Here is the promised white paper on wiki-map project. This is basically
a summary of the discussion we had on this subject with some extension
added to the discussion.
---------------------------------------------------------------------
WIKI-MAP
--------
1) Objective:
a) To create a free open source collaborative Geographic Information
Knowledge base tool to enable community participation in Information
generation, processing and dissemination
b) Wiki_Map will enable local participation to create GIS database
in areas presently not covered adequately. This will be useful for
proper planning, navigation, collaboration by different agencies. This
will also make possible to integrate small bits of information to a
comprehensive information base.
2) Features:
a) Prime feature of the project is a Map server. It will allow
uploading and integration of local detailed maps into a bigger base to
enable seamless navigation.
b) It will allow uploading of maps by remote user with some
mechanism for authentication, calibration and validation.
c) It will also allow hand drawn maps/vector information to be
incorporated in the map.
d) There should be flexible geo-coding to enable numbering of
houses. To cope up with frequent renaming of roads the system should
maintain past histories for the names also. e.g. It should be possible
to search for a house with old as well as new street names
e) The system should be accessible over web. Preferable interfaces are:
i) Thin client with any popular web-browser (based of
javaScript, dHTML)
ii) Client-server model with OS independent software like
interpreted language (tcl/tk, python, Java etc.)
iii) Client-server model
f) Landmark/ waipoint database to be maintained e.g. Railway line,
Metro line, Monuments, Hospitals, hotels, etc.
g) Display of landmark should be filterable. i.e. user should be
able to decide which feature to display
h) The serving of landmark information should also be based on
authentication. (It may not be wise to share all strategic information
with people at large)
i) The projection should integrate satellite images, jpg/gif maps
and vector drawn map, generated through GPS or entered by user
j) The system should be able to generate navigation information
k) Navigation information may be processed through speech processor
to enable guidance to blind people (Simputer+Festival - ??)
l) The navigation should be possible both in terms of best of
distance as well as time
m) It should be possible to feed road congestion information
dynamically and through a time dependent geo-model
n) Client version of the software should be usable independent of
the server.
o) The map should be able to keep both 3D and 2D data.
3) Interfaces
a) The map should be downloadable to:
i) Local PC/Palmtop
ii) GPS devices
b) The site should allow integration with other similar websites.
i.e. It should be possible to use maps from site as input to another
website as overlay. It should also be possible to select different
server for different portion of maps. The client software should
integrate it seamlessly.
c) It should be possible to maintain history of changes in the map
to show chronological changes as well as searching on historical names.
i.e. the map should be able to accept both Calcutta/ Kolkata in address.
4) Standards to follow:
a) NACS - Natural address system to be used for all location in the map
b) GPS data format for vector information
5) Requirement & Possible sources:
a) Data:
i) Images: Preferably we need pretty detailed images of Kolkata.
The image may be processed and postarized to bring out the roads
clearly and hide strategic details. We need to get down to details of
each lane. Satellite pictures will be a good starting point. Goggle maps
may be a starting point if nothing better is available. (The quality of
images near Kolkata is terrible)
ii) We need geo-coded maps of Kolkata. To start with a vector
data will do. Along with the maps and data we need the house numbering
logic. In some places the logic is pretty difficult to comprehend. To
complicate the matter more, there are frequent changes in street names.
This is the part we expect to be filled-up by local collaboration. We
will facilitate feeding the information and expect people to collaborate
to feed the detailed map.
iii) Need vector data of the map of Kolkata. We have some
members with GPS. They may provide us with data on major streets. This
will come handy for geo-coding rest of the streets.
iv) If we can get some digitized maps it will give the project
some head start, otherwise we have to build those data..
b) Software:
i) This is the big problem. There are too many options
available, but none is perfect match for our requirement. We need to
pick our base tools carefully. It will be pieces of existing software
with some addition from our side to integrate the pieces and fill the gaps.
We have put the list of softwares collected in wiki_map site at
yahoogroups. This needs somserious evaluation before a decision on the
selection can be made.
c) Server:
i) As Indra has wisely pointed out the production level system
requirement is a tall order and not going to come cheap. But then
Kolkata is poor city with lots of rich man around. We have got to get
some sponsor and may need to go commercial.
ii) The revenue may be generated through advertisements and
through paid position markers. I think we can reserve discussion on the
revenue model till the development work is somewhat progressed.
iii) IP issues - Indra has made it clear the domain name
ilug-cal.org is paid by Russell McOrmand. But then who owns ilug-cal ?
We/Indra/Russel McOrmand. These questions may become very crucial the
moment there is revenue generation or potential of revenue generation
from the site. Both by the participants and from Government agencies -
who is going to collect the tax from ilug-cal revenue Canada/India. Who
will be responsible for the profit/loss from the operation Russel
McOrmand/ilug-cal/we as individual members.
iv) As ilug-cal.org is already there and we are going to do this
work as part of ilug-cal, there is no need to register a separate domain
for our work. It can be named something like http://maps.ilug-cal.org
. But it also opens a question on commercial exploitation, which will
become necessary for our production server. Tomorrow there should not be
any ugly law suit on us on the revenue generated from ilug-cal domain.
v) We need to register our organization under the category of
professional society like ICWA, IIE etc and get clear understanding on
ilug-cal ownership. Registration of society is a costly affair, it is
going to cost some where over Rs. 15,000/- If need be, let us pay for
the domain name and server from ilug-cal account. Probably this is more
important than participation in events. I think we need to take views of
some legal expert on this subject.
vi) For the development, our best bet is to try the pieces at
our own machines and consolidate it in sourceforge. Sourceforge has
impressive infrastructure for our kind of work - MySQL/Mailing
list/Compilers/Linux/Task trackers/CVS and .... Submitted project
proposal to Source Forge for hosting. The site should get approved in
3-4 days.
6) Timeline
a) I know the above list is a tall order. Let us first try to
categorize the requirements into:
i) No Go: The ones without which the project is not possible
ii) Essential: The ones which will enable proper usage of the
software
iii) Desirable: The tall order, which is the icing on cake. To
make every one happy but can be dropped if we have to.
In the present scenario IMHO it will be better to follow the process of
itiriative development. We should aim at setting up a working system at
the shortest possible time and gradually develop it to match the full
feature list. As most of the bits and pieces are available I think 3
months may be a reasonable time to set up a alpha version of wiki_map.
--
To unsubscribe, send mail to [EMAIL PROTECTED] with the body
"unsubscribe ilug-cal" and an empty subject line.
FAQ: http://www.ilug-cal.org/node.php?id=3
