[Live-demo] 5.5 Beta 1 Status

[Gavin Treadgold]

On 2012-02-02, at 10:28 , Hamish wrote:
> Fran wrote:
>> In there is the almost-trunk version which I
>> think works fine for this purpose: a chance to
>> preview functionality.
> we like to use tried-and-tested versions (warts and all) over latest-greatest versions as there are
> less unknowns to deal with.

I'd actually say it is the opposite in Sahana Eden's case. Deploying an old release will have an increasing number of warts compared to the later versions, and any support issues will almost be pointless to deal with as the code base will have moved on. The real problem is that we're not packaging up releases regularly that fit nicely with the OSGeoLive release schedule, instead focusing on development. That said, we're not going to deploy bleeding-edge trunk either, which is why Fran chose an 'almost-trunk' version. I trust Fran's choice as being appropriate. This also brings benefits of better geo integration as Fran points out.

>> If we were going to spend more time then it should be on setting up suitable prepopulate demo data,
>> but this is certainly more time than I have available currently.
> 
> I'd certainly encourage Sahana to take advantage of the live disc as a general purpose already-set-
> up-for-you demo and educational tool, but understand too well about time pressures. I'll just say that
> the opportunity is there if you want to use it.

However, to build up a sample scenario, and relevant data, I will likely need to focus on creating it for a set geographical area, e.g. it will likely be data that will be fixed from release to release, as we need to have fixed locations for things like evacuation centres etc. It won't be able to be changed for each move of the underlying geodata to new OSGeo conference locations.

I may look at getting it based on real data from the Canterbury/Christchurch set of earthquakes as I can include things such as the 10000 or so earthquakes that we've had. It is far easier to build up this information post-event, than to create mock data, and we have a pile of data about what has happened in Christchurch that is accessible and open, including post-earthquake raster imagery. The problem here is of course that space is tight enough already on the Live image to probably allow this.

> fwiw one of my main motivations for putting sahana on the disc in the first place was not to get it
> more deployments, rather to get it into the hands and under the noses of the collective osgeo
> developer brain trust, to maybe attract some new devs to your projects

Absolutely, and this makes me more comfortable that we have chosen the right option of a stable-enough, near-trunk version - particularly as we're focusing more on geodevs than end users. Not to mention that it is a better way of trying to keep up with the state-of-the-geo art that we're trying to promote. We should be promoting new features to geo developers, ahead of relying on stability in this sort of demonstration.

Hopefully we're all on the same page now :)

Cheers Gav