I came to the conference with precious little knowledge about G-cloud, except from a few pages that I had read (e.g. John Suffolk and Cloudbook) but just an impression that perhaps G-cloud by definition should/could be totally open (open source, open data, open data structure, open ….).
Why my interest?
Well working in the public sector I have become increasingly aware that our ability to be agile is frequently hampered by the software systems that we use. Specifically, most of these systems cause problems due to vendor lock-in. The main impacts of vendor lock-in is that we are dependant on the vendors for development (e.g. we have been waiting for 16 months for a working connector between our casework system and ePlanning) but also it is extremely difficult to switch between vendors . The other issue is that in Scotland the legislation etc. is generally different and we often end up with software that is ‘Scottishised’ i.e. usually does not quite meet our needs.
So our current situation is that our systems sit within our network, difficult to connect to the outside world, difficult to create additional functionality …. you get the picture.
My dream … our future systems would probably be outside our network and would be modular, so that all users use a core business logic but can ‘pick and mix’ the modules to bolt on to the core. The data would also meet agreed Open Data Standards/Structure (see PODS), which would enable us to easily switch vendors and if necessary develop functionality in-house. Also other user groups that interact with us (e.g. Community Councils) could also more easily develop functionality linked to our systems.
So when I heard about G-cloud I wondered if it might be my ‘dream’ solution. Could this deliver what I want (an end to vendor lock-in) and help develop shared services?
The discussion that followed (in no particular order) was really interesting:
- There was consensus from the other people working in the public sector that vendor lock-in is a real problem (i.e. it isn’t just me!!)
- Open Data Standards may be too restrictive and perhaps a better approach might be the Semantic Web (Ontologies, RDF …)
- Business Logic can be proprietary (businesses need to make a profit) but can be small
- Systems have ‘got to be open’ (i.e. Open source is what we should be striving to use but open APIs etc….) but there was some debate about issues
- Security may or may not be a barrier (not conclusive)
- The code could be ‘Government source’
- Procurement policies in relation to open source.
- There are other models for developing public sector systems (e.g. SEEMIS)
My take when I went to the conference was that we should push the UK government to ensure that the G-cloud can meet our Scottish needs (i.e. componentised and open so that it be developed).
However, my impression was that perhaps there is not much appetite for taking on the UK government (it’s probably just me). So perhaps what we could do is start by getting a few Scottish public sector agencies together and set up our own ‘cloud’ for testing purposes. This might then provide an opportunity for developing our own Government Open Systems (GOSCON might be a starting point)? Who knows, our existing vendors might then be interested in developing more open solutions?
To sum up, this was a great opportunity for me to air some of my daft ideas and fortunately I wasn’t totally shot down 🙂