----------------------------Original message----------------------------
 
Last week Charlie wrote the memo copied at the bottom of this
response.  Though Charlie was writing about copywrite and data
lineage, there was a brief comment on generating files out of TIGER.
Charlie, I understand what you were getting at there, but please
allow me touse it to springboard to another issue.  The issue of
whether the "user" wants bitmapped images, either on paper or on
disk, or wants the data which they can manage and analyze on thier
own system.
 
On MAGIC, we have invested in distributing spatial information as
geo-spatial data, not as maps.
 
I tried the "reference desk" approach back at the beginning of the
ARL Project.  I used Dr. Doolittle to create census block groups
and tracts of the counties of Connecticut, extracted general data
variables from STF1 for population and housing, and made them
available using AV1 in the Map Library, the Reference Desk area and
the Geography Department's Computer Lab.  It was a non-event.  The
users really wanted coverages.  At this point, I'd like to ask for
honest feedback from the participants.  Is this just a UConn thing?
The times when someone came to "make a map" they really wanted, and
did, import their own data into census and line data MAGIC had
available.  Have others found this sort of thing to be true?
 
This January we converted all of the TIGER 95 data to Arc and MIF and
found some cheap HTTPD software to run on our Novell server.  We have
been allowing FTP transfer of our data since 5/94, but the interface
was terrible.  HTTP has enabled us to build a graphical,
cartographical interface to the data.  All the data are in the public
domain, none from Wessex or ESRI is available online.  The numbers
speak for themselves.  There is a LOT of interest.
 
 Month        Bytes       Files     Minutes
 
6/30/94    54,754,676      222       30,278
8/30/94    66,827,165       66          982
10/31/94    7,851,801      239       12,619
7/30/95     8,502,348        9          101
8/30/95       161,738       19        2,659
9/30/95     5,978,020      474       15,827
10/31/95   50,273,595      167        5,941
11/30/95    7,554,192      366        4,171
12/31/95   20,199,325      254       11,138
1/30/96    20,364,861      120       11,138
2/29/96   108,656,028    1,024       57,273
3/31/96   355,994,877    3,100       67,610
 
Some further comments.  These data are zipped at an 85% compression
ratio, so multiply for a true value.  When the user downloads, they
engage in an active decision to own the data.  The data and program
run on their machines.  They make the map.  We might, and do, print
them here when they want a better printer than what they have handy,
but we are printing files that have been layed out, not holding hands
to lay out.  In general, only 1/3 of the users are academic, here at
UConn, Yale and the Connceticut State Universities,  more are
commercial and public.  The major use in in the state, but
consultants as far away as Virginia have spent time and
telecommunication costs downloading files.
 
Discussion?  Is it our role to make maps?  Is it our role to enable
map making?  There are a lot of questions inherent in the path MAGIC
is taking.
 
Forum?
 
 
Patrick McGlamery
Map Librarian
Map and Geographic Information Center
Homer Babbidge Library
Storrs, CT 06268
TEL: (860) 486-4589
FAX: (860) 486-3593
 
 
> Date:          Thu, 28 Mar 1996 23:05:19 -0500
> Reply-to:      [log in to unmask]
> From:          [log in to unmask]
> To:            Multiple recipients of list <[log in to unmask]>
> Subject:       Web maps
>
> Hello friends!
>
> One of our crew has asked a question about copyright and a student creating
> maps to put on a web page. Since this is such a good item to discuss, I
> thought I should do this broad-scale.
>
> It is a somewhat difficult question in general terms. We've been going over
> this one with our legal folk, trying to iron it out once and for all, and
> we're discovering that it's a bit like nailing jello to the wall.
> Technically, the issue is not so much a function of the software but of the
> data. That is, what are the permissions for the given data sets? If they are
> ESRI-generated versions of PUBLIC DOMAIN DATA, such as DCW (originally DMA)
> or ArcUSA vector stuff (originally USGS) or the like, there is no problem
> using the data in a web GIF, and all that is desired is some attribution.
> However, if someone uses tabular or vector info that is not from ESRI *AND
> PUBLIC DOMAIN*, other rules may apply.
>
> For instance, some tabular items in ArcWorld are from the World Bank.
> Technically, the license for use that we secured from these folk did not
> anticipate the "re-publication of data" in web GIFs, and a strict interpreter
> might find such use outside the bounds of the license. Similarly, some of
> the data in the ArcView sample data set was not licensed for use in anything
> other than individual tutorial mode, so if you decided to make a picture
> using the streets of Atlanta and your favorite circuit from the various
> shopping malls, that might be interpreted as outside the bounds of the
> license.
>
> The whole murky matter revolves around (a) "how big" is the entire universe
> from which this one portion comes, (b) how much of the "whole thing" is the
> person using, (c) how much is the originator being deprived of reasonably
> expected revenue by the release of this derivative work, and (d) what has
> the original creator said is permissible with the given material?
>
> So, what's a body to do? The best answer to this one is to know the data
> sources absolutely. Know which ones have a directly traceable heritage that
> includes (a) public domain data and (b) preparation by a person or
> organization who has granted permission for this use. Wessex has generously
> made their data sets usable on the web for our Bessie project; they
> originally came from the Census Bureau, and that's that. But it might go
> beyond the bounds of fair play (as example) for me to say "I'm going to
> create as my home page a 12x18" electronic map of all the streets of Ramsey
> County (MN) using the Wessex data." (Certainly it would be stupid, and one
> of the dullest home page graphics imagineable, even for someone as blatantly
> parochial as me, but it is doable.) However, if I really wanted to do this
> and also wanted to be perfectly safe, I could get the original TIGER data,
> crunch it into coverages myself, and *then* create the map. (This would be
> of course even more stupid, and, I'd wager, a fair bit less clean and
> attractive, given my technique, but it would be legal. As they say,
> unfortunately, it isn't against the law simply to be stupid.)
>
> If you know the heritage of the data sources, and you can speak to their
> being public and/or permitted for such use, then you're pretty safe. If you
> aren't sure, I've found it best to ask; most sources are very happy to have
> you use the stuff and just give them the attribution.
>
> I'm sorry that there is no neat clean rule on this stuff, but, really, the
> best guide here is knowing the questions to ask of the data.
>
> Charlie
> ----------
> Charlie Fitzpatrick
> ESRI K-12 Education & Libraries
> 3460 Washington Drive, Suite 101
> St.Paul, MN  55122-4301
> v: 612/454-0600 x.26
> f: 612/454-0705
> e: "[log in to unmask]" or "[log in to unmask]"
> info: "[log in to unmask]" or "http://www.esri.com/resources/k-12/k-12.html"