The format is there and it's working. The format is GML 2. I know, that GML files is very large (approx 10 times bigger than proprietary binary formats, but they compress well with simple tools like ZIP), they are a bit complex and there is several different way to structure them (eg. the British OS Master map and the Dutch TOP10NL).
But the format is also human readably so it's possibly to correct errors in a ASCII editor, it's structured and relatively easy to translate the structure to other versions and there is a streaming version (WFS). And with WFS it's primary parts of a Map which is requested for further work and then the interoperability is much more important than size, network bandwitch etc. One of the problems has been the complexity and that it requires much more structure, than we has been used to in desktop GIS (but we also need more structure, if we are going to build e-government solutions and to use data from a lot of different sources). Anther problems has been lack of data in GML - there is very few clients (Mapinfo has a OS Master map import, because this is the only way their British customers can get topographic maps) and who will release a client without a data ? EU has tried to approach the problems and is proposing a European spatial data infrastructure (INSPIRE http://www.ec-gis.org/inspire/, where I'm observer in the expert group, representing the regional administrative level). The large US GIS companies is getting a significant amount of their net revenues from Europe and a large part of it is from the public sector. We has started to request interoperability and it's a matter of survival to deliver it (why use a GIS, which can't access the data you need or can't distribute them ?). The first phases is about discovery and viewing, where simple WMS services are a big step forward (where commercial consultant easily can perform an online integration of data from different sources - there is already several Mapserver based WMS services her in Denmark). There is interesting news about the Mapinfo commitment at the UK site http://www.mapinfo.com/industries/government/ogc.cfm and there are rumours about very interesting interoperability solutions in the coming releases. Our price is, that we have to structure our data and that we has to learn a new technology (XML, web services and object orientation). But it's very promising and it's just a matter of starting to use it (there is free client and servers at http://deegree.sourceforge.net/ but I'm looking forward for the Mapinfo solutions for full scale production) Ole Gregor, Staff, Dept. of Environment, planning, Nature conservation and highways Viborg Amt (a county like regional authority) Denmark |---------+----------------------------> | | Neil Havermale | | | <[EMAIL PROTECTED]| | | ems.com> | | | | | | 01-08-2003 09:30 | | | | |---------+----------------------------> >-------------------------------------------------------------------------------------------------------------------------------| | | | Til: "Mapinfo-L (E-mail)" <[EMAIL PROTECTED]> | | cc: | | Vedr.: SUM: MI-L I want a new format..... Or insight | >-------------------------------------------------------------------------------------------------------------------------------| Well I guess I am not alone on the general frustration of increasing dealing with .SHPs.... But there may be some good news...? >>>>>>>> Just what is a WMS or WFS? I just heard yesterday about WMS - Web Map Services and WFS - Web Feature Services as the "new" common formats for making data available over the web. ArcExplorer and ArcMap evidently support this already, with MI to offer compatability with 7.5 (possibly) or 8.0 (definitely). Could this be what you are looking for? Cheers M.... >>>>>>>> and some other comments ..... Buy design? > - no internal info on projection nor datum, all geographical > attributes default to black lines, black dots, and white interiors. > Is this buy design? Yes, it is by really bad design (if you think about how young the format is, you must think how intelligent the developers were). Shape is an extremely stupid format, no projection info, no colors, no text etc. Just points, lines and polygons. In addition it's working with DBASE version 0.00001, means column names are limited to 10 chars and all uppercase (just ULTRA_UGLY). > data environment, I was wondering if some of you with dual MapInfo > and ESRI citizenship, might clue me/us in on how to better deal > with .SHP data? Newer Shape versions come along with two additional files with the extensions "prj" and "xml" containing first the projection and second some metadata. Those are, as far as I know, only supported by V8 of all the ArcStuff ... If you count as well the indexing files ("sbn" and "sbx") it totals up to 7 files per layer (more to come?). Almost as ugly as ArcInfo coverages which are even directories containing a gazillion of files (!) ... And if you look further, ESRI has a very nice collection of formats itself: Coverages, E00 (pure fortran!), Generate, Shape, Personal Geodatabase (Access), Geodatabase (SDE), ArcSDE V2.x, ArcSDE V3.x/8.x and others and probably more to come. And i.e. the project files (workspaces) are binary (holy MapInfo ASCII- Workspaces and holy ASCII MIF/MID which allows even humans to fiddle around if needed ...). At the end there is only one answer to that: FME (www.safe.com). >>>>>>>> and as all ways a voice of sanity in the confusion.... > ... Is a retarded .SHP format just the millstone we, as MapInfo users, must > carry? I imagine that ArcGIS users must have the same problem. If there's no coordinate system info associated with a SHP file, how do they get it georegistered correctly when loaded into an ESRI product? IMHO, it's a mistake not to include coordinate system information with spatial data. As to the lack of font, symbol, pen and brush styles, I think ESRI is correct about separating those from the geographic elements. A geographical entity logically has a one to many relationship with graphic styles, so it makes sense to me that graphic style is not a fixed part of the SHP file data. Information about that probably belongs in a data file as one or more attributes so it can be readily accessed by the DBMS query engine. These same data could be accessed in a similar fashion by a thematic mapping function to apply graphic styles. Externalizing graphic information makes legend creation more modular as well. Portability issues are simplified too when style information is not buried inside a map object's definition. Regarding the SHP file's becoming the de-facto data interchange standard, all I can say is that I'm glad ESRI published the specs. It makes a good point about open standards. They were savvy about that too. Now all the government agencies that provide data to the public and use ESRI's software have no ethical or political problems providing data only as SHP files -- that format is an open standard. This provides govt. agencies with less reason to develop their own open standards (note the USGS has even abandoned their ill-fated SDTS standard in favor of using ESRI's ArcGrid format in distributing NED data), and they become increasing dependent on ESRI software, because that's the only one that easily produces the de-facto standard. This bias cascades through the industry and ESRI multiplies their advantage. Soon the govt. agencies will no longer have the people and resources to maintain independent standards, and they'll just buy their solutions -- from ESRI. -- - Bill Thoen >>>>>>>>>>>>> As far as I know you just have to live with those issues. If you are lucky the data provider will have created a seperate *.PRJ file that sits alongside the SHP and contains projection info. This is one that MapInfo can work with and also the Universal Translator writes out when creating a shapefile from MapInfo. Thats the first bit - and a crucial one to be sure - how many hours have I wasted trying to sleauth out the coordinates from a shapefile ([EMAIL PROTECTED]@#$). Usually I preview it in ArcView or ArcExplorer and have enough sense to guess from a sample coordinate pair. Alternately you fish through a seperate metadata file to find the projection and coordinate info. Or you harass the data provider until they give it up. With regards to the styling, yes thats a bummer too. There is a nifty additional optional file in ArcView 3.x called a *.AVL - a legend file - that defines a style setting or thematically driven setting that loads when you open the SHP file. In the newfangled Arc8 there is something called a Layer file - sort of the same idea - an umbrella with settings to sit over the data in the SHP. Neither of these can be used by non-ESRI tools as far as I know, but if you could read them and apply their contents in MapInfo that could be a fix to this annoying non-feature of the direct SHP. As an aside, SHPs are also approx 50% fatter than TAB (.MAP) with large complex spatial layers like topographic contours. I am very interested in more insight on this topic as well - particularly with regards to Open GIS data formats and standards - the geodata.gov site is very interesting and some of you know a lot more about the Geospatial One Stop initiative and also web-based OGC compliant services than I do...I sit on a state GIS board here in Maine where we are planning to implement a new state GeoLibrary - and data distribution, standards, and formats are all on the table. You find SHP files everywhere for downloadable data, and often those ArcExport files too which have a whole host of their own peculiarities. At least the SHP files are published and not propietary - I fear the day when the download will be a Arc...only or other proprietary format - but open formats can still win. Wish I knew more... Will >>>>>>>>>>>>> Will, Snip>>>>>>>>>> With regards to the styling, yes thats a bummer too. There is a nifty additional optional file in ArcView 3.x called a *.AVL - a legend file - that defines a style setting or thematically driven setting that loads when you open the SHP file. In the newfangled Arc8 there is something called a Layer file - sort of the same idea - an umbrella with settings to sit over the data in the SHP. Neither of these can be used by non-ESRI tools as far as I know, but if you could read them and apply their contents in MapInfo that could be a fix to this annoying non-feature of the direct SHP. As an aside, SHPs are also approx 50% fatter than TAB (.MAP) with large complex spatial layers like topographic contours. <<<<<<<<< I did manage to write some code that read an avl and loaded some basic style values into arrays for later use (i never got around to finishing the code that converted the blank objects to styled objects 'on the fly' in Mi but its 'doable')... If you or anyone else are still interested in this kind of thing give me a prod and i'll try and finish it off...As I recall, matching Region styling wasnt too bad (only one or two unmatched pattern fills) but lines styles was proving difficult with only line weighting and colouring really 'emulateable'(hence my abandnement of the idea), symbols were also problematic because of the need to get the fonts installed.... If anyone has a *.lyr file reader (or *.mxd !?!) then i for one would be keen to see it........ regards Russell Lawley geologist Brit.Geol.Surv --------------------------------------------------------------------- List hosting provided by Directions Magazine | www.directionsmag.com | To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] Message number: 7862 --------------------------------------------------------------------- List hosting provided by Directions Magazine | www.directionsmag.com | To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] Message number: 7863
