to, 2009-06-11 kello 15:01 +0200, Giacomo A. Catenazzi kirjoitti: > I think we need: > - one tool that generate the new copyright files. People forget to check > and update files; and the non-tiny packages need such tools (if we need > the DEP5 format). (the tools as an helper, ev. overwritten by maintainer > decisions)
Writing a tool to automatically extract copyright license info from source trees is pretty difficult to do, if you want it to be reliable. I suggest we don't need it to continue the discussion. > - one or more tools that decode the new format. I agree that it is useless > to have "it can be parsed automatically" if nobody "parse it > automatically". If DEP5 gets into a reasonable shape, I'm going to write a tool (and a Python library) to query debian/copyright files. I don't have any firmly held opinions on the format or content of the files, though, except that I don't particularly feel the need to mandate inclusion of more or more detail information than what is in the files already. However, I don't want to write the tool as long as the spec keeps changing whimsically. > - and a reason That's the killer point we should concentrate on. I know commercial derivatives of Debian can benefit from machine-readable debian/copyright files: their customers may need to get a list of licenses used in the (subset) of packages the derivative provides them, and this would get easier with DEP5. For Debian itself, this is not sufficient reason to bother. Would Debian benefit from being able to easily query for things like "packages linking to OpenSSL, licensed under GPL, but without an exception"? -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
