Topic: Drawings That Have Deviated From The Standards

[M-dub]

Hey All,
We've got thousands of drawings that go back to '65 for the manual drawings and electronic drawings that go back to about '82 - '83.  They've been drawn or worked on by hundreds of different people, under many different supervisors, for many different companies.  Beginning to get a picture here?

My scenario is this:
A company has a set of standards.  For a few years, those standards are followed, until the CAD Supervisor gets 'replaced', etc.  At which time, the standards mysteriously change as well.  Now you've got draftsmen who are used to 'The Old' way of doing things and a new standard.  This trend continues for FOUR DECADES (or is that three?  Anyway...).  What we're left with is a virtual kaleidoscope of CAD drawings ranging from 100% correct and in accordance to the Standard (of the day) to drawings that are barely recognizable AS drawings.  I would say most of the drawings here are of the 'Baskin Robbins' standard...(31 different flavours).

Of course, Layers are only one tiny little aspect of standards, but it's like the pea under the mattress that got me to get up and post this topic.  I'm in a drawing where everything is on Layer 0 and Layer 1.  Nice    (Note the sarcasm).  I find myself 'on the fence' with this subject because half of me wants to move everything to the correct standard, but the other half (which is beginning to take over) is saying "Screw it!  Garbage in, Garbage out!".  I mean, if I start drawing to the current standard, that stuff will stick out like a sore thumb.

I'll leave it to you folks...Whaddayathink?

[Slim©]

Personally, I would only revise, update, or redraw only the ones that are needed right now.
If you have an intern drafter, this would be a great job to help them get experience, learn your company's standards then become a valuable employee. Or you could get a junior drafter to do it, for the same reason. Great entry level work, just make sure they are well supervised.

[CADaver]

We've had the same CAD manager for the last 12 years... me. It's rare now to find a file in house that hasn't been brought up to current standards.  We did, though, recently have a client that was purchased by another company, as we work on legacy files we bring up to the new client's standards.

[Bob Garner]

To me, it all comes down to:

Got Budget?

[Dent Cermak]

I fix it, then I fire them.Works like a charm!

[hendie]

I'd have to update the blue poodle, simply because it would annoy the hell out of me later in case anyone thought I was responsible for it.


and then there are days when I just don't care.

[hudster]

I as CAD senior HAVE to audit every drawing being issued, if it don't comply I have to make it comply prior to issue.

[SPDCad]

I change the unstandardized files as I come across them. If there is a drawings or a piece of a drawing that I need to complete the current work I have.  I will take the time to update the drawing.   The new updated drawing is then saved with the old drawing name but -UPD is added to the name.  If there comes a time, and they are rare, that a whole projects has been converted over or is close to it (usually the students job to complete a update) the old drawings are archived.
    That was the standard practice at my old office which was fairly large and a general standard was in practice..

[diarmuid]

here is a perfect example of how an extremely elaborate lisp routine (childs play to dome of you guys) could be used to good effect.

the lisp could be made to.  identify all objects on various layers and sort them accordingly.  (sounds simple)

if you adapt the attitude on how you'd make the layers it can be done.  you wont get the drawing completly right but it will be in a dame sight better shape than before.

1. text.

select all text in model space. set the text colour to bylayer.  get the routine to create text layers in accordance with standard text heights.  get the routine to put text of various scales in to their relevent layer, and apply a "fuzz" size i.e. text that is near to the scale. or any "rogue" sizes on its own layer called "rogue_text" for example

2. Dimensions

a bit simpler.  apply the same philosphy as above.  You may want leaders to appear on their own layers. you could then, using layer merge merge those layers with their corresponding text layers.

3. Linetypes,  select eveything that is not bylayer linetype.  move them to their corresponding layer.  when they're on the layers  "hidden_lines" or what ever  make them bylayer but change the layer's linetype to hidden or what ever.  you could probably elaborate on those selection sets as well having them called "hidden_lines_nbyco" the nbyco would be the layers with the entities colour are not bylayer.

4. ploylines

5. viewports

6. titleblocks, north arrows

etc.

and so on.

there are about 30-40 different types of enities that can be placed bylayer. it would be a case of establisihing what sould be placed on what layers.  the standard hatches could go all on one layer or one for each type...BUT you then have 20 layers for 20 hatch patterns, but you can still whittle that down even more. by "hand" at a later stage.

A poorly  standardised or organzed drawing, is for the most part, poor layering.  the lisp that would be composed would be a very large undertaking.  but at the end of it all, it would be a cruise missile of a routine.

If your company uses standard blocks  it could be made to select all the block names from a standard text file( which could be updated regularly, )search for them and move them to a layer for processing. by a junior technician. get them to do it. delegate, delegate

It can be Done.  i'd have done it before but i'm no good a lisp routines etc.  i understand whats involved in bringing a drawing up to scratch , lord knows i've spent time doing it.  there is a lot of repetition involved in it.

the 2-3 weeks you take to put together a routine  it will save you years in both your own manhours and or someone elses.  it seems like a tall order but,  it is a very sound proposal.

you will end up with a drawing, that will be in a hell of a lot better shape than before, possibly even in good condition.

simpify each section, break each set up into its layer "groups"  then set about whittleing those down by "hand" you will probably find that's when you revise the "cruise_missle.lsp" (i've named it, patent pending   ) to do the whittling.

in short.  divide and conquer.

HTH

i'll put the long version up next week.      

[Slim©]

Nice idea, diarmuid. I believe I can work up something like that for where I work. Although we are working on getting some updated standards in place, so it may be a short while.

[yyou]

I would fix the dwg up to standard if the dwg is small like details.
Many times we have to use details from the structural firms.

If the dwg is a large as-built or existing plan from other engineering firm.  
I just try to change the color and the font so it plots right.  The rest of it stay as is.

[Keith™]

To me it depends entirely upon whether or  not these drawings will ever be used again.
For example, you note that some drawings are from 1982. I would hope that by now the project has been completed. In that case, I would say, leave them alone....If the drawing is one that is used regularly, and needs updating, then I would update it completely, including standards.

My point is, that if you have a drawing that is not within your current standards, but it is no longer needed except for archival purposes, forget about it, it is a losing proposition, because when standards change 5 years from now, you will still be trying to revise and update that drawing.

If the drawing is a current drawing and I find it is significantly out of compliance with standards, it must be redone, completely, not reworked, as this is where mistakes are made, but completely redone from scratch. The offending party will remember the problems and will not make the same error twice, if they do, it is the last time they make that error.

[whodo]

I would fix the new parts I was working on
and leave the rest, unless you have the time and
budget then fix the whole drawing.
The problem I see for you is that you say fix but
at the time they were created they were not incorrect
only now after a lot of time has gone by and many new
cad managers are they now incorrect.
So I would have to agree that it comes down to time
and budget, got money got better drawings otherwise
your screwed.

[M-dub]

Yeah, You know exactly what I'm talking about, Ron...You've had the 'pleasure' of working on lots of these 'masterpieces'.  They were all correct, or close to it, at the time they were originally drawn at, but now some of them are just a dog's breakfast.

I think for this client in particular, it's definitely time and money.  The company and the guys who are using the drawings don't give two you-know-whats about what layer / colour / etc. everything was created on...as long as they can read the drawing and that it's correct.

[t-bear]

By "correct" I'm assuming that you mean you can build from it....not that it meets any kind of proper "dtandard".  If the client doesn't have any problem with it, and doesn't KNOW it's a basket case, I guess there's no sense in blowing your profit margin fixing the stupid things.....much as it may hurt your sense of pride....it's NOT your work, after all.....

Sound like a cop-out? ... maybe, but you need to draw the line somewhere......