Date |
Type |
Reported by |
Description |
State |
Maintainer |
18.07.2000 |
UPDATE |
ml |
at startup nothing can be seen in the aquarium |
DONE |
ml
|
19.07.2000 |
ERROR |
rs |
sensors are counting things in range more than once |
DONE |
ml
|
19.07.2000 |
ADD |
rs |
SNNS-netfiles export of the ANNs |
DONE |
ml
|
19.07.2000 |
ADD |
rs |
add additional transfer-functions for the ANNs |
DONE |
ml and cb
|
19.07.2000 |
ADD |
rs |
add possibility to select an agent with the mouse |
DONE |
ml
|
19.07.2000 |
ADD |
rs |
show informations for the mouseposition |
DONE |
ml
|
19.07.2000 |
ADD |
rs |
add splines for each neuron instead of using only one for the whole ANN |
DONE |
|
19.07.2000 |
ERROR |
rs |
textfields in the settings-dialog (width/height) are set to width 0 when the setttingsdialog is resized |
POST |
|
19.07.2000 |
ADD |
rs |
add the following things to the settingsdialog: min. energylevel for mitosis, min. energylevel for meiosis, min. energylevel for pheromone output, max. phermone amount/pos, max. substrate amount/pos, light distributon function, light cycle function, number of input neurons, number of output neurons, numer of total neurons, numer of cubic splines points, radius for pheromone output, transluency factor for agents, energy consumption for standstill and substrate for dead agent |
DONE |
ml and cb
|
19.07.2000 |
ADD |
rs |
add names to the different units of the ann (in the SNNS-net files) |
DONE |
ml
|
20.07.2000 |
ADD |
rs |
add an additional line in the statusbar area |
DONE |
ml
|
20.07.2000 |
UPDATE |
rs |
improve display of different world elements (phermone, substrate, light) |
DONE |
ml
|
20.07.2000 |
ADD |
rs |
no substrate visible/available at startup |
DONE |
ml and cb
|
20.07.2000 |
ERROR |
rs |
the lightworld is not calculated properly |
DONE |
ml
|
20.07.2000 |
ADD |
rs |
agents should be partialy transluent (not dark below) |
DONE |
ml and cb
|
20.07.2000 |
ADD |
rs |
add tags to the statistics (for sensors for example) |
POST |
|
20.07.2000 |
ERROR |
rs |
sometimes agents are drawn and placed at the same pos |
DONE |
ml and cb
|
20.07.2000 |
ERROR |
rs |
consoleoutput shows wrong agent number |
DONE |
ml
|
21.07.2000 |
ERROR |
rs |
when mitosis/meiosis occur, the consoleoutput should show the indices of the involved agents |
DONE |
ml
|
21.07.2000 |
ADD |
rs |
the ann output should be displayed in the infowindow |
DONE |
ml
|
25.07.2000 |
ADD |
rs |
add checkbox to toggle the display of light |
DONE |
ml
|
27.07.2000 |
ERROR |
rs, ml, pe and cb |
agents are moving into the same direction at the start of the simulation (only the initial ones) |
DONE |
ml and cb
|
27.07.2000 |
ERROR |
rs, ml, cb and pe |
sometimes an outofmemory-error occurs (even with 256MB of RAM) |
POST |
|
03.08.2000 |
ERROR |
ml |
sometimes the ANN has connections from output-neurons (only seen from node 22 until now) back to others and from inputneurons to node 17 (an inputneuron) |
DONE |
ml and cb
|
03.08.2000 |
ERROR |
ml |
the ANNs have very often only one connected output-neuron |
INPROG |
|
04.08.2000 |
ERROR |
pe |
if the agent consumes light, the program doesn't take care of how bright it is, instead a const is consumed |
DONE |
pe
|
04.08.2000 |
ADD |
pe |
it would be nice, if the agents are drawn colored according to their current energylevel |
DONE |
ml
|
09.08.2000 |
UPDATE |
rs |
change the substrateworld to a more realistic model (substrate should dissolve into the water, instead of coming up from bottom), the user should be able to choose between substrate coming up from the ground and instant, homogenous distribution of substrate |
DONE |
ml and cb
|
09.08.2000 |
ADD |
rs |
make the FOV a line if an angle of 0 for the FOV is specified |
DONE |
ml and cb
|
11.08.2000 |
ERROR |
pe |
the substrate world contains a little bug (if no more substrate left, it's endless regenerated) |
DONE |
pe
|
16.08.2000 |
ERROR |
cb and ml |
the pheromone at a certain position resolves to values below 0 (in fact it's endless resolved) |
DONE |
ml and cb
|
23.08.2000 |
ADD |
rs |
the number of crossover-points should be selectable |
DONE |
cb and ml
|
23.08.2000 |
ADD |
rs |
make the number of cubic-spline nodes changeable |
DONE |
ml and cb
|
23.08.2000 |
ADD |
rs |
add mutation rate and crossover rate to the settingsdialog |
DONE |
ml and cb
|
23.08.2000 |
UPDATE |
rs |
add a seperate transfer function to each neuron in the ANN (except the input neurons) |
DONE |
ml and cb
|
23.08.2000 |
UPDATE |
rs |
add a bias to the ANN to make it more realistic. this bias should mutate and be inherited to child agents |
INPROG |
ml and cb
|
30.08.2000 |
ADD |
rs |
to speed up things make debugging optional (debugging should be implemented in a class with different debugging levels) |
POST |
|
30.08.2000 |
UPDATE |
rs |
to speed up things the combobox for selecting agents should only be enabled if the simulation is stopped |
POST |
|
04.12.2000 |
ERROR |
ml |
the wrong agent id is written to the statistics file |
DONE |
ml and cb |