Brisbane Meetup of Aust-NZ OSGeo Chapter

In a fit of ill-advised post FOSS4G enthusiasm I set up a meetup page. In my defence it sounded like a lot of the fun to be had in open source spatial is happening at the OSGeo chapter level.
I was actually concerned that I was taking on too much; that whole feeling of throwing a party and wondering if anyone is going to show...
Turns out we had quite a crowd; and an excellent location looking over the Brisbane river:

Rather than try and provide a summary the first Brisbane "Hack and Yak" here are the links brought up on the big screen:

• Release Notes - GeoTools - Version 8.0-M2
• Gource Controls
• CommonMap
• ozBlog - The online portfolio of Levi Putna
• Asset Management Solutions for Asset Intensive Industries | Mincom
• OpenScales
• OpenStreetMap-in-a-Box - Open StreetMap Wiki
• Welcome to the Quantum GIS Project
• NGIS :: Home
• Home - Brisbane City Council
• spatial@gov :: Home
• OSDC 2011 | Nov 14-18 Canberra
• YOW! 2011 Australia Developer Conference
• AusGOAL
• Discuss -- OSGeo Discussions
• Australian Ocean Data Network Portal
• BlueNet
• OSGeo Aust-NZ Open Source Geospatial Chapter (Brisbane)
• ELOGeo Repository
• Queensland Government Information Service

If any of that looks like fun you are welcome to join us on Friday Nov 11th. I will be on hand with a preview of an upcoming GeoTools Workshop. I also hope to start round 1 of uDig vs QGIS challenge (to try and encourage some creativity).

FusionIO IODrive OSM

I always run into people doing interesting work at LISAsoft - this one was startling enough I decided to share.

A bloke here on the hardware side of the street has hooked up a FusionIO IODrive to OSM.
What is that thing? Terabytes of virtual memory backed by happy little clustered NAND chips.

What happened?

• To pre-seed the top zoom-levels of OSM using mapnik are *close* to 50% quicker (based on tests done seeding a subset of the world).
• Lower zoom levels seem to be more CPU limited at this stage
• The IODrive should allow for many more concurrent parallel seeds to occur before we saturate IO

Yes this post has low open source content - but golly that is just so fast. And I don't want to think what one of those costs. Think we could enter it in the WMS shoot out - or is that like handing contestants a motorbike for a marathon.

Aside: This same hardware group was interested in donating hardware to an open street mapping party for FOSS4G but so far nobody returned their call? Can we make introductions or something?