I was chatting with someone a short time ago about my recollection and involvement (or evolvement?) of Computer-Aided Design (CAD) and how it paralleled with the U.S. Navy shipbuilding design industry.
WARNING: This is long-winded. I'm stuck in the house with slush and ice outside and everyone else is sick or sleeping. I'm bored. I ramble. I'm leaving out some trivial parts. Strap in. Let's go…
1984 - I was hired at Gibbs & Cox as a Junior Drafter. Trained from scratch. Bowl of mush molded into a semi-functioning brain by a group of professionals with bad hair cuts and a missing fashion sense. Nice people nonetheless, and I learned a great deal from them.
Everything was done by hand, using mechanical pencils with E0 or E1 plastic lead on frosted Mylar film. Occasionally we would have to work on Sepia prints, which sucked holy crapcakes, especially if you came into work with a hangover (hey, I was young then) and reaching for the coffee cup, you accidentally knocked over the opened bottle of "Eradicator Fluid". Guess what that stuff does to Sepia? I would letter hundreds of lines of text every day, and became intimately close to my drafting tools:
- Triangles (30/60 and 45/90)
- Letter guides (at first these were printed sheets of paper slid under the translucent Mylar)
- Letter slots (plastic slots to constrain lettering height, and make them uglier)
- T-square on modified household door slabs (high dollar shit there)
- Drafting machine with NASA-approved cabling system and tilt board
- French Curves
- Shape templates
- Standard and Metric 3-sided scales
- Ships Curves and "Duck" weights
- Coffee cup
- Playboy pinup calendar (these were later confiscated)
- Cassette Tape Sony Walkman with latest rock songs
- Visine (weekend recovery tools)
1985 - I was hired at CDI Marine Company as a Senior Drafter. It only took a year to hone my hand drafting Kung Fu skills to get from "Junior" to "Senior". My pay went up a bit as well. Still doing the Mylar and plastic lead drafting. This involved "front sheets" with tabular material lists, general notes, references, title block, the usual. All of it done by hand, even the lines in the BOM table, as well as all of the lettering. I was pretty good at it and was blessed as the go-to idiot for the tougher jobs or those that were facing crunch-time delivery.
In 1986, we were introduced to the cutting edge mainframe CAD system called AutoTrol. It was like ENIAC Junior (you can Google that reference). Cutting edge meant it could do lines, arcs, circles and polygons. A hotfix to the system added colors for the "layers" (or "levels" I can't remember now). Another hotfix added associative dimensions. The Navy played with it, then stopped the car and pushed it out over a mountain pass, letting it fall to its death.
1987 - I was hired at M. Rosenblatt & Son. We had several CAD systems in use by this time. All were mandated by various U.S. Navy or commercial (Chevron) contract vehicles. These included Intergraph EMS/VDS and PDU/PDM, Computervision CADDS 4X (aka "CV4X"), Personal Designer PD-something.
At first, the Navy forbid the use of CAD on anything "design" related, allowing it only for use on front sheets. They were not yet convinced it met their stringent accuracy and precision requirements. That last sentence should make you laugh if you know anything about Navy design standards in the 1980s.
- Still had cassette tape Sony Walkman
- Playboy Lingerie calendar (a new one)
- Our own coffee mess next to the "CAD Station" (four machines in close proximity)
1990s - Over the following years, we added Microstation, and AutoCAD R10 to the mix. I was relegated to CV4X and Intergraph because our bald-headed asswipe department manager didn't like me and refused to let me train on AutoCAD. Probably something to do with me calling him a bald-headed bastard while talking to a co-worker at the urinals, without realizing bald-headed bastard was in the bald-headed bastard toilet stall behind us. Oh, did I mention he was bald? and a bastard? Never mind.
We modeled 3D setups for various naval ship "spaces" to design retrofit changes. Things like clearing an engine room or machinery space to install newer equipment and re-route piping, ductwork, structure and all the stupid insignificant wiring bullshit (that's a cheap jab at my electrical design buddies - ha ha). All of this, at the time, was done in wireframe. There were no decent flat, Phong or Gouraud shading systems at the time. These were mostly $30,000 to $50,000 workstation UNIX-based monstrosities. They took forEVER to boot up and load a part or model for working. Here's a sample of a typical workday:
- 6:30 AM - arrive, turn on office lights, take off coat, apply Visine drops, rub eyes
- 6:35 AM - initiate UNIX login, crack knuckles, stare into empty coffee cup and try to identify what the dried up substance at the bottom is composed of.
- 6:55 AM - login completed. Initiate FILE / OPEN / OPEN PART <select part file>
- 7:00 AM - system indicates part is loading "regeneration in progress…"
- 7:01 AM - go scrub coffee cup, fill with mud (coffee), stir crap in (sugar, creamer powder substance) and watch to see if metal spoon disintegrates in the liquid immersion
- 7:05 to 9:00 - shoot crap with co-workers, most of which are also waiting for their parts to load so they can get to work
- 9:15 AM - part regeneration is complete and is now ready to be worked on
- 9:18 AM - Kevin walks up, and reaches across my pen tablet and clicks "DEL ALL" and laughs. These shitty systems had NO UNDO feature in those days. So such incidents (or accidents) required EXIT PART NO FILE (no save) and start over.
- 11:20 AM - part back up and ready to work
What's interesting is that this was a model of a single auxiliary machinery space. Complete with structure (all forms thereof, including chocks and doublers, hatches, ladders, etc.), piping, valves, flanges, collars, couplings, sleeves, unions, elbows, laterals, tees, and simplex or duplex strainers as well. Then there were the HVAC system items: ductwork, transitions, turns, vaned turns, collars and coamings, in-line heaters and cooling coils, fan coil units, the usual. Then came the big equipment: pumps, strainers, compressors, diverters, scuppers, chiller plants, and separators. Finally, we had the electrical systems, junction boxes, panels, wireway racks, coamings, switches and all that, and let's not forget the furniture and incidental crap.
All of this in wireframe. No solids or surfaces at all. No constraints or associative dimensions. This took TWO HOURS to load. That's right: TWO. FRIGGIN. HOURS. TO. LOAD.
SIDE NOTE: This same set of data, modeled as actual solids and surfaces, with associative dimensions and constraints, can be loaded on a $2000 x86 PC with AutoCAD in less than a minute (on the slowest day I can imagine) - that is progress.
It was during this time that Bald-headed Bastard gave me no other option to teach myself AutoCAD R11/R12 using a few books and staying after hours and coming in on weekends.
We eventually moved everything to AutoCAD, per guidance of the U.S. Navy. We hired an MIT graduate to program some kick-ass LISP apps to automate the design process. Mr. MIT wore sandals and shorts in the Winter. Mr. MIT was a dopehead. Brilliant at coding, but many of his apps suffered bugs. Mr. MIT soon became Mr. Unemployed. His apps became Mr. Unsupported soon after.
I taught myself AutoLISP using Gene Straka's amazing book "AutoLISP Programming by Example" (now out of print). This is (or was) the best programming book ever written, for any language, for any time, for all of the Universe to infinity. You can quote me on that. I wrote our own HVAC and piping systems design apps for AutoCAD R12. I continued on this path for years to follow.
1996 - I was hired at Tenneco / Newport News Shipbuilding to "head up" the installation and management of their first-ever "AutoCAD Network". Prior to this, they were a Netware shop for WFWG users doing clerical stuff and engineering calculations. The CAD work had been done for decades with CADAM and a few other UNIX-based CAD systems. Rice bowls and Fiefdoms aplenty. This was a big fishbowl with a political machinery like the inside of a complicated wristwatch. The tools to be put into motion…
- Windows NT 3.51
- AutoCAD R13
This was my first cross-training into LAN administration, printer and plotter management, as well as continuing on with AutoCAD management and customization. The other systems remained in operation, as well as newer players like IDEAS, Pro/Engineer, MicroCadam and Helix, Unigraphics, and a few others from places I can't pronounce. They spent millions of dollars. We spent a few thousand. They had to send employees to get trained for thousands of dollars. We hired high schoolers, pre-trained on AutoCAD. Any sane MBA would stop right here and make the obvious decision to cut losses. Nope.
2000 - I was hired at SAIC-AMSEC to manage their AutoCAD systems. I continued on with my Windows administration gyration consternation prognostications. I got a certification.
I'm too tired to continue on with this, but now we have AutoCAD 2011 on Windows 7 and things are good. I wish the prices were more reasonable, but it is what it is (don't you hate that overused phrase?).