when a summary report is generated from a large model (long time-series of data and reasonable number of nodes), MUSIC frequently runs out of memory. Summary reports are needed to generate treatment efficiency curves for guidelines and this constraint is a major nuisance. It can probably be easily fixed by closing the data used to calculate mean annual loads for each node before calculating the next one while the results are generated (I'm assuming they're all kept open simultaneously, it's about the only way it could possibly consume 3GB+ of memory).

Also, proposed upgrades such as those to facilitate calibration and upgrade the input/output functionality are sorely needed in industry. It would be good to hear some information from the research and development team and as users to provide feedback and suggestions for the next version.

From Dale Browne