9

We're at the investigation phase of creating a web based GIS. We're assessing options for an ORM to use in all of our backend import/export and processing routines. From my research, GeoDjango appears to be more mature, but Rails/RGeo looks to be well thought out and reasonable comprehensive.

Does anyone have experience with both and can recommend one or the other?

M Schubert
  • 103
  • 1
  • 6
  • 2
    Hmm.. I may be alone in thinking this, but I think changing the title to Pros and Cons of using Ruby/Rails/RGeo vs Python/GeoDjango for a web based GIS might get a little more traction. – elrobis Apr 17 '13 at 00:12
  • 2
    Yes. The last thing I want to start is a Ruby/Python or Rails/Django war. – M Schubert Apr 17 '13 at 00:17
  • Can you define or include link to definition for ORM, please? – PolyGeo Apr 17 '13 at 00:40
  • I think we can not separate the language/framework from application. I think you'll need to specify a one or more use cases so you'll get the answer you need. Web based GIS is rather broad. Or at least specify some aspects you want compared. – R.K. Apr 17 '13 at 04:18
  • @PolyGeo, ORM = Object Relational Mapper. A python example of this is SQLAlchemy. It is basically a level of abstraction that allows you to seamlessly change out your web mapping engine or SQL engine without having to rewrite the code to match system-specific architecture. – RyanKDalton Apr 17 '13 at 04:42
  • @R.K. I respectfully disagree---there are questions asking for contrasts between PostGRESql and MySQL, or QGIS and ArcGIS, and those questions are certainly answerable. It stands to reason that popular middleware can also be described in terms of their strengths and weaknesses. At the lowest common denominator, though, my opinion is Python offers greater leverage in the GIS space. – elrobis Apr 17 '13 at 16:07
  • Worth pointing out here that GeoAlchemy is in active development if you're after a well-respected ORM: https://github.com/geoalchemy/geoalchemy – Charles Roper Apr 19 '13 at 14:08

1 Answers1

8

I'll offer a perspective: Python is in more widespread use in the geospatial arena. It is the scripting language of choice for ArcGIS and QGIS and there are a wide variety of high quality libraries available for it, plus community.

Python/Django/GeoDjango are a mature combination, with a somewhat slower, steadier development pace than Ruby/Rails/RGeo, which may be a pro or a con for you. Arguably, the GeoDjango documentation is better than RGeo's.

Ultimately I believe it comes down to developer preference, experience and productivity. If you look at the work of Vizzuality (github), they do some spectacular stuff with the Ruby stack. I'm not aware of any GeoDjango shop with such an impressive portfolio, but please post here if anyone knows of one!

This post by Jeff Atwood on why he chose Ruby to build Discourse is salient:

But why Ruby? Well, the short and not very glamorous answer is that I had narrowed it down to either Python or Ruby, and my original co-founder Robin Ward has been building major Rails apps since 2006. So that clinched it.

Community
  • 1
Charles Roper
  • 1,002
  • 3
  • 12
  • 15
  • I like this as an answer. Vizzuality has a very impressive portfolio. I also discovered OpenStreetMaps itself uses a Ruby stack. – M Schubert Apr 24 '13 at 00:35