The Rise of the Web Programmer

The web programmer is increasingly becoming more important to the geospatial community. However, where are these programmers coming from?  Either our GIS trained people move into web programming or web programmers move into geospatial applications arena. Most of the traditional GIS product vendors are embracing the Internet and are delivering web development tools. They may be reluctant to open their technologies to the broader development community but they have recognized that Internet technology is an integral and important part of their product mix. So maybe we can expect that people trained in these products could also be trained in some level of web development. But is this an optimum outcome?

The web programmer is increasingly becoming more important to the geospatial community. However, where are these programmers coming from?  Either our GIS trained people move into web programming or web programmers move into geospatial applications arena. Most of the traditional GIS product vendors are embracing the Internet and are delivering web development tools. They may be reluctant to open their technologies to the broader development community but they have recognized that Internet technology is an integral and important part of their product mix. So maybe we can expect that people trained in these products could also be trained in some level of web development. But is this an optimum outcome?

From my experience, its clear that people trained in web development are quite different creatures to those trained in GIS development and their Internet components. Typically GIS trained people come from Cartographic and Drafting backgrounds where building spatial products is an important business activity and are typically not trained as software engineers like most IT programmers. I started out as a Cartographer and Photogrammetrist and decided that I liked computing so gravitated towards programming in the 1970’s. Fortran was the language of choice in technical computing and the Internet did not exist. But of course along came the many different GIS products and I like many others got engaged in software development at the scripting or macro level. This was simply to make the user’s GIS experience more convenient. Historically many people move into the programming world via this macro world on many different GIS platforms. This was simply pragmatics and was never a way to develop rigorous skills in programming.

The universities were all teaching software development skills like crazy back then as they needed to churn out programmers to quench the thirst of industry that was embracing computer technology at an alarming rate. Some of us mapping folks got retrained in this surge and some successfully migrated to the Computing profession, however not many people had both programming and GIS skills back then. But as the GIS marketplace grew there was a new brand of GIS user evolving, one that could not only use the software extremely well but could also modify these tools to do more – they needed only GIS product skills which were invariably proprietary.

When the Internet arrived all the GIS vendors started looking very closely at the technology but only adopted it when customers pushed them. They much preferred the controlled environment of a closed enterprise implementation. As the Internet developed further and as web service technology began to appear they all started to augment their products with Internet technologies. Then along came OGC and most of the major vendors joined in support albeit in some cases paying lip-service support only. This gave us a framework for development on the Internet in an interoperable way. Could the GIS-developer easily adapt to this new paradigm of web application development?

New graduates from university are trained on this type of technology. Not only do they understand how to build robust object oriented applications but they also understand the ‘glue’ that holds it all together in the Internet. GIS-centric programmers know enough to muddle through on their specific GIS product platforms. But they tend to specialize on products and its rare to find a GIS-centric programmer that is master on many different products. This is not to diminish the role of the GIS-centric programmer; they are critical to the success of exiting GIS implementations. No, its about the new web applications that potentially are going to access these GIS published datasets and geo-services that is the focus.

If we peer into the future its clear [at least to me] that increasing numbers of geospatial applications are going to arrive on the Internet for the broader community to use. These will cover a vast spectrum of functionality from simple points on maps to sophisticated geospatial analysis. We are already seeing some of this stimulated by Google, Microsoft and others. These are all geospatial applications requiring a web programmer to develop and publish. And sure, GIS users and custodians of data can and are publishing applications on the Internet that are available to the broader community. But they are typically what the GIS custodian wants the users to have and use and many look like GIS products which can be intimidating to non GIS trained users. Many people from the GIS industry defend this situation by saying “spatial is special” and to some degree they are right when you look at issues like cartographic projections, generalization, symbolization and heavy spatial analysis. However, these are issues that can be standardized or commoditized making them less of an issue to web developers – for example, who needs to geocode addresses anymore, this can be done in the cloud. But the real area of growth and potential is the unfettered availability of geospatial resources to an army of web developers who may or may not understand all the nuances of the geospatial world. This is not as farfetched as one would imagine and, if it is just around the corner, who will own the skills that will exploit this environment best.

By way of example, my son is a web developer with years of experience in developing websites and web applications with a variety of frameworks. He had no experience in the geospatial domain and he and I work together in a business that publishes the ABS Census data online as thematic map overlays [see www.numaps.com.au]. He has been able to build excellent geospatial applications that people can use in a GoogleMaps-like environment or users can ingest these overlays into their own GIS desktop via OGC web service requests. These applications could be considered to be GIS-like and can also be tailored to be function specific for targeted markets/problems – just like mashups. The concept of publishing data in the cloud in an open way is second nature to him. He understands open standards that makes data available to the broadest community possible. He’s open to experimenting with new open source tools etc. These are typical of attitudinal or cultural trend in the younger generations. A trend towards the belief that the Internet is a way to democratize the world in all things including geospatial. I’m sure those who work in Google have a similar view albeit monetarily supported.

The point is that these skills in web programming far out weight the lack in knowledge of geospatial issues which can be easily navigated. These skills and even the attitude towards the Internet as an infinite resource of information and services is hard to find in the traditional GIS-programmer. It is these skills that will realize the potential of a geospatial web. What is missing from this scenario is the availability of spatial data and services beyond those supplied by Google, Microsoft, etc.

Governments either consciously or not recognize that there is a huge pool of web programmers out there and have instigated several mashup competitions [spatial and non-spatial] with significant prize monies for them to show off what they can do with Government data. Whilst this has only scratched the surface of the web developer community it would appear that there were few entries from traditional GIS-developers even though a high percentage were geospatial mashups – not many of which looked like GIS products. The other thing is a lot of these were from individuals who most likely were from this younger generation consistent with their voluntary and democratic culture.

So its clear to me that the web programmer is important going forward in their own right. Its also important that if Governments want to harness this huge potential they start publishing their data on the Internet or give it to someone who can. Sure they will have to cover their backsides when it comes to liabilities; sure they will have to set expectations on service levels; sure they will have to improve the data quality and currency; and sure they will have to protect Government intellectual property. But they need to get on and do it so we can all start to tap into this resource.

GIS programmers typically have their hands full keeping enterprise systems operational which is still a very important task and for which they are more than qualified. But when it comes to the future of published geospatial data sets and online applications such as geo-mashups, it is my belief that the web programmer with an IT background will not only cope with the geospatial application area but has the potential to build more efficient, fresher, targeted and relevant applications for use in the broader community – spatial is not special when it comes to web applications.

Brad Spencer, General Manager NuMaps

 

From my experience, its clear that people trained in web development are quite different creatures to those trained in GIS development and their Internet components. Typically GIS trained people come from Cartographic and Drafting backgrounds where building spatial products is an important business activity and are typically not trained as software engineers like most IT programmers. I started out as a Cartographer and Photogrammetrist and decided that I liked computing so gravitated towards programming in the 1970’s. Fortran was the language of choice in technical computing and the Internet did not exist. But of course along came the many different GIS products and I like many others got engaged in software development at the scripting or macro level. This was simply to make the user’s GIS experience more convenient. Historically many people move into the programming world via this macro world on many different GIS platforms. This was simply pragmatics and was never a way to develop rigorous skills in programming.

The universities were all teaching software development skills like crazy back then as they needed to churn out programmers to quench the thirst of industry that was embracing computer technology at an alarming rate. Some of us mapping folks got retrained in this surge and some successfully migrated to the Computing profession, however not many people had both programming and GIS skills back then. But as the GIS marketplace grew there was a new brand of GIS user evolving, one that could not only use the software extremely well but could also modify these tools to do more – they needed only GIS product skills which were invariably proprietary.

When the Internet arrived all the GIS vendors started looking very closely at the technology but only adopted it when customers pushed them. They much preferred the controlled environment of a closed enterprise implementation. As the Internet developed further and as web service technology began to appear they all started to augment their products with Internet technologies. Then along came OGC and most of the major vendors joined in support albeit in some cases paying lip-service support only. This gave us a framework for development on the Internet in an interoperable way. Could the GIS-developer easily adapt to this new paradigm of web application development?

New graduates from university are trained on this type of technology. Not only do they understand how to build robust object oriented applications but they also understand the ‘glue’ that holds it all together in the Internet. GIS-centric programmers know enough to muddle through on their specific GIS product platforms. But they tend to specialize on products and its rare to find a GIS-centric programmer that is master on many different products. This is not to diminish the role of the GIS-centric programmer; they are critical to the success of exiting GIS implementations. No, its about the new web applications that potentially are going to access these GIS published datasets and geo-services that is the focus.

If we peer into the future its clear [at least to me] that increasing numbers of geospatial applications are going to arrive on the Internet for the broader community to use. These will cover a vast spectrum of functionality from simple points on maps to sophisticated geospatial analysis. We are already seeing some of this stimulated by Google, Microsoft and others. These are all geospatial applications requiring a web programmer to develop and publish. And sure, GIS users and custodians of data can and are publishing applications on the Internet that are available to the broader community. But they are typically what the GIS custodian wants the users to have and use and many look like GIS products which can be intimidating to non GIS trained users. Many people from the GIS industry defend this situation by saying “spatial is special” and to some degree they are right when you look at issues like cartographic projections, generalization, symbolization and heavy spatial analysis. However, these are issues that can be standardized or commoditized making them less of an issue to web developers – for example, who needs to geocode addresses anymore, this can be done in the cloud. But the real area of growth and potential is the unfettered availability of geospatial resources to an army of web developers who may or may not understand all the nuances of the geospatial world. This is not as farfetched as one would imagine and, if it is just around the corner, who will own the skills that will exploit this environment best.

By way of example, my son is a web developer with years of experience in developing websites and web applications with a variety of frameworks. He had no experience in the geospatial domain and he and I work together in a business that publishes the ABS Census data online as thematic map overlays [see www.numaps.com.au]. He has been able to build excellent geospatial applications that people can use in a GoogleMaps-like environment or users can ingest these overlays into their own GIS desktop via OGC web service requests. These applications could be considered to be GIS-like and can also be tailored to be function specific for targeted markets/problems – just like mashups. The concept of publishing data in the cloud in an open way is second nature to him. He understands open standards that makes data available to the broadest community possible. He’s open to experimenting with new open source tools etc. These are typical of attitudinal or cultural trend in the younger generations. A trend towards the belief that the Internet is a way to democratize the world in all things including geospatial. I’m sure those who work in Google have a similar view albeit monetarily supported.

The point is that these skills in web programming far out weight the lack in knowledge of geospatial issues which can be easily navigated. These skills and even the attitude towards the Internet as an infinite resource of information and services is hard to find in the traditional GIS-programmer. It is these skills that will realize the potential of a geospatial web. What is missing from this scenario is the availability of spatial data and services beyond those supplied by Google, Microsoft, etc.

Governments either consciously or not recognize that there is a huge pool of web programmers out there and have instigated several mashup competitions [spatial and non-spatial] with significant prize monies for them to show off what they can do with Government data. Whilst this has only scratched the surface of the web developer community it would appear that there were few entries from traditional GIS-developers even though a high percentage were geospatial mashups – not many of which looked like GIS products. The other thing is a lot of these were from individuals who most likely were from this younger generation consistent with their voluntary and democratic culture.

So its clear to me that the web programmer is important going forward in their own right. Its also important that if Governments want to harness this huge potential they start publishing their data on the Internet or give it to someone who can. Sure they will have to cover their backsides when it comes to liabilities; sure they will have to set expectations on service levels; sure they will have to improve the data quality and currency; and sure they will have to protect Government intellectual property. But they need to get on and do it so we can all start to tap into this resource.

GIS programmers typically have their hands full keeping enterprise systems operational which is still a very important task and for which they are more than qualified. But when it comes to the future of published geospatial data sets and online applications such as geo-mashups, it is my belief that the web programmer with an IT background will not only cope with the geospatial application area but has the potential to build more efficient, fresher, targeted and relevant applications for use in the broader community – spatial is not special when it comes to web applications.

Brad Spencer, General Manager NuMaps