What do you think are the top current trends to watch in GIS Development? Which capabilities/tools do you think will be most in demand to both developers and end users in the future?
Web Services - for both data and tools. On desktop, web, or mobile apps. For users (transparently) and developers.
The amount of effort for each GIS project to collate data, import it to a geodatabase, trackdown metadata, add in updates is a huge cost in time and budget.
Well implemented WMS/WFS services (and ArcGIS web services) mean this only has to be done once and maintained by the owners. The national SDIs in Europe are a large part of getting these implemented.
Projects can then focus on project related datasets, but with a myriad of (hopefully current) related datasets added in in a few lines of code.
GIS "cloud" tools are not too far behind. For example why go to the trouble of buying routing data, installing and configuring a routing module on a server, customising and testing it when you can make a one line call to a dedicated web service and get better results.
I think low-tech desktop GIS analysis is one area that will be become very important. Google Earth really exposed many people to the visualization side of things, but is still hard to do a buffer and overlay analysis (technically and data-wise). I interact with many planners and non-profits and they struggle to do simple analysis of their problems using GIS due to technical hurdles. (One major flaw is that many of those who are familiar with GIS are educated in universities that only have ArcGIS; when they start working, only large companies can afford it, or they have the most basic version which does not allow raster analysis).
I think QGIS is one of the easiest to use desktop tools out there (but there are others), so hopefully in the future there will be a wide range of free or low-cost desktop GIS tools that people are trained on in universities, and that they can apply in their jobs.
In my opinion, another trend is that on client side end users will used web browser-based GIS applications, while users with a more technical or professional profile, will also use desktop, feature rich applications (gvSIG, GeoMedia, ArcGIS, OpenJump...).
I think this makes sense because end users need less features that GIS "power" users. Other reasons are the cost of non-free applications (paying a licenses for an underused software makes no sense) and the fact that web based GIS systems have no installation or deployment on client side.
A problem to be solved is access control to geographic information. When end users edit or add attributes to features, there will be the need of restricting or allowing to read, edit or delete sets of features depending on user role. As far as I know neither WMS nor WFS address this issue. An interesting question about this was post in this website: "WMS with Access Control (ACL) for different users? Or solution for thousands of points on web map?"
"Open" web products targeted for niche markets
I think we've all had fun putting maps on the web, the next generation of mapping products will consist of sophisticated apps geared towards niche market applications. Instead of these apps being closed (data, source, or access) though, they'd be "open" to some extent. Some will be opensource, open data, free use, etc that allow for PPGIs, GISci, researchers, policy makers, Engineers, technologists, and others to all communicate under one medium.
Imagine the following Scenario:
Designing a highway: Engineers in "real-time" can determine the best location of right of way while working with Environmental experts depending on environmental, first nations, flooding, wet lands, etc situations. This is crucial as the current situation involves a procedural / hierarchal approach, where the right of way is bought first, then the it'll be sent for review by other experts. A real time accessible GIS (system and data) will allow those professionals to work together.
Mobile Technical Applications: I imagine that there will be a great need for apps that involve sophisticated modeling, using mobile devices only, whether installed on the phone or through a web interface
I usually take the kiss method. Stuff like geotagging, Of course as mentioned by geographika the cloud services are only going to expand. I also see a huge increase in "personal" image processing. What I mean is that more sofware (especially open source) is making image processing and the extraction of data from sat or aerial photos more desktop centric. Or the lay-geotech is more likely to be successful learning how to accomplish these tasks now. "Whew!" I think I got it out. You also cannot count out advancements on the mobile/phone technology front. Stuff is happening so fast there that it is hard to track or keep a finger on it.
True 3D data models / datasets. Particularly in applications such as Hydrology where knowing the xy extents of an aquifer is certainly not enough to get a true picture of the volume, or solid geology, where provinces have volume and map overlap/undercut each other at different depths below the surface.
Very rash thing to say (and I may look back at this and be wrong), but IMO (and a lot of others) HTML5 will be the dominant language for Online Mapping APIs.
As a programmer who has used GIS from time to time, I am amazed at the number of geographers that stick to using text-based file formats. I am really looking forward to seeing tools that can converge a diverse set of data, but I think we still need to fix the problem of folks still tossing CSVs around. It is very inefficient.
In short, I think the next big thing is the next big thing from 20 years ago.