. Its good to know your background to understand the approach you have taken. and it sounds good.

Yep, that's what we are finding... I agree that there's definately efficiencies in keeping it in one file.

I have a problem with some people at my office... in that they have a tendancy to place all the notes and annotation for detail drawings in the model file, and then throw the sheet into paperspace for printing.... What would you recommend I say to them to pursuade them that annotating in PS would be a better plan? If they know that the drawing will only exist in one drawing, and at one scale, is there any other reasons for keeping annotations in PS other than the nuisance of having to have multiple layers for different scaled text and dims?

Even for general plans, sects, and elevs, it's more likely that most annotation is placed in the model file.... (sigh)

We work with a folder of BASE drawings which are xref'ed into a series of SHEET files, which contain the annotations and the sheets... your system basically negates the need for a BASE folder as all your sheet drawings have their own live data in MS as necessary. It sounds like we could simplify our system somewhat by working with your approach.

Do you have any links to examples of project directory structures ? (or even an example from youself) ....(sorry if this has strayed offtopic a tad ) ...

thanks.

 

disclaimer: I work in Civil

My rule is no annotation in the base files unless its stationing or survey callouts. Everything else goes into the
sheets in model space. The exception would be an annotation base file. Say for rough grading where all the base files
are referenced in and the only thing that is in the file is annotation. Then that annotation base file is referenced
into the sheets. That only works in instances where all the sheets use the same view twist. The only other exception I
can think of is profiles as everything goes into that base file except the few things that have to go in paperspace on
the sheet.

patrickmacdonald <>
|>Thanks for the description of your work. Its good to know your background to understand the approach you have taken. and it sounds good.
|>
|> >Tried that, it requires maintenance of many more files and hinders fluidity of design. Changes must first be made to the model then the XREF re-loaded in the sheet file before continuing changes there. Uses the layout space in the model file allows for rapid editting when changes occur and greatly reduces the instability of (R2002) DIMASSOC=2 trans-spatially.<
|>
|>Yep, that's what we are finding... I agree that there's definately efficiencies in keeping it in one file.
|>
|>I have a problem with some people at my office... in that they have a tendancy to place all the notes and annotation for detail drawings in the model file, and then throw the sheet into paperspace for printing.... What would you recommend I say to them to pursuade them that annotating in PS would be a better plan? If they know that the drawing will only exist in one drawing, and at one scale, is there any other reasons for keeping annotations in PS other than the nuisance of having to have multiple layers for different scaled text and dims?
|>
|>Even for general plans, sects, and elevs, it's more likely that most annotation is placed in the model file.... (sigh)
|>
|>We work with a folder of BASE drawings which are xref'ed into a series of SHEET files, which contain the annotations and the sheets... your system basically negates the need for a BASE folder as all your sheet drawings have their own live data in MS as necessary. It sounds like we could simplify our system somewhat by working with your approach.
|>
|>Do you have any links to examples of project directory structures ? (or even an example from youself) ....(sorry if this has strayed offtopic a tad ) ...
|>
|>thanks.


David

 

Our directory structure is pretty simple
<Client> / <Project> / Drawings / <Discipline>

File names are the drawing number, layout names are sheet numbers.

If you're working 2D, there less advantage to annotation in PS.

 

Splitting the design effort along discipline lines, then along process/function lines reduces the need for more than one designer to work a given task. Even back in maual drafting days we tried very hard to keep the tasks controllable enough for one designer to see it through to completion.

Example would be separating the structural framework for all Architectural efforts makes the structural part an extremely simple drawing package as wouls separating the HVAC into it's own package. Using XREF's to keep all the associated cross-discipline files "fying in formation" maintains accuracy.

Some of our project managers, manage more carefully than others but even the least careful is careful enough to avoid major break-downs in efficiency. Our best tool for maintaining that is well-trained designers at the keyboard level.

 

I read all the threads about putting many sheets in one, I guess the conclusion is that its ok to do if it works for
you.
I still dont see the point though as it screws up your annotation in modelspace.
Every time I investigate what is actrually going on with people that have what they call "complex" drawings, they turn
out not to be and that is why they get away with crunching stuff in one drawing.

Try taking a topo, five base files with everything from centerlines to swales around houses, and then base files from
architects done in ADT, and you will see the instability I am speaking of. Then have four cad operators all editing
different sheets. Doing things "old school" is not old. Its just that the "improvements" to A2000 impose a lot of
disadvantages to get multiple layouts. No big deal, now each person can choose how to do things.

I am really looking for the person who deals with the situation above as I am sure I could use some tips on cleaning
drawings from ADT. For those that don't have the instability or multiple team members working on the same thing, enjoy
it.

"Greg McLandsborough" <>
|>What is the best way of doing things.
|>
|>1. Dimensions & text in Modelspace Veiwed in Paper Space
|>or
|>Dimensions & Text in Paperspace.
|>
|>2. Multiple layout tabs or Multiple Drawings.
|>
|>Has anyone conducted a study, or has Autodesk released any papers on this.
|>
|>Cheers
|>
|>Greg
|>

James Maeding
Civil Engineer/Programmer

 

One DWG = 1 sheet method is hardly "old school," since it makes a heck of a lot
of sense in any version of AutoCAD. Multiple tabs in the same DWG file means one
person controls the entire plot set, and no one can one-off a plot of a DWG
file.

Batch plotting largely takes care of the "its easier to plot the entire series
of layouts" and in A2k5 you can batch plot a whole set of separate DWG files in
the background, behind the scenes, without bumping anyone out of a drawing.

Also, if someone loses or deletes the single file with all 20 layout tabs,
you've lost a GREAT deal of work - especially if you annotate in a layout tab. 1
DWG = 1 sheet minimizes the risk of data loss.

Every office is different, and what works for small teams may not work well in
large firms with many players. As you stated, just because you can do something
in AutoCAD doesn't necessarily mean you should. Every office should explore
different ways of doing things and come up with their own methodology that's
tied to the way they work best.

Matt

 

bind.

Unless your pesky client specifically asks for bound files, which happens...

Matt

 

Once we've explained the file size issue, only one continued to demand bound xrefs. We explained the additional cost, he agreed to pay, we bound the files, and choked his email to death, literally. Hey, that's what he wanted. We then zip 'em to several dozen CDs and FED-X'd to his office (all at his cost). When his only drafter finally got a hold of the files he nearly had a stroke. Takes a minute to open a 200meg drawing file on an old PII running Win98 and 256M ram.

 

Have you tried CDGPurge? It does a good job of cleaning up large files....

Sorry, but I don't have a link...

 

I agree, binding Xrefs is absolutely the dumbest thing that CAD users have to
deal with. Educating the client is one step, but we have had several (and one
yesterday, which is why I chimed in) that simply want one DWG file, no matter
how screwed up it gets with bound Xrefs.

Matt

 

One drawing per file also reduces the amount of data that can be lost if the
file becomes corrupted.

 

I am the trainer and even I screw up views once in a while. I am very careful and catch stuff <<

What kind of screw-ups are you talking about, since viewport locking eliminates zoom errors?


I haven't a clue what you mean here. What does the zoom scale have to do with a callout?

proper training takes time<<

We take that time. For experienced users, we have a 16 hour procedural training class that focuses on how we do things and what we expect. For those less experienced we have a 24 hour class followed within a month by another 16 hour class. It's well worth the investment.


Bear in mind we don't upgrade immediatly, and when we do, it's piecemeal. We leave each project on the release that was used when the contract was initiated. So a release may be out for some time and have several patches (or bug warnings) in place prior to our first install. The only corruptions we've seen came from a network hardware (old) glitch or from "dumbing-down" drawings from outside contractors using vertical products like ADT. Now that's noted and solved/worked-around, we don't see many corrupted files.

the same as MS to avoid text style and dimscale problems.<<

Doesn't help with multiple scales on one drawing, nor would it be possible to plot the different views of a model.

Annotations in MS require a different set of annotation layers for every floor plan, every elevation and every detail scale in the file so that they can be controlled in the PS viewports. They also require a multitude of UCS controls to insure the annotations reside in the proper "space". We've tried this method in the past and the massive clutter of these overlapping layers made the model extremely difficult to manipulate.

 

Man, I've BTDT!

Matt

 

I'm sorry I asked.

The debate in the office is equal to one online. Does anyone have any stats,
or Autocad feed back on which is better.

Thanks for all your comments, as I say, it got the boys in the office
talking.

 

actually I have one use for binding xrefs and thats architects base files. As a civil all I want is a base file to plop
into my drawings. So the architect sends me 50 drawings that make up a floor plan. Now what the F am I going to do
with 50 files xrefed into every sheet. And they got their own color / drawing naming conventions, then all the colors
are not by layer, I could go on and on. Oh and they insist in attaching instead of overlaying. ITs a nitemare trying
to decifer their system. So I just bind away until I get one file. Too bad bind does not always work.

Matt Stachoni <>
|>I agree, binding Xrefs is absolutely the dumbest thing that CAD users have to
|>deal with. Educating the client is one step, but we have had several (and one
|>yesterday, which is why I chimed in) that simply want one DWG file, no matter
|>how screwed up it gets with bound Xrefs.


David