Following up with a few points Sandro brought up while testing the previous release candidate...
- in LICENSE files I’ve seen that are listed all (single) images used, and > it’s Ok (I think too verbose), but isn’t it too much in detail ? > Actually, this is the recommended approach (as per a discussion in the general@ list several months back). Otherwise, users wouldn't know which licenses applied to the resources. > And for example if we miss one used image what are the risks ? > The licenses are managed as SVN properties to make sure that we don't miss an image. http://svn.apache.org/repos/asf/incubator/pivot/infra/bin/check-svn-props.shchecks for the existence of these properties. Then I run a shell command to spit out the list for each license before each release, ensuring that each release has the most recent list of files. > - in the source of many (generated) html pages, I see that for example in > pre tags, > > all < and > chars are generated (by xslt transformations I think) in the > html escaped way, that's Ok > > (browsers don't complain on them) but it's less human-readable (only > manually reading page sources) ... > Not escaping them breaks the DTD that those HTML pages use, so they must be escaped.
