The Geomnesia-system has to be platformindependent, the mapdata has to be vectordata to keep the transfertimes low. There has to be the possibility to fade specific kinds of routes and mappoints in and out. There has to be the possibility for smooth zooming of the map and map printing/exporting. There has to be the possibility to export the routes and mappoints to different file-formats. There has to be the possibility to import the routes and mappoints from different file-formats. That means the fileformats of other map- and routplanning software and the fileformats of GPS-devices. More and more motorcycleriders, bycicleriders, hikers etc. are nowadays logging their tours with GPS-devices to share them afterwards with webcomunities or to store them in their own archives. All these peoples have to able to effortless transfer their material from and into Geomnesia (platformindependently).
The technology of vectorized mapmaterial already exists, among others the company Mapsolute GmbH (u.a. www.Map24.de), offers such a system. Because Mapsolute gives access to their APIs, the system is always expandable. Because the mapdata of mapsolute is constantly updated there has to be a functionality in Geomnesia, that archives "old" local circumstances (mapareas) together with waypoints/routes, to avoid senseless Infopoints at waypoints/routes, that don't exist anymore.
What Geomnesia adds to that as corefeature, is the possibility to refer to "geo-objects" (mappoints/routes) via hash-code. So you create a point and get a unique combination of characters/numbers, for instance: wr-A45gZ3b. This hash-code leads by input/referring inside the geomnesia-system to a dedicated, single "geo-object". From there you can open additional informations. It has to be possible to link from extern sources to these points (for instance: www.geomnesia.de/card?wr-A45gZ3b).
Addendum:
An alternative for producing the mapdata (but this would be a total seperate project) would be
a kind of "open-geodata-project". Because free useable mapdata is not available (and that although
land surveying offices got the mapdata paid from taxes; except Osanbrücks FRIDA;
the USA are ahead of us, mapdata produced from tax-money has to be made available to population there)
and many people own GPS-devices with logging-function, a kind of "networked-germany-surveying"
would probably be quite exciting. Each participant would upload the routes, that are already on his
system and step by step this would build the streetmap of germany. If one would couple that with
a kind of ranking system á la SETI, this could quite quickly assemble the important streets.
The verification of the Data would probably a problem, and exact positioning too, but that should
be solvable with more or less big effort. Maybe the Wikipedia-way (better geopedia ... OK,
"Geopedia" already exists, but they
don't occupy with streetdata; so more likely "Routepedia" or "Streetpedia").