AutoCAD R12/R13/R14 DWG Format Archive

[#1]

What are the pro and con of using X-clip (versus using a view in Paper
space)

[#2]

Another advantage to X-clipping has to do with demand loading. If you set
INDEXCTL = 1, 2 or 3 AutoCAD creates either (and I'm not sure of the order
here) layer indexces, spatial indexces or layer and spatial indexces. If
demand loading is enabled in the drawing you've got the file referenced into
(XLOADCTL=1) then AutoCAD only loads the information it needs to display the
drawing.

The long and short of this is that if you create layer and spatial indexces
and enable demand loading you will see an improvement in performance if
you've got the layers frozen in the referenced drawing and/or the drawing
X-clipped.

[#3]

One of the main advantages of using paper space for multiple views of the
same drawing is that you only xref the source drawing once, so there is only
one set of layers associated with that drawing. If you xref the drawing
multiple times, using xclip to create your views, you end up with an entire
new set of layers for every version of the xref. You can overcome this
inconvenience to a certain extent by using the Bonus Tool Layer Manager, but
its still a little bit of a hassle.

Another major advantage is that plotting is so much easier if you always
xref your titleblock into paper space, and display your drawing in a paper
space viewport, set to the proper scale. For some reason, the average
draftsman has a really hard time understanding the concept of scale factors.
With paper space, plotting is always 1:1 and nobody has to do any thinking
to get the plot to come out right every time.

There are downsides to using paper space. In my opinion, paper space is
unfit for anything except arranging views and placing drawing titles. For
one thing, most Autocad commands cannot be invoked transparently while in
paper space, so forget dimensioning a large drawing in paper space.
Secondly, the more viewports, the more the computer bogs down. And thirdly,
depending on your LTSCALE/PSLTSCALE strategy (no, I'm not going there now),
the line types may not come out right in a paper space viewport. When this
situation comes up, I usually xref the drawing into paper space at
1/dimscale to make the linetypes come out right. But this means that you
have to do all of your drawing in the source file, which sometimes negates
the advantage of xref'ing in the first place.

Keep in mind, all of my opinions are based on producing architectural
drawings, and these points may not be valid for anybody else. With that in
mind, I use paper space viewports about 90% of the time (to display
different views of the same drawing) . I use clipped xrefs about 10% of the
time (when the linetypes are a problem).

JT

Metz Engineers wrote in message <89j4df$2r718@adesknews2.autodesk.com>...
>What are the pro and con of using X-clip (versus using a view in Paper
>space)
>
>

[#4]

The pros of xclipping are that you have unlimited control over the shape of
the base drawing (even in a viewport) Rectangular viewports may interfere
with notes that are going into paper space, so it is very convenient to have
a means for "erasing" parts of the xref so that they don't show through into
the viewport. There is one drawback that I have found, however (it is a
question that I posted to this newsgroup a while ago and got no response, so
I'll assume that there isn't a known way to get around it). Linetypes that
have dots in them (such as dashdot) will only generate with a space where
the dot should be. So dashdot looks like dashed with bigger spaces. When
xclip is turned off, the dots come back. So my advice would be to be
cautious when you know you are going to xref and xclip, and pay close
attention to the drawing once you have clipped it. We didn't figure this
problem out until the client asked us why the linetype didn't match the
legend, and that's not the most enjoyable way to find problems with AutoCAD!
Metz Engineers wrote in message
news:89j4df$2r718@adesknews2.autodesk.com...
> What are the pro and con of using X-clip (versus using a view in Paper
> space)
>
>

[#5]

Opps... ignore the accidental post.

I feel an urge to clarify...

> If you xref the drawing multiple times, using xclip to create your views, you
> end up with an entire
> new set of layers for every version of the xref.

Rather than xref the drawing multiple times copy the first reference and
re-clip. The xref is only loaded once. Use demand loading (INDEXCTL=3 in base
drawing XLOADCTL=2 in sheet drawing) so AutoCAD doesn't try to regen what it
doesn't need to.

> Another major advantage is that plotting is so much easier if you always
> xref your titleblock into paper space, and display your drawing in a paper
> space viewport, set to the proper scale. For some reason, the average
> draftsman has a really hard time understanding the concept of scale factors.
> With paper space, plotting is always 1:1 and nobody has to do any thinking
> to get the plot to come out right every time.

As an alternate you could xref your titleblock at 1:1 and scale all your xrefs
down appropriately. I don't do this but I've been thinking about it. But I
agree... the average drafter doesn't understand scale factors let alone the
average Architect!

> There are downsides to using paper space. In my opinion, paper space is
> unfit for anything except arranging views and placing drawing titles. For
> one thing, most Autocad commands cannot be invoked transparently while in
> paper space, so forget dimensioning a large drawing in paper space.
> Secondly, the more viewports, the more the computer bogs down. And thirdly,
> depending on your LTSCALE/PSLTSCALE strategy (no, I'm not going there now),
> the line types may not come out right in a paper space viewport. When this
> situation comes up, I usually xref the drawing into paper space at
> 1/dimscale to make the linetypes come out right. But this means that you
> have to do all of your drawing in the source file, which sometimes negates
> the advantage of xref'ing in the first place.

Which commands do you have problems working transparently in PaperSpace? I know
in R13 there were all kinds of issues with PaperSpace and I avoided it like the
plaque... try panning or zooming in R13 - man what a pain! As for dimensioning
in PaperSpace... it works great! Just set your DIMLFAC to the scale factor of
your drawing and go to town. Viewports do tend to bog a computer down somewhat
but I routinely work with drawings that have 20 or more viewports and haven't
noticed a major problem. I think it has a lot to do with how I use x-clipping.
If I've got a base xref that I'm viewing through viewports to create enlarged
drawings I make sure that there is only enough information clipped to show up in
the viewport. I copy the xrefs and re-clip as needed to fill out my sheet.
With demand loading working for you things go really fast. But the linetype
problem has me stumped... I never have any problems getting them to look like I
want. Both LTSCALE and PSLTSCALE are set to 1 and the viewport handles scaling
of linetypes.

What are your thoughts on my thoughts?

[#6]

Greg:

>> If you xref the drawing multiple times, using xclip to create your views,
you
>> end up with an entire
>> new set of layers for every version of the xref.
>
>Rather than xref the drawing multiple times copy the first reference and
>re-clip. The xref is only loaded once.

Copying the xref multiple times doesn't accomplish what we need to do. We
draw a base floor plan that contains walls, doors, windows, and dimensions.
Dimensions for all the different scaled views are placed on this base plan
top of each other on separate layers. Then, layers are selectively turned
on or off in each view. If we use paper space, we use the "freeze in
current viewport" command to get the proper visibility. If we use multiple
xrefs, we have to rename each xref so we get a unique set of layers for each
view so we can selectively turn dimension layers on or off.

>As an alternate you could xref your titleblock at 1:1 and scale all your
xrefs
>down appropriately. I don't do this but I've been thinking about it. But
I
>agree... the average drafter doesn't understand scale factors let alone the
>average Architect!

Yes, we do xref our titleblock into paper space at 1:1. And we xref our
base plan into model space at 1:1. Then we place notes and keys on top of
the model space xref.

>Which commands do you have problems working transparently in PaperSpace? I
know
>in R13 there were all kinds of issues with PaperSpace and I avoided it like
the
>plaque... try panning or zooming in R13 - man what a pain! As for
dimensioning
>in PaperSpace... it works great!

I have to admit, it was R13 where I decided not to use paper space for
drawing because of the transparent command problem. I got so out of the
habit back then, that I haven't seriously tried to see what R14 can or
cannot do. But its really academic. We find that, for us, dimensions are
much better in model space in the base plan. A typical architectural base
plan gets tweaked many times before it goes out to bid. Having the
dimensions right there in the same drawing file with the walls you are
changing makes changes much quicker and accurate (for us).

>With demand loading working for you things go really fast. But the
linetype
>problem has me stumped... I never have any problems getting them to look
like I
>want. Both LTSCALE and PSLTSCALE are set to 1 and the viewport handles
scaling
>of linetypes.

I know this is kind of a repeat of the other thread, but for the sake of
continuity, here's how I do it. Set LTSCALE to the plotted scale factor,
PSLTSCALE to 0, and use linetypes that display correctly at 1:1. Its clear
to me that your method works just fine, too. The only difference I can see
is that with my system you have to place line entities at a scale factor of
1 and with your system you have to place line entities at a scale factor
equal to dimscale. Is this the way you see it, too? And can you think of
any inherent advantage (or disadvantage) to either approach? What do you
think?

JT

[#7]

> Copying the xref multiple times doesn't accomplish what we need to do. We
> draw a base floor plan that contains walls, doors, windows, and dimensions.
> Dimensions for all the different scaled views are placed on this base plan
> top of each other on separate layers. Then, layers are selectively turned
> on or off in each view. If we use paper space, we use the "freeze in
> current viewport" command to get the proper visibility. If we use multiple
> xrefs, we have to rename each xref so we get a unique set of layers for each
> view so we can selectively turn dimension layers on or off.

That's more or less what my firm does too... but I hate it. So much so that I
took it upon myself to modify the system on this project (much to the alarm of
the old school). So far everthings working great. My only beef with the "old"
system is that I can't modify my annotations when I modify my text.
Consequently I have to move back and forth between the files to accomplish what
I want. Furthermore I find it nearly impossible to really organize a sheet
compositionally with my dimensions in the base file. There is a system put out
by the AIA called ConDoc which sort of brings together all the best practices
for dimensioning and noting a drawing. It was put out years ago and doesn't
directly refer to CAD but it does talk about things like placing your dimensions
on the edges of your drawing area to free up space internal for annotations.
When you put dimensions in the base file you (we) typically place them just
about as close to the walls as we can to ensure that they actually show up on
the plottable sheet. What you end up with is, IMO, a jumbled, hard to read
mess. Maybe if our users were more skilled this wouldn't be as much of a
problem. Then there's the whole issue of layer management that always seems to
come up for us when we try to put too much in the base file... YUCH!

> Yes, we do xref our titleblock into paper space at 1:1. And we xref our
> base plan into model space at 1:1. Then we place notes and keys on top of
> the model space xref.

Actually what I was suggesting was that I could make an arguement for getting
rid of PaperSpace altogether. Sequence... xref titleblock in ms at 1:1... xref
base files in ms at 1/scale factor... xclip as needed... layer control as
needed. Viola! This will work as long as you don't need some layers on and
some layers off (sorry... make that FROZEN... off bad...) in the same sheet file
(such as multiple scales or floor plans and reflected ceiling plans). If you're
doing that then yes... you either need to use ps or xref the file multiple times
with different names (not a great idea).

> I have to admit, it was R13 where I decided not to use paper space for
> drawing because of the transparent command problem. I got so out of the
> habit back then, that I haven't seriously tried to see what R14 can or
> cannot do. But its really academic. We find that, for us, dimensions are
> much better in model space in the base plan. A typical architectural base
> plan gets tweaked many times before it goes out to bid. Having the
> dimensions right there in the same drawing file with the walls you are
> changing makes changes much quicker and accurate (for us).

I really think that the methodology my office uses has a direct relationship to
how previous versions of AutoCAD worked. I'm like you... I used to hate
PaperSpace with a vengenance. Then, with R14, I decided it was tolerable. Then
I, quite by accident, figured out that I could dimension in PaperSpace and still
have it associative. And that, as they say, was that. I was hooked. Still am.

> I know this is kind of a repeat of the other thread, but for the sake of
> continuity, here's how I do it. Set LTSCALE to the plotted scale factor,
> PSLTSCALE to 0, and use linetypes that display correctly at 1:1. Its clear
> to me that your method works just fine, too. The only difference I can see
> is that with my system you have to place line entities at a scale factor of
> 1 and with your system you have to place line entities at a scale factor
> equal to dimscale. Is this the way you see it, too? And can you think of
> any inherent advantage (or disadvantage) to either approach? What do you
> think?

About the only disadvantage I can see is that you actually have to do
something. I mean if I create a new drawing (without a template) LTSCALE and
PSLTSCALE are already 1. I never even thought about doing it your way. What
advantages do you see to your system?

I'm not sure I get what you mean by "place line entites at a scale factor." In
my base file I set my LTSCALE to whatever my DIMSCALE is (we've actually got
buttons that take care of a lot of this for me). I create my lines with what
ever linetype I think I need. If I need it smaller or wider I either pick a new
one or modify my ACAD.LIN file to give me what I need. I know I could modify
the entity to have a different scale factor (and I assume that's what you mean)
but to me that's like making an entiity something other than BYLAYER which is
for me (almost) a cardinal sin. (I say almost because I actually do it quite
frequently but only when I don't feel that an entire new layer to manage is
necessary for only a few lines.)

[#8]

Quinton...

This reminds me of a topic that I would like to talk about. How to arrange
the information in model space, so that it is comprehensible for other people
who work on the same drawing.
I draw rectangles in model space, on the defpoints layer. These rectangles
are arranged in rows. They have a tab on the upper left corner, so that they
look like the icon that Windows uses for directories. In the tab, I write in
big letters one or two words describing the contents of the rectangle. For
instance, DETAILS , another rectangle´s tab can have something like
ELEVATIONS, and another one says MASTER XREF-(XCLIP). Below every detail I
write a letter or a number: A , B , 3 , etc... also on Defpoints. When I go
to Paper Space and try to create a viewport, I turn on the "Airplane" that is
"Aerial View". Then I use the pan option inside Aerial View. Then, it is very
easy for me to find, let´s say Detail #3 inside the DETAILS "folder". It
helps other people find information in my drawings, too. If they want to add
a new frame type or door type, they find the rectangle or "folder" with that
name on the tab, and put them there. The tabs also say, as I explained above,
if an x-ref is an x-clip.

By the way, another advantage of using X-clips is that it reduces the size of
the drawing. Compared with the same drawing with the same X-refs but without
X-clips.

Alfredo Medina

Quinton Mcdaniel wrote:

> Well first you want to make sure that who ever might work on the drawing
> knows the drawing has been x-cliped.
>
> Pro Xclip in R14 you can clip in polygon where a viewport can only do
> rectangles this is the best feature although I hear that 2000 can do
> irregular shaped viewports. I makes it possible to hide information on a
> drawing that you might not need to see.
>
> I really cant see any Cons besides just make sure that who ever opens the
> DWG knows that it has been X-cliped.
>
> Metz Engineers wrote in message
> news:89j4df$2r718@adesknews2.autodesk.com...
> > What are the pro and con of using X-clip (versus using a view in Paper
> > space)
> >
> >

[#9]

Well first you want to make sure that who ever might work on the drawing
knows the drawing has been x-cliped.

Pro Xclip in R14 you can clip in polygon where a viewport can only do
rectangles this is the best feature although I hear that 2000 can do
irregular shaped viewports. I makes it possible to hide information on a
drawing that you might not need to see.

I really cant see any Cons besides just make sure that who ever opens the
DWG knows that it has been X-cliped.

Metz Engineers wrote in message
news:89j4df$2r718@adesknews2.autodesk.com...
> What are the pro and con of using X-clip (versus using a view in Paper
> space)
>
>

[#10]

Greg:

>There is a system put out
>by the AIA called ConDoc which sort of brings together all the best
practices
>for dimensioning and noting a drawing. It was put out years ago and
doesn't
>directly refer to CAD but it does talk about things like placing your
dimensions
>on the edges of your drawing area to free up space internal for
annotations.
>
ConDoc used to be our bible in the board drafting days. And you're right,
there's much about sheet composition in there that is just as valid for CAD
drafting as board drafting.

>When you put dimensions in the base file you (we) typically place them just
>about as close to the walls as we can to ensure that they actually show up
on
>the plottable sheet. What you end up with is, IMO, a jumbled, hard to read
>mess. Maybe if our users were more skilled this wouldn't be as much of a
>problem.
>

The old standard was, the first string of dimensions should be 3/4" away
from the object, and each additional string should be 1/4" from the
preceding one. And all walls that touch an exterior wall are dimensioned
with an exterior string, rather than a string running through the interior
of the building. But even with all these rules, dimensioning is still as
much an art as a science.

> Then
>I, quite by accident, figured out that I could dimension in PaperSpace and
still
>have it associative. And that, as they say, was that. I was hooked.
Still am.
>
This is brand new information for me. I'm going have to go check it out.

>I create my lines with what
>ever linetype I think I need. If I need it smaller or wider I either pick
a new
>one or modify my ACAD.LIN file to give me what I need.
>
Does this mean you have a different defined linetype for each scale factor?
If so, this seems like more trouble than using a single set of linetypes
that work for every scale via the varying LTSCALE setting.

> I know I could modify
>the entity to have a different scale factor (and I assume that's what you
mean)
>but to me that's like making an entiity something other than BYLAYER which
is
>for me (almost) a cardinal sin. (I say almost because I actually do it
quite
>frequently but only when I don't feel that an entire new layer to manage is
>necessary for only a few lines.)
>
I couldn't agree more.

By the way, thanks very much for the lively debate. It seems that whenever
I suggest to some people in this NG that their LTSCALE/PSLTSCALE strategy
may not be the only one or the best one, they seem to get offended and just
cut off the discussion. I really appreciate your frankness and your ability
to explain and analyze. I have found this thread very helpful.

JT