View source for Subjective answer on selecting a free GIS

>>''return to [[Open Source GIS Software]]''
----

On this page I will do some very subjective comments on the desktop GIS I know since people keep asking me what I would chose for a GIS.
		
'''If you would ask me the following question:'''

''Dear Stefan,''<br>
''We have a projectX in that we use softwareY and would like to use as well some free*  GIS software for VeryShortPurposeDescritpion? What free GIS would you recommend?''<br>

''sincerely,''
''...''<br><br>
+) probably free as free-of-cost = "free beer"


'''then I would answer with a bunch of questions:'''

* What does VeryShortPurposeDescription contain?
* Do you want to do just maps, or do you need analysis functionality as well? - Ie. do you need a simple data viewer, or do you need a data editor and analysis functions too?
* Are your data in raster or in vector format?
* Are your data stored in a database or just files? and, do you want to write your data to a database or is storing in files ok?
* Will your users work with Microsoft Windows only? ...or is it better to have a cross platform product that runs on MacOSX and Linux as well? (this may become important for adding custom functionality)
* What will be the language of your users? (this is important for user support on email lists/forums)
* Can you deliver detailed case studies / uses cases? I.e. what should the user be able to do with the software and what not? (see for instance the 9 GIS tasks listed in Steiniger and Weibel 2009)
* Is it an option for you as well to use free-of-cost but proprietary GIS, or even low-cost but proprietary?
* If you have some time for a serious evaluation the the several FOS GIS I would recommend to read Atle F. Sveen masters thesis, which presents an approach to evaluate and select a GIS software based on the (functional) requirements that need to be established previously.

I hope these questions limit already the set of GIS software that can be used for your purposes.

As a note: in 2008 - when I wrote this, open source GIS were not that good in mapping tasks yet. But if the software could export for instance in SVG format (vector) or tiff/jpg format (raster) then it should be doable to do some fine tuning with the help of special layout programms such as Inkscape or Adobe Illustrator and GIMP or Adobe Photoshop. However, now in March 2014, I think I can safely '''recommend [http://www.qgis.org/ QGIS]''' for almost every GIS tasks - '''except for''' some '''specialized tasks''' where OpenJUMP, GRASS or gvSIG may be better.


----

	
However, Larry was asking me [in January 2009] a specific question on which I gave my very personal opinion. Maybe this answers helps somebody too.

==  Larry's question: ==

''We have a field school to help Belize with disaster management. Belize cannot afford expensive software, especially the communities.  Nor can students.  I wish that key software types could be available to developing countries at reasonable costs. We are interested in hazard and vulnerability assessments related to floods and storm surges.  Belize uses ESRI software, but the country has limited GIS resources.  They have experienced three major storm disasters in the last two years.''

== My longish answer: ==

(note, in Larrys answer I learned that Belize has English and not Spannish as official language ;)

Hei Larry,

well.. how to start? It is good to know the background you gave in the second email. From that I assume now that a certain amount of tasks will cover GIS analysis. And it is good to know that we talk about Belize - i.e.a spannish speaking country as this changes priorities. I should also tell you that I am a bit biased for open source due to my research background and that I am participating in the OpenJUMP project.

I should start mentioning low cost alternatives to ArcView. There are for instance Manifold GIS and Tatuk GIS that I would consider looking at. However, I can't really tell you how easy they are to use and how the (spannish) support will be.

Looking now on the open source and free software, then I like to note first that support is available via companies (see here: http://www.osgeo.org/search_profile) when custom functionality/ or programming is needed. But in terms of questions (How/Does it work?) email lists are the first and sometimes the only option. However, most of the email lists have discussions only in English - but it does not really matter how good or bad the English is - as long as people can understand, since people from open source projects have different language backgrounds too.

'''Free-of-cost GIS software''' (but not open source) that may be of interest, but which I don't know well are:
* Spring GIS, which is manufactured by the Brazilian Space Institute
* SavGIS, which seems to be maintained by an official body too
* some people have been also satisfied with Forestry GIS for editing purposes

'''Free and open source software:'''<br>
# Quantum GIS (www.qgis.org): has great potential (in terms of already a huge user community), but so far it's available functionality is a bit restricted with respect to your needs (note I haven't had a look on Version 1.0 yet)
# OpenJUMP GIS (www.openjump.org): small user community, but highly specialised on vector data editing and ease of use. I.e. it is quite good when adding new vector data to a system.
# GRASS GIS (grass.itc.it): an overwhelming functionality, but difficult to use - from my perspective ( I couldn't load data without looking into a manual). However, QGIS can be used as an interface to Grass GIS functions. And you may consider Grass GIS if it comes to raster analysis for flood prediction and so on.
# gvSIG (http://www.gvsig.gva.es/): I actually think You should have a first look on this GIS for several reasons 1) it is supported by the region of Valencia (i.e. I assume that support will continue for a couple of years) and 2) it has even a spannish speaking user lists (as far as I know). However, I am not sure how advanced the functionality is, but the goal of the project was to replace ArcView type software in the regional administration - which gives some hints. Note: raster analysis functionality is added from the SEXTANTE plugin, a separate project .
# another free software from a spannish company is KOSMO (http://www.opengis.es/) - with basically all documentation and user support in Spanish. Functionality is on ArcView level, but I think they don't have raster analysis functions.

There are more free GIS such as uDig, MapWindow and SAGA, where the latter is strong in raster/terrain analysis. So you may consider SAGA as well for your terrain analysis tasks. But I am quite unsure in terms of (spannish) support and the future of uDig and SAGA as I have not that much information if they both have developed user communities of reasonable size that are able to maintain the projects over the next years.

What might be interesting as well is Kalypso due to your planned application background in flood prediction. Kalypso has several modules on flood prediction and evacuation. But I am not sure how easy it is too use and how good any support is.

* Note 1: In general there is a tendency by people to use several of the GIS mentioned above for different tasks. Which is ok, as they are free - but of course it takes time to get to know them.

* Note 2: Basically all of the GIS above should be able to read and write shp files - so it should be possible to use the data later in ArcGIS/ArcView.

So.. finally, you see there is a lot out there. And I can't really recommend anything - as it also depends how far the needed analysis functionality reaches. But I would suggest (if you tend to free-of-cost software), have a look first on Spring GIS and gvSIG, then QGIS or  MapWindow or uDig.<br>
However, if you need a GIS for vector data editing I recommend you to use OpenJUMP. The only drawbacks I know of (i) it can not handle very large shp files and (ii) the use of different coordinate systems is not supported out-of-the-box. I.e. OpenJUMP takes the data as is - which makes it difficult to integrate data from very different sources using different coordinate systems. (well.. this was probably a biased comment  ;)

I hope the information did help, somehow...
Stefan

	
	
	
sstein: 27 Jan. 2009 (24 March 2009: added Atles thesis)