Data.Gov.In - Open Data Formats for GIS data
753 views
Skip to first unread message
Thejesh GN
Feb 6, 2013, 3:43:01 AM2/6/13
to data...@googlegroups.com
Hi All,
Team Data.Gov.In wants suggestions for GIS Open Data Formats. They would like to use some of those formats to publish GIS data on Data.Gov.In website. Please do send your suggestions by replying to this chain.
Mine:
GPX - GPS eXchange Format
- Its an OpenFormat
KML
- Started by Google
- But it became an Open Standard http://www.opengeospatial.org/pressroom/pressreleases/857
Sajjad Anwar
Feb 6, 2013, 4:43:29 AM2/6/13
to data...@googlegroups.com
GPX and KML are interesting and open. These formats are quite commonly
used during data collection through GPS devices.
GPX can represent waypoints, tracks and routes efficiently. When it
comes to complex geometry like polygon and multi-polygons, GPX is not
preferred. KML can take care of these geometries but I've always found
it difficult to deal with KML as a form of geospatial information
exchange. Plus, KML was introduced to work entirely with Google Earth
and their upward geospatial stack. I'm not quite sure whether we would
want to endorse them.
Shapefiles (https://en.wikipedia.org/wiki/Shapefiles) are extremely
efficient and widely used, but ESRI has a handle on them and they can
flip the switch to make it a closed standard any day (hopefully not).
If we are ignoring this, I would highly recommend Shapefiles.
In fact, Riju and I spoke about this briefly yesterday and we think
the Geographic Markup Language
(https://en.wikipedia.org/wiki/Geography_Markup_Language) is a good
choice. It was developed by the Open Geospatial Consortium
(https://en.wikipedia.org/wiki/Open_Geospatial_Consortium) and is
widely recognised as an information exchange format. The GDAL OGR
(http://www.gdal.org/ogr/) library can help you deal with it in
whatever ways you want. Moving to a web paradigm, I prefer GeoJSON
over anything. http://geojson.org/
https://en.wikipedia.org/wiki/GeoJSON. They are fast, simple. Extended
from JSON. Modern maps libraries can easily store and represent
GeoJSON.
My preference will go like this:
1. Shapefiles
2. Geographic Markup Language
3. GeoJSON
4. KML.
Cheers,
Sajjad.
> --
> For more details about this list
> http://datameet.org/discussions/
> ---
> You received this message because you are subscribed to the Google Groups
> "datameet" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to datameet+u...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
--
Sajjad Anwar | W: http://sajjad.in | T: @geohacker
used during data collection through GPS devices.
GPX can represent waypoints, tracks and routes efficiently. When it
comes to complex geometry like polygon and multi-polygons, GPX is not
preferred. KML can take care of these geometries but I've always found
it difficult to deal with KML as a form of geospatial information
exchange. Plus, KML was introduced to work entirely with Google Earth
and their upward geospatial stack. I'm not quite sure whether we would
want to endorse them.
Shapefiles (https://en.wikipedia.org/wiki/Shapefiles) are extremely
efficient and widely used, but ESRI has a handle on them and they can
flip the switch to make it a closed standard any day (hopefully not).
If we are ignoring this, I would highly recommend Shapefiles.
In fact, Riju and I spoke about this briefly yesterday and we think
the Geographic Markup Language
(https://en.wikipedia.org/wiki/Geography_Markup_Language) is a good
choice. It was developed by the Open Geospatial Consortium
(https://en.wikipedia.org/wiki/Open_Geospatial_Consortium) and is
widely recognised as an information exchange format. The GDAL OGR
(http://www.gdal.org/ogr/) library can help you deal with it in
whatever ways you want. Moving to a web paradigm, I prefer GeoJSON
over anything. http://geojson.org/
https://en.wikipedia.org/wiki/GeoJSON. They are fast, simple. Extended
from JSON. Modern maps libraries can easily store and represent
GeoJSON.
My preference will go like this:
1. Shapefiles
2. Geographic Markup Language
3. GeoJSON
4. KML.
Cheers,
Sajjad.
> For more details about this list
> http://datameet.org/discussions/
> ---
> You received this message because you are subscribed to the Google Groups
> "datameet" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to datameet+u...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
--
Sajjad Anwar | W: http://sajjad.in | T: @geohacker
Pranesh Prakash
Feb 6, 2013, 7:57:26 AM2/6/13
to data...@googlegroups.com, Sajjad Anwar
Shapefiles will not meet the the requisite features of standards used by
the government as per the National Open Standards Policy. GML and KML
would work well, and perhaps GeoJSON too.
From the NOSP:
4.1.1 Specification document of the Identified Standard shall be
available with or without a nominal fee.
4.1.2 The Patent claims necessary to implement the Identified Standard
shall be made available on a Royalty-Free basis for the life time of the
Standard.
4.1.3 Identified Standard shall be adopted and maintained by a
not-for-profit organization, wherein all stakeholders can opt to
participate in a transparent, collaborative and consensual manner.
4.1.4 Identified Standard shall be recursively open as far as possible.
4.1.5 Identified Standard shall have technology-neutral specification.
4.1.6 Identified Standard shall be capable of localization support,
where applicable, for all Indian official Languages for all applicable
domains.
The desirable characteristics are:
4.2.1 Open Standard having multiple implementations from different agencies.
4.2.2 Open Standard widely used in India for which technical expertise
and support exists in India.
4.2.3 Open Standard that has Extensions and / or Subsets meeting
mandatory characteristics of section 4.1.
The 'Interoperability Framework for E-Governance' (IFEG) doesn't mention
any standard for geographic data.
Sajjad Anwar [2013-02-06 15:13]:
--
Pranesh Prakash
Policy Director
Centre for Internet and Society
T: +91 80 40926283 | W: http://cis-india.org
PGP ID: 0x1D5C5F07 | Twitter: @pranesh_prakash
the government as per the National Open Standards Policy. GML and KML
would work well, and perhaps GeoJSON too.
From the NOSP:
4.1.1 Specification document of the Identified Standard shall be
available with or without a nominal fee.
4.1.2 The Patent claims necessary to implement the Identified Standard
shall be made available on a Royalty-Free basis for the life time of the
Standard.
4.1.3 Identified Standard shall be adopted and maintained by a
not-for-profit organization, wherein all stakeholders can opt to
participate in a transparent, collaborative and consensual manner.
4.1.4 Identified Standard shall be recursively open as far as possible.
4.1.5 Identified Standard shall have technology-neutral specification.
4.1.6 Identified Standard shall be capable of localization support,
where applicable, for all Indian official Languages for all applicable
domains.
The desirable characteristics are:
4.2.1 Open Standard having multiple implementations from different agencies.
4.2.2 Open Standard widely used in India for which technical expertise
and support exists in India.
4.2.3 Open Standard that has Extensions and / or Subsets meeting
mandatory characteristics of section 4.1.
The 'Interoperability Framework for E-Governance' (IFEG) doesn't mention
any standard for geographic data.
Sajjad Anwar [2013-02-06 15:13]:
Pranesh Prakash
Policy Director
Centre for Internet and Society
T: +91 80 40926283 | W: http://cis-india.org
PGP ID: 0x1D5C5F07 | Twitter: @pranesh_prakash
Arun Ganesh
Feb 6, 2013, 8:50:42 AM2/6/13
to datameet, Sajjad Anwar
My vote is for GeoJSON because:
- Full support on OGR, so trivial conversion to pretty much any other existing GIS format. http://www.gdal.org/ogr/ogr_formats.html
- Web application friendly
- Human readable
S Anand
Feb 6, 2013, 9:00:16 AM2/6/13
to Datameet, Sajjad Anwar
+1 for GeoJSON.
Having said that, conversion between these formats wouldn't be an issue given GDAL. Perhaps the bigger challenge would be for departments to comply with a format we suggest. Maybe we could say
a) "Here's a suggested format, but any other format that's readily available is fine too", or
b) Offer them support to convert across these file formats. I'm sure some of us will be happy to help them with this. I certainly would -- just to see the data :-)
Regards,
Anand
prashant khare
Feb 6, 2013, 10:16:02 PM2/6/13
to data...@googlegroups.com
I go with GML, for the rich content support it provides in addition to multi platform support being Open source XML format. However, it then also demands the content provider to ensure marking multiple fields (mandatory) like CRS etc.
I'd also advocate of publishing the data and creating API for more then one format, and for this I vouch for GeoJSON. Both the formats have their own attributes, on one hand GML provides an extra leverage of linking with addition datasets on the network (since this being XML schema and could easily be transformed or linked with Linked data formats) and create content-rich analysis and application. On the other hand, GeoJSON is a robust format, ready to use and access the meta data types format which prevents unnecessary trails. I do not see a point in publishing a data in a format like Shapefiles which on routine might not serve purpose since the data would then after be again needed to transform into a format which the other programming platforms would find themselves compatible with.
Hence, my choice: GML and GeoJSON. (Data in both formats be made available)
Shashank
Feb 6, 2013, 11:26:57 PM2/6/13
to data...@googlegroups.com
I vote for KML because:
1) It's an open format.
2) Conversion to any other format is trivial for a relatively skilled computer user.
3) It is easy to use for most unskilled computer users (point and click) such as the vast majority of rural Indians.
4) It is an easily transmissible file (no shapefile-like bundle of multiple files)
Most people I work with in rural India are extremely familiar with Google Earth and Google Maps; it's changed the way they view their landscape, and as a result they are very familiar with the KML format. If the data on data.gov.in is meant to be accessible by all, it shouldn't be in a format that only a minority are comfortable using.
1) It's an open format.
2) Conversion to any other format is trivial for a relatively skilled computer user.
3) It is easy to use for most unskilled computer users (point and click) such as the vast majority of rural Indians.
4) It is an easily transmissible file (no shapefile-like bundle of multiple files)
Most people I work with in rural India are extremely familiar with Google Earth and Google Maps; it's changed the way they view their landscape, and as a result they are very familiar with the KML format. If the data on data.gov.in is meant to be accessible by all, it shouldn't be in a format that only a minority are comfortable using.
Reply all
Reply to author
Forward
0 new messages