[Live-demo] new apps for 4.0

Hamish hamish_b at yahoo.com
Thu Mar 25 06:55:27 EDT 2010


Alex:
> > I'd like to formally invite all projects that are on the Live* to
> > become at minimum OSGeo Labs projects to encourage their participation
> > in OSGeo and make a soft rule that being an OSGeo Project, Incubation
> > Project or Labs project is the only way to guarantee a spot on the
> > disc (note becoming a labs project is really easy).
> >
> > Now that we've reached our current size we can't just keep adding, as
> > it won't fit on common media beyond what we've got and there needs to
> > be some way to pick, filter, and trim.

I have updated the  http://wiki.osgeo.org/wiki/Live_GIS_Build  wiki page
a bit to better describe our packaging priorities (official osgeo projects
get first chance at slots, bloaty things risk being left out, etc.).

At some point we should try an do a more formal analysis of how much
(compressed) space each package is using on the disc to get a better
idea of what the hogs are. I'm quite confident that I can find ~ 500mb
(uncompressed) to remove in a crunch. There's a bunch of low hanging
fruit we haven't touched yet.

To start with it would be good to make a wiki page with a list of the
included datasets (and their sizes).

Having a USB version which is identical to the DVD version sans the
Mac/Win installers seems like it will work well, especially because of
the minimal extra effort required to produce it that way.


> > We should work more with the Board on this idea of approaching
> > projects to strengthen ties to OSGeo.

If you like. I was looking at it as a simple matter of having a few seats
left in the car and offering some folks who're heading to the same
direction a ride. As long as they tick the "open source" (osi approved)
and "geospatial" boxes I'm happy to just call it informal "outreach".
shrug


Cameron:
> Hamish, these two disaster support projects do look very interesting, 
> and I expect will be good additions to the LiveDVD.
> 
> I also agree with Alex that we are in a position where we will need to 
> start putting limits in place, are we are starting to hit our disk
> capacity limits, are we are starting to gain a level of quality which
> is very valuable and will take managing to maintain.

It is possible (and desirable) to have some functioning scripts around
which are operational but not on the main disc. Then you could quickly
build a number of remixes for e.g. desktop gis for the classroom; field
work; nautical edition; disaster response; technology preview (aka beta);
etc. just switch out the install_list as wanted & rebuild.
 
> As such, I think we need to revisit our guidelines for responsiblity
> for projects on the list, possibly categorising projects between:
> "Core / Stable / Quality" projects, who meet all our guidelines
> and "Alpha / Beta" software, which is labeled as such, and
> doesn't tarnish the name of the established projects.

See what you think about what I wrote about beta software on the Build
wiki page. Categorizing by "quality" can be a rough business. :)
This probably isn't the venue for alpha software (moving targets are a
nightmare to maintain/support).


> I suggest that we discuss and refine these guidelines over the next 
> month or so, probably in conjunction with other OSGeo committees, in 
> particular the marketing committee. Once refined, we can go forth to 
> projects asking them to align with the guidelines.

If you like. My general take is to keep it simple: osgeo-aligned projects
get first dibs, then 3rd party projects which bring pre-packaged quality
software and have shown an active involvement.


regards,
Hamish



      


More information about the Live-demo mailing list