Sep. 2122, 2006

1
v
Sep. 21-22, 2006
FME Worldwide User Conference - Vancouver
Keynote History and Currency of the FME Mark
Sondheim Peter Friesen, Integrated Land
Management Bureau, Government of British Columbia
2
Keynote
History and Currency of the FME, as told by
Customers 0 and 1
Really! This is true!!
3
What was the problem?
We needed to share data among members of a
diverse community.
And we didnt know how
4
How big was the problem?

• It was local.
• It was global.
• It was an impediment to many business operations.

• How could we think about the problem?

from 1990 documentation
5
The Vision
from 1990 documentation
6
Geomatics Information Architecture
Levels of Abstraction,with interfaces between
them
We needed a way to describe geospatial data
models and the data sets adhering to the models
from 1990 documentation
7
SAIF emerges

• Mark and Peter chair national committees looking
  for a Canadian standard
• They also decide to enter the fray
• SAIF is developed by the BC govt.
• It is designed to meet criteria defined by these
  committees
• The last entry in the race, it competes against
  various international standards

8
Serious Competition!
9
Serious Competition!
and the winner is ...
This was in 1991. A few years later DIGEST
wasadded as another standard for Canada.
10
SAIF, Berkeley and the US Army Corps of Engineers
The initial work on SAIF was shown to David Skea
of Minerva Research
David Skea

• SAIF became the basis of the early work by the
  Open GIS Foundation, which was formed in 1994 by
  Kurt, Kenn, David Schell and others, and which
  later was renamed the Open Geospatial Consortium
• Mark took part in these efforts from 1991 through
  1995

11
Evolution and Convergence
SAIF
12
CGSB

• Canadian General Standards Board

CGSB 171.1-95-CAN/CGSB TitleCGIS-SAIF Canadian
Geomatics Interchange Standard - Spatial Archive
and Interchange Format Formal Definition
(Release 3.2) Canadian General Standards
Board Publication DateJan 1, 1995
13
Class Syntax Notation
ltGeographicObject    subclass       
StreamSectionMS    attributes     
substrate                CompositionMS   
                    bankFullWidth         
Real32         restricted     
position.geometry   ArcDirectedgtltEnumeration
    subclass        CompositionMS   
values           gravelly sandy fineGrained   
comments     "Three options for the composition
of the                          bottom of a
stream are available."gt
from 1995 documentation
14
Object Syntax Notation
Dale and colleagues at MDA developed OSN
15
How translators worked

• A big correlation table
• Used typically to transfer feature coded data
  from a data producer to a data consumer
• For example, a mapping company produces
  topographic data to a given specification and
  delivers it to the government
• Hard coded and conceptualized as a thin pipe
  between systems

16
How we thought they should work

• A CSN/OSN file is an intermediary
• Because SAIF allows for semantic richness, we
  dont have to look for the lowest common
  denominator.
• In fact, we can add intelligence to the data

• We wanted to support relational to relational
  mappings
• Object to relational mappings
• Object to object mappings

17
An insight
SystemA
SystemB
SystemC
SystemD
Dale and Donrealized that the intermediary
could bein-memory and need not be a file.
SAIF
18
An ugly secret !!

• In 1991 when SAIF was approved as a national
  interchange standard, we could describe data but
  we had no practical encoding scheme. We could
  neither archive nor interchange data!
• We tried ISO 8211, originally developed at the
  Oak Ridge National Laboratory to hold chemical
  data. It was used by SDTS. We worked with its
  inventor, but to no avail.
• We then tried eXternal Data Representation from
  Sun Microsystems.
• (XML hadnt been invented yet.)

19
XDR

• In 1992 Dale and his colleagues at MDA carry out
  this effort
• It does not prove practical either

20
ASCII?

• (Note the company that did this work!)
• The idea was to stay with ASCII, but zip it into
  a series of blocks, as zip had size limitations
• The approach proved effective

21
MoF Data Exchange Pilot

• 1994 RFP announced for translation
  package to convert industry formats
  to/from MoF format using SAIF as the exchange
  intermediary
• RFP won by Safe Software
• Did the deliverable from this project provide
  impetus for the creation of the FME?
• 1995 - FMEBC purchased by the province to provide
  translation capability to/from SAIF for an
  increasing number of formats

22
FMEBC and GDBC 1995-1996
TRIM ESRI ArcGen
TRIM - SAIF 120000 Base Map
TRIM MOEP
TRIM MapInfo
TRIM ESRI Shape
TRIM Microstation
TRIM ESRI E00
The TRIM Translator Completed!
23
FME and GDBC
The Tool of Choice
Weapon of Mass Transformation!
24
Natural Resource Information Centre
FME is a major part of the solution
25
Land and ResourceData Warehouse Statistics

• Over 300 datasets
• 2,200 spatial data layers currently available
  including associated attribute data
• Over 14,500 metadata records
• Supports 160 applications
• 5 gigabytes downloaded daily
• Users
• Public
• Industry gt 5,000 BCeID
• Government

26
iMapBC
27
FrontCounter BC Solution
Directional Arrow on a Stream?
28
Then and Now

• Consistency, year after year
• Dependability, year after year

29
Some final comments

• The FME is an outgrowth of the efforts on SAIF in
  the early 1990s
• Dale and Don contributed to those efforts and
  with insight and effective engineering developed
  the FME
• Because the FME is so enormously useful for model
  to model transformations and for many kinds of
  geoprocessing, it has become a cornerstone
  element of an ever increasing number of
  geospatial developments
• The FME will continue to play a key role in
  interoperability, as it has now for a decade

30
Thanks for listening!
peter.g.friesen_at_gov.bc.ca
mark.sondheim_at_gov.bc.ca