NOBLE APE MAILOUT - MARCH 2003
This month has been one of the busiest in Noble Ape's recent history! It featured a new version release, a change of webhost, a BBC radio interview and a substantial increase in the viewers and the types of viewers looking over the Noble Ape site.
NOBLE APE SIMULATION 0.659
The Noble Ape Simulation 0.659 was released earlier this month. This was a wide-coverage release and as this mailout is sent, it has been downloaded more than 4,000 times thanks to Sourceforge, Freshmeat.net, macupdate.com, macosxapps.com, download.com, zdnet.com and info-mac.
The more I looked, the more download sites I found that were willing to direct their users to the Noble Ape Simulation Carbon version and the new manual. The speed and veracity of the downloads was quite overwhelming.
The release of version 0.659 provided a litmus test for the press release in many ways and also introduced a sample audience to the Simulation. On the actual application itself, the feedback has been positive through open source developers and mixed from the commercial-download-site public. The Simulation was hosted above Ghost Recon on download.com. I am a great admirer of the Tom Clancy/Rainbow Six series games - but you can't get anything more different than the Noble Ape Simulation.
In software terms, the big difference between 0.658 and 0.659 was low memory handling. A dialog-warning was added when memory got low. Also, a fix around the colour window allocation was introduced to stop memory from running out before the colour windows hit the screen in low memory conditions. The big addition for 0.659 was the revised manual which brought the manual up to pace with Ocelot and the four window interface.
OPEN SOURCE EXPANSION
After some deliberation and introducing a number of active links from the Simulation page, I have expanded Noble Ape's use of Sourceforge to cover feature tracking, bug tracking and additional release hosting, as well as CVS. The Noble Ape Simulation was listed on Freshmeat.net, a Sourceforge-connected open source site, and substantial traffic came through with this listing - roughly 550 in the first day, and still at least forty per day.
The active integration with Sourceforge/Freshmeat of the Noble Ape Simulation and progressively other aspects of the development, such as the iON and potentially noblemak(e) and logtrack, bring Noble Ape into mainstream open source review. My former webhosts were not prepared for the Freshmeat.net volume of visitors and my email went out completely throughout the first and third day of the Freshmeat listing.
The Simple-End-User-Linux Educational software group (seul-edu) has raised a request amongst its members for a port of the Noble Ape Simulation to Linux (with my assistance). This kind of contribution is made possible by the developments of Noble Ape being shown in the broader open source community. Although no developer has stepped forward to assist with the port.
The viewing patterns of the open source developers/users was quite different to the viewing patterns of people coming from commercial sites or other links. Most of the users from Freshmeat reviewed the documentation on the graphics, introductory programming and the history of the Simulation. Few read the source log and few read the mailouts. In contrast, a general user would look at the log information, if they got to the front page. Many are now initial directed to the Simulation page.
More open source developers and users coming to the site, has changed the average site viewer's profile considerably. The documentation written over the past three months now needs to be updated inline with the source code. Passive documentation - such as the source log - appears not to be referenced. Part of this is a problem with passive linear log information. It is very hard to cross-reference. The information in the log describes the development historically. If you wanted to look at the early development of Ocelot, much of this information is now totally out of date.
So both with regards to current information and linked referencing the source log fails. I am looking at adding more active indexing to the logs with active keywords that are indexed passively with the source log. Logtrack, the software for creating the source log and barbalet.net, is active and open source. Changes can be made.
On the programmer documentation, I think the solution is to move some programmer HTML files into source control. Thus, treat the HTML documentation as source code in the development of the Simulation. This parity between the code and the developer documentation is currently tracked by a Sourceforge request.
PLANET NOBLE APE
Recently, I dusted off and colourised a version of Planet Noble Ape. This should be online in the next week or so. Developing Planet Noble Ape together with the Simulation is quite important. An objective of the next year is to integrate the two environments creating a Simulation environment over the entire Planet Noble Ape.
As with the Simulation, the visualisation in Planet Noble Ape gets equal weighting for development (and processor time). The dream is to replace the map window in the Simulation with the sphere of Planet Noble Ape. Surprisingly, this requires more integration and development on the Simulation end, but this has already started by separating land form and biology (0.657), object based simulation components (0.658) and improved memory management (0.659).
SCIENCE OVER SOFTWARE
As posted through the mailout, I appeared on BBC Radio 4's Material World program last week. It was interesting to talk about Noble Ape in terms of science rather than software. Much of the development of Noble Ape in the past couple of years has been about modernising the software.
Talking about the software in terms of science rather than improvements in the user interface was refreshing. It made me think quite critically about the development of the science in the Simulation rather than the latest interface improvements.
Ric Colasanti, the digital ecologist on the program, mentioned before the program that the interface to the Noble Ape Simulation was very well refined when compared to the standard scientific simulation tools available. Ric worked with Steve Grand OBE and Cyberlife Research, and said what I was doing with Noble Ape was what they wanted to do with Cyberlife Research.
In contrast, to the questions of science in the development, I received a number of emails following the program asking for a Windows version of the Simulation. I have been working this weekend to get a GPI Windows port ASAP.
Hope all is well with you all,
Tom Barbalet, 29 March 2003.