Re: [Gml4j-devel] IRC discussion for opensource GML development

[Cameron S. <csh...@bi...>]

Comments inline:

On Thursday 21 February 2002 08:48, Aleksandar Milanovic wrote:
> Hi All,
>
> I wouldn't mind participating on the conceptual level, but I will have no
> time to do programming myself (unfortunately).

I expect your knowledge of GML is better than ours, and some well place 
advice from you at the design stage could save us lots of work downstream.

>
> A few comments:
> 1. The freefs project, VFNY's open-source WFS hosted on SourceForge,
> already covers parts of the goals here. It is a WFS and has a SAX GML
> parser, although I have no clue if the parser is exactly what you had in
> mind. Also, check the license, it may be different from yours.

Yes, you are right on track.  The freefs project has been renamed  GeoServer, 
and Rob H has already been joining us for discussions.


> 2. Some IRC time slots are inconvenient for me. For example, 8am in NYC is
> 5am in Vancouver.

The timeslot we are looking at is 10am in Vancouver.  Is that OK by you?
See, the following meeting planner URL:
http://www.timeanddate.com/worldclock/meetingtime.html?day=27&month=2&year=2002&p1=136&p2=240&p3=262&p4=179&p5=256

> 3. Although I am an author of GML4J, I am not convinced that it should be
> extended as is. The current implementation is somewhat inefficient in that
> it depends on the DOM tree (or rather sits on top of it). It'd be better to
> have GML objects that are independent of other object models, but can be
> generated, perhaps, from other object models such as DOM. In this sense,
> GML SAX parser would be useful.

OK, it would be good to have your ideas on this.