[Historic] Hello list

[Brad Thompson]

It's good to see you join this list, Matt - welcome!

The historical GIS work you're doing at the NYPL is one of my favorite
things on the internet. I consider the holy grail of historical mapping
(and one of my main goals) to be a canonical set of map imagery, covering
as many times and places as possible. And here you have the two main tools
required for getting there, right there on the NYPL site - web-based tools
for both georectification and vector tracing and object classification. I
agree that OSM shares the same kind of goals of being crowdsourced, open,
etc, and hence seems a natural fit (especially for immediate historical
data, as you say). But if OSM is to play a role in historical maps, then
those larger questions (relation to current OSM, modeling of time) you've
identified will need to be answered.

I imagine, at a minimum, a basic framework with the semantic rules for
embedding a historical map just like a normal OSM map, in which a developer
could simply include additional parameters to specify the time to be
displayed by the map. Then, if imagery is available for the exactly
specified time, the map tiles would be served, and if not, the temporally
nearest available imagery would be served. Imagery could be fed to this
system, as I imagine it, could come from one of a number of different
sources: a Tim Waters/NYPL style set of warping and tracing tools for older
imagery, and/or an OSM-style editor.

Looking forward to more discussion!

Brad



> Hi all,
>
> I'm Matt Knutzen, Geospatial Librarian at the NYPL. We're a collection of
> about 1/2 million sheets, 24k books and atlases and a growing collection of
> digitized historical maps (25,000 by the end of 2013), mostly of NYC.
> Together with our friends at EntropyFree (now Topomancy), we've pushed the
> development of the Map Warper and Digitizer (tracer), documented the
> application, secured NEH funding to run a whole  bunch of our maps through
> the scan/warp/transcribe process, and are now building a historical
> gazetteer of NYC using a hodgepodge of data from outside NYPL and
> transcribed from our own collections (eg 57,000 building footprints with
> historical addressing from 1852). The hope is to flesh this out to a full
> blown reference work for the research of historical geography and
> ultimately put it to use in the service of historical geocoding, geoparsing
> and the like.
>
> I've known for some time that there's great opportunity to engage a wider,
> cartographically inclined community and openstreetmap has always seemed a
> natural fit. To that effect, Topomancy have enabled our maps to work in OSM
> editors. Many questions articulated here remain: as to the relation
> historic data has to current OSM data, the modeling of time, the policy
> around the immediate historical nature of OSM data, and how that data is
> tracked, stored, archived, preserved.
>
> I'm very happy and grateful (thanks Mikel) we're convening this list, but
> know we could benefit even more by meeting up in person (an OHM gathering).
>  I can offer meeting space at the NYPL and am happy to offer a hand in
> planning and logistics related to bringing folks together. Also, since some
> of this conversation will be happening at SOTM, it would be excellent if
> those facilitating/attending there could give a synopsis when they return.
>
> Talk to you all soon,
>
>
> Matt
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/historic/attachments/20121001/2e600867/attachment.html>