In this paper we introduce two project evolution indicators. One is showing an increase of downloads of the project and therefore a growing interest of users in the results of the project. The second indicator is predicting the future evolution of the project with the submission of new revisions to the concurrent versioning system. Both indicators can provide evidence of the sustainability of a software project. We used the General Linear Model method to statistically formulate the two linear equations that can be used to predict the two indicators. The predicting equations were build by using two stratified data samples one of 760 projects and the second of 880 projects extracted from the SourceForge repository. The six metrics included into the final version of the two models were extracted from a set of thirty project and product metrics as: the number of downloads, the number of developers, etc. We have validated the discriminant and the concurrent validity of the two models by using different statistical tests as the goodness-of-fit and we have used the two models to predict the indicators on two hold-out validation samples. The model predicting the increment of downloads was correct in 75 percent of the cases, the model predicting the submission of new revisions was correct in 93 percent of the cases.