Geographical Information System (GIS) was designed for engineers. That was good because engineers always work with standards which make interfaces open. Web Map Service (WMS) is one of the open interface standards. WMS focuses on maps portrayal via the Internet. A WMS client will request map graphics from the server which acquires, stores and prepares the maps using GIS back ends. A WMS server can be a full blown GIS engine or it can just be a gateway for other GIS engine.
Due to the GIS engineering perspective, many GIS implementation was done by hardcore programmers who work with C/C++ or Java programming languages. The legacy survives till today. The choice of programming language determines the type of server the GIS requires. Unfortunately, most GIS implementations only work with dedicated servers which incur high operating cost. A dedicated server need a special work force to overlook on the daily operation.
We have to be realistic if we want to bring GIS into the mainstream. Initial cost can be a ticked off. I have friends who keep on coming to me asking for GIS solutions. They simply could not bear the cost. I said to them over the years I had been trying to find some solutions that will reduce the cost. I even have an open source solution stack for GIS. However, my solution requires a dedicated server. A dedicated server may not be really an issue. What will be an issue is when they want to see a demo. They hadn't invested anything and they wanted to have a demo. And I have to get a dedicated server up, stuff the GIS solution stack and upload some spatial data just to show a demo, which nobody will pay. If I tell them that there are a lot more to do then they will go away. I just told them there were really a lot more to do to implement GIS.
There was a case when a dedicated server I was working on a GIS project was taken away.
How can I implement a GIS without a dedicated server? A cloud solution can still be expensive in term of time and money. How about a shared web hosting server? The PHP programming language and MySQL database are usually available on shared web hosting server. PHP has never been treated on par with C/C++ or Java. And the MySQL database will not be installed with any spatial indexing engine. Well, PostGIS for MySQL is not yet full blown. No, we are not allowed to have the PostGIS on MySQL in a shared web hosting server. Anyway, I built Camar using PHP and plain MySQL.
Camar is a WMS server with GIS back ends and should be able to run on any shared web hosting server with PHP and MySQL. I can now create GIS demos with no worries on missing servers. In fact, I can co-locate many Camar implementations on the same shared web hosting server.
Due to the GIS engineering perspective, many GIS implementation was done by hardcore programmers who work with C/C++ or Java programming languages. The legacy survives till today. The choice of programming language determines the type of server the GIS requires. Unfortunately, most GIS implementations only work with dedicated servers which incur high operating cost. A dedicated server need a special work force to overlook on the daily operation.
We have to be realistic if we want to bring GIS into the mainstream. Initial cost can be a ticked off. I have friends who keep on coming to me asking for GIS solutions. They simply could not bear the cost. I said to them over the years I had been trying to find some solutions that will reduce the cost. I even have an open source solution stack for GIS. However, my solution requires a dedicated server. A dedicated server may not be really an issue. What will be an issue is when they want to see a demo. They hadn't invested anything and they wanted to have a demo. And I have to get a dedicated server up, stuff the GIS solution stack and upload some spatial data just to show a demo, which nobody will pay. If I tell them that there are a lot more to do then they will go away. I just told them there were really a lot more to do to implement GIS.
There was a case when a dedicated server I was working on a GIS project was taken away.
How can I implement a GIS without a dedicated server? A cloud solution can still be expensive in term of time and money. How about a shared web hosting server? The PHP programming language and MySQL database are usually available on shared web hosting server. PHP has never been treated on par with C/C++ or Java. And the MySQL database will not be installed with any spatial indexing engine. Well, PostGIS for MySQL is not yet full blown. No, we are not allowed to have the PostGIS on MySQL in a shared web hosting server. Anyway, I built Camar using PHP and plain MySQL.
Camar is a WMS server with GIS back ends and should be able to run on any shared web hosting server with PHP and MySQL. I can now create GIS demos with no worries on missing servers. In fact, I can co-locate many Camar implementations on the same shared web hosting server.
Comments
Post a Comment