Three messages on how libraries should handle GIS data. --------Johnnie ------------------------------------------------- Wed, 08 Dec 93 07:54 PST "Virginia R. Hetrick" <[log in to unmask]> Re: Responsibility to get data to user Speaking from the users' perspective on this issue, I think the responsibility of a librarian ought to be to have the data available in a form roughly equivalent to a file, compressed if necessary, that can be sent over the network from the library's computer(s) to my computer (preferably as an FTP put). Unfortunately, most of the computers that I've encountered in libraries and that have CD-ROM based information are not networked. So, I must sit in the library and literally construct the files in some kind of mishmashed subset form that will be small enough to fit on (fortunately in most cases) HD 3.5in floppy disks. I don't believe that librarians should have the responsibility for supplying a full-blown GIS with all its analytical capabilities because, as Jim F. points out, that's really a lab responsibility. I can't remember the last time I did a search of a CD-ROM database which resulted in a single file of less than 3.5inHD floppy size, i.e., 1.44MB. Where the way the information is supplied to the library is some kind of non-standard format, the supplier (NOT the librarian) is, in my view, obligated to provide a mechanism for the library to give its users extracted files which are readable by more-or-less standard methods, such as comma-delimited database files which can be imported into databases and spreadsheets or standard header defined images or drawings which can be imported into image processing (.PCX, .TIF, or .GIF formats) or drawing packages (.DFX, .CGM, .WMF, or .CDR formats). (The specification of file types is intended to be illustrative not definitive.) Related to this topic, I really appreciate the way our library's online system allows me to email the results of a search to my "regular" userid. It saves me an enormous amount of time typing references for papers because I wrote a little REXX program to parse the search results into the format I need to use for the places where I regularly submit papers. It should be possible to do something similar (preferably FTP put as mentioned earlier) with data searches. HTH. virginia --------------------------------------------------- "Matthew Gilmore (DC Pub. Lib)" <[log in to unmask]> Re: Responsibility to get data to user Wed, 8 Dec 93 11:11:25 EST Well, here at DC Public Library we are very optimistic ( :-) ) that we can provide at least some GIS service/information as an integral part of our reference service. We've ArcView on order and Tiger and Census coming. Maybe we are _too_ optimistic? We are only providing this service for the District/Metopolitan area, BTW--we won't have any data beyond that. Matthew Gilmore PS We were inspired by what St, Louis Public Library is doing. According to J Fieber: > > ----------------------------Original message---------------------------- > > To me GIS or spatial data literacy implies having a computer, software and > > knowledge. These users have access to a lot more information, but they > > need to know what and how to use it. The problem is, that these users are > > not necessarily as literate as I had assumed. In fact, as "non-GIS" people > > start using the data, those economists, real estate, sociologists and so on, > > they need a lot of help. So, though they have the computer, software and > > application knowledge, how raw can the data be; or do we have a responsibility > > to cook it a little. > > Another way to look at GIS is comparing it to a piece of advanced > laboratory equipment. I really think a GIS system has much more in > common with, say, an electron microscope than a bibliographic > database. Maybe GIS use should be taught in the laboratory by > professors instead of in the library by librarians? With this > approach, the library provides a basic system suitable for > locating and maybe some quick-and-dirty previewing of data which > the patron then takes to their lab to actually use. > > Naturally, in small specialized technical libraries the line > between the library and the lab may shift a little. > > > For example, the DOQs for Connecticut probably need to be stored on the > > server in compressed as well as expanded mode. The library might need > > Hmmm. I just counted up CT and found 120 7.5' quads. The DOQ's > are 3.25' so that makes 480. Multply by 50 megabytes per > expanded 3.25' quad makes 24,000 megabytes. Large disks can be > had for around $0.85/meg so that comes to about $20,000. I think > you would be much better off putting more money into CPU power and > uncompressing on demand. > > > to devise ways of "clipping" that portion of the image the user needs, or > > joining and clipping when soil and vegetation types don't cooperate with > > our county lines. We probably have a responsibility to supply the user > > with what they need, not just with what we have. We either need to supply > > the data in appropriate format or supply a graphics converter. > > When a student comes in with a paper topic, I don't typically sit > down with PsychLIT and print them out a list of relevant articles > (what they need). Rather, I show them the database (what we > have) and give some starting tips on how to begin a search. > > We have to identify where providing information stops and > performing research begins. With respect to GIS, I think that > line is pretty poorly defined at this point. > > *Personally*, I'd love to have a fully equipped GIS lab here in > the library where I work. I'm a computer junkie at heart. :) > However I think it is somewhat tangential to the main mission of > the library. > > -john > > === [log in to unmask] ================================================ > ======================================= Come up and be a kite! --K. Bush === > Matthew Gilmore [log in to unmask] D.C. Public Library Washingtoniana Division 202.727.1213 ________Celebrating the Bicentennial of the District of Columbia________ 1791-2002 * * * --------------------------------------------------- Wed, 08 Dec 93 11:57:40 EST Patrick McGlamery <LIBMAP1@UCONNVM> Responsibility to get data to user I think the discussion is moving into three areas: Bibliographic Instruction, Collection Development and Reference. If the discussion continues, perhaps we need to break it out into three 'Subject:' headers. =====Bibliographic Instruction===================================== >> To me GIS or spatial data literacy implies having a computer, >> software and knowledge. These users have access to a lot more >> information, but they need to know what and how to use it. The >> problem is, that these users are not necessarily as literate as I had >> assumed. In fact, as "non-GIS" people start using the data, those >> economists, real estate, sociologists and so on, they need a lot of >> help. So, though they have the computer, software and application >> knowledge, how raw can the data be; or do we have a responsibility to >> cook it a little. > Another way to look at GIS is comparing it to a piece of advanced > laboratory equipment. I really think a GIS system has much more in > common with, say, an electron microscope than a bibliographic > database. Maybe GIS use should be taught in the laboratory by > professors instead of in the library by librarians? With this > approach, the library provides a basic system suitable for > locating and maybe some quick-and-dirty previewing of data which > the patron then takes to their lab to actually use. There is not doubt in my mind that spatial information has little in common with bibligraphic databases. Nor do I feel the library should provide much more than quick-and-dirty previews of data, but this assumes a level of literacy that is not there. It is not the geographers or environmental folks I am talking about here, it is the historians, psychologists, etc. who are beginning to use GISes as application tools. Referrals are an important part of what we do as is instruction. Maybe where I say "they need a lot of help" I should amend it to say they need help with some tasks. I think How-tos and other BI tools are appropriate, though labor intensive. One way of dealing with this is to share among ourselves BI tools as they are devised. ========Collection Development====================================== >> For example, the DOQs for Connecticut probably need to be stored on >> the server in compressed as well as expanded mode. The library >> might need > Hmmm. I just counted up CT and found 120 7.5' quads. The DOQ's > are 3.25' so that makes 480. Multply by 50 megabytes per > expanded 3.25' quad makes 24,000 megabytes. Large disks can be > had for around $0.85/meg so that comes to about $20,000. I think > you would be much better off putting more money into CPU power and > uncompressing on demand. Thanks for doing the math. It is daunting isn't it. I expect that I will only add those areas with direct research/teaching applications here at UConn. For example the area surrounding campus which probably should be augmented by getting the Civil Engineering folks to scan and GPSs (! what's the verb?) the air photos for the area. I guess I'm not sure I would be better off putting my money into CPU power. Storage is what we do. CPU power for retrieval rate, but not necessarily for crunching. I mean, even on a network, 50Mb is a lot to "give" someone. Which brings us to: ========Reference===================================================== >> to devise ways of "clipping" that portion of the image the user needs, or >> joining and clipping when soil and vegetation types don't cooperate with >> our county lines. We probably have a responsibility to supply the user >> with what they need, not just with what we have. We either need to supply >> the data in appropriate format or supply a graphics converter. > When a student comes in with a paper topic, I don't typically sit > down with PsychLIT and print them out a list of relevant articles > (what they need). Rather, I show them the database (what we > have) and give some starting tips on how to begin a search. > We have to identify where providing information stops and > performing research begins. With respect to GIS, I think that > line is pretty poorly defined at this point. PsychLIT is a poor example. It is a bibliographic database and we've agreed GIS data isn't. I do "clip" quad coverages for users; topos, surficial and bedrock geology (and anything else that exists), and I "clip" adminstrative polygons; states, towns, tracts, etc. and I would like to be able to "clip" other geographies; basins, geologic areas, etc. What the user does with these paper "Clips" is up to them. The same is true of digital data. I don't think clipping is research necessarily. I think it is data acquisition. ================Other============================================== > *Personally*, I'd love to have a fully equipped GIS lab here in > the library where I work. I'm a computer junkie at heart. :) > However I think it is somewhat tangential to the main mission of > the library. I'm becoming less and less a computer junkie at heart and more and more a librarian as I begin to sense the enormity of the job. There are really some great ideas and issues here. I'd held having an PC ArcInfo key at for years thinking that there was little that a library could do with it. I've had one for the past year and a half and couldn't live without it, just to manage information, not to do analysis. Other Comments?