X-Git-Url: http://wagner.pp.ru/gitweb/?a=blobdiff_plain;f=geography%2FfGIS%2Findex.html;fp=geography%2FfGIS%2Findex.html;h=9fe5611a0ab2fb99137b1d0fdf76698a1731c15c;hb=e8890a31f945acef02c617264aa041a6ace03366;hp=0000000000000000000000000000000000000000;hpb=9766a16a9d842ad56bfda8c0444ffa96e14cfe37;p=sites%2Fhome_page.git diff --git a/geography/fGIS/index.html b/geography/fGIS/index.html new file mode 100644 index 0000000..9fe5611 --- /dev/null +++ b/geography/fGIS/index.html @@ -0,0 +1,131 @@ + + +fGIS project + + +

f(GIS) project

+ +fGIS stands for "functional GIS". It means that fGIS uses +functional data model rather than object-oriented. Spatial phenomena +are interpreted as functions of spatial coordinates, defined on +some subset of earth surface. Such subset can be contineous area, +or set of lines, like transport or hydrography network, or even +set of isolated points. +

+ +I hope that such approach would allow to develop powerful map operation +language much like simple relational model of tabular data give birth to SQL. +

+Idea of fGIS was born when I've worked on DOS with EPPL7 and Arc/Info GIS systems. I've found that +on most tasks, related with natural sciences, EPPL7 outperforms Arc/Info +hundred times or so. So, I wanted to have system which is open-source, +have purified ideology, and is free from some EPPL7 limitations, caused +by 16-bit nature of DOS. Later I've found that GRASS already implements most of my ideas, +but GRASS is huge and knowing how other raster GIS like IDRISI are outperformed +by EPPL7, I have very few hopes that its existense makes my efforts useless. +

+Concepts of project, both data model and program design are described +on separate page +

+If you could read Russian, you can read old conceptual paper on functional GIS. + +

Project structure

+ +To avoid inventing of yet another, incompatible scripting language, we +decided to develop f(GIS) as extension to Tcl/Tk. It give us not only +full-featured macro language, but also cross-platform GUI and large set +of extensions, including interfaces to major RDBMS servers. +

+fGIS project now consists of following parts: +

    +
  1. Portable C library to access EPPL7 data files.
    + Documentation in russian language and PDF format + is available. Later, I've decided to drop EPPL DGT vecrtor format in + favor of my own, which would be more suitable for fGIS concept. DGT + is bit too object oriented for me, while it is most compact and fast + of non-topology vector GIS formats available. +
  2. Environmental planning Utilities - reimplementation of major EPPL7 + commands. EPPL7 stands for Environmental Planning Programming Language, + but I don't want special language for it. There are lot of existing ones. + I prefer to use Unix toolbox philosophy, so it is set of independednt + utilities. +

    + Not all of this utilites are written yet. + +

  3. Tcl extensions to visualize data + It is simple and most portable way to write map viewer. Planned + features include drawing of raster maps as semi-transparent patterned + images, which allow to overlay raster over raster, but now only + colored maps work. This extension includes commands for editing raster + files and for spatial queries + +
  4. Set of Tcl libraries to implement high-level tools. + It should lead + to complete graphical application with powerful macro language (TCL itself), + access to most of SQL servers (all which have Tcl interface libraries, + including ODBC standard on Win32 platform). + This libraries should include wrappers + for utilities, to hide from user two important things: +
      +
    1. that raster values are always integer numbers rather than text strings +or real numbers and "legend files" are used to map one to another +
    2. that logical map layer is not nessecary stored in separate raster file. +It can share raster with other layer and differ from it only in value +remapping (reclass) table. +
    + This libraries consist of several abstraction layers, which are not properly + documented yet. +
  5. Tcl application which provides graphical user environment. Idea behind + it is that most everyday tasks should be performed by few mouse clicks, +but if you want more you always can open Tcl console and have a full power +of Tcl at your fingertips, including ability to modify GUI. + + +

    Current state

    + +f(GIS) project haven't reach evel alpha stage yet. +I would consider it alpha not sooner than at least 80% of features would +be implemented. But it is already usable for those who own EPPL7 and want portable viewer for their maps +or utilities which overcome some limitations of EPPL7. You can browse +TO DO list. I would be grateful if you would peek +something from it and attempt to implement it. + +

    +There is no snapshots and releases. This page have only historical +interest +

    +Be sure to read installation instructions +before attempt to install fGIS. +

    +Of course, it is not guaranteed even to compile, nevertheless to run, +although it have been used successfully on Linux (GCC 2.7.2), +Solaris-SPARC (GCC 2.7.2) and (utilities only) DOS (Watcom C 10). +It would almost undoubtely fail to compile on 64-bit platform (any +brave soul with Alpha or UltraPenguin to port?). +

    +Current version of visualisation library works with Tcl/Tk 8.0p2. +It should also work with Tcl 7.6p2/Tk 4.2p2. + +Note that dash patch is +almost neccessary, becouse without it Tk would unable +to print (save to postscript) raster maps, produced by f(GIS). + +Img extensioncan also +be helpful, if your raster data are not limited to GIS maps. +Note that +current version of patch, accompaniing this extension, allows to make offsite of raster maps transparent. In future, this would be requirement, but +for now, while fGIS is not fully functional. I don't like to force you to + apply patches which could break your +other Tcl scripts or extensions. + +

    Distribution policy

    +Once f(GIS) would be released, it would be distributed +under GNU Public license. +Pre-alpha versions are freely downloadable, but can be used for evaluation +only unless you are planning to take part in development. If f(GIS) is useful for you even in such limited state, contact me and I would consider marking some parts of project alpha, +thus allowing their redistribution. + + + +