[freearchitecture] Detail libraries

[Steve Hall]

Bruno Postle wrote:
 >
 > [snip]
 >
 > I think a modern-day online pattern book of detailing "best
 > practice", produced by and for architects; but accessed, maintained,
 > documented and improved using techniques drawn from the
 > free-software world could be incredibly effective - The problem
 > would be getting other people to agree :-).

Interesting, we have a volume called "Architectural Graphic Standards"
in the US which is a dead tree version of what we're discussing. I'd
not considered a detail library akin to pattern books, or to the
one-step-removed standards books like AGS, but it really is the same.
Great insights.

Of course, the book obtains its credentials from being both
long-standing and referenced on each page by So-and-so, AIA (American
Institute of Architects). Any kind of detail system we cooked up would
need to document sources some how. A sort of CVS for drawings could be
helpful. It might convey the many variations a given detail underwent
by architect and application.


 > (There's another more serious problem, which is the absolute
 > unsuitability of current CAD file formats for this sort of thing)

I assume you mean the on-line library system, due to the brower's
inability to render a CAD file, meaning thumbnail/blowup nightmare?
Wouldn't it be nice if we could develop an SVG CAD system? ;)


 > > Building component companies, generally make digital details
 > > available for inclusion into architects drawings
 >
 > In my experience, these details are universally garbage, either the
 > manufacturer tries to push their own product by introducing specific
 > construction techniques, or they are simply useless, displaying a
 > complete lack of understanding of the real requirements.  Any usable
 > detail library needs to be produced by the people who actually need
 > and use the information.

Yes, I agree with your assessment of proprietary details. But doesn't
a CAD detail library's quality ultimately boil down to it's author(s)?
Graphic Standards has maintained its quality by naming individuals for
each detail. This instills a sense of ownership and pride to
contributors. Expertise is also directed, rather than anonymously
scattered through the volume, hit or miss.

I think a Free detail library would require both an organizer-in-chief
and section owners. I coined the term "Library Czar" in our office.
This individual has the power to eliminate entire details if their
quality is low or they are unhelpful. Everyone else (is supposed to)
contribute and file as they wish, the whole system is open. You need
some way to keep the quality high while at the same time encouraging
submissions.

Since the body of work is mostly a collection, rather than one big
solution like most GPL projects, any one piece may stand alone.
Without incredibly precise and automated standards, it will be nearly
impossible to collect uniformly useful and quality details from around
the world. But perhaps a project like this isn't intended to regulate
quality, maybe simply organizing them is half the battle? Hmm... seems
that's where I left off on my last mail.


Steve Hall  [ digitec___ndspring.com (insert "t at mi") ]