We are working on creating an 'image' of WASCE (2.0), with the application already bundled. Initially we have a clean install of WASCE and deploy the application to it, then upload that to a server for the end user to download and use locally. We don't want to do that on every code change, so we just replace the corresponding modified (or add the new) jar in the expanded EAR structure in the repository entry. On a full repository entry change (means we created the full expanded EAR structure and replace that in the server) we had failures saying that the application couldn't be found. That's when knew about this config.ser, config.sha1 and config.id files. What I did then, was to hand over the files of another WASCE local installation I have (I had deployed the application through the Eclipse plugin, so that generated those files), to my surprise that made the application work, even though we all have very different install directories (I can see in the config.ser that there is a part that looks like the web.xml that has a context parameter refering to my local install) and mostly what surprised me is that my workspace was outdated, meaning that the EAR that was generated for the server version had more jars than my local one.
That made me think about the question on the subject.. is the config.ser ever used at all? or just some code checking that the serialization file is there? Thanks in advance Luciano
