On Monday 05,April,2010 09:09 AM, Alexander Kojevnikov wrote: > Hi list, > > There is a blog post [1] and a discussion on HN [2] explaining how > long headers in source code files are not really necessary. After > discussing it with Gabriel on IRC he suggested to post this here. > > So the idea is to get rid of the headers. > > * Duplicating the license in each file is excessive and has no legal meaning. > * Git does a much better job at telling who is the author of a file > than the list of authors in the header. > * Same for the years in which the file was developed. > > Gabriel proposed to replace all headers with just this: > > // This code is released under the MIT/X11 license; see COPYING > > Any objections or insights? > > Cheers, > Alex > > [1] http://hackerboss.com/get-rid-of-templates/ > [2] http://news.ycombinator.com/item?id=1239810 I object to this, unless you plan to document the copyright holders and their respective years of each and every file in COPYING. Files can be lifted from other sources, and excluding the original copyright holders from these files is unfair, and in some cases, not legal. Git does a better job of telling who *modifies* what and when, but cannot/does not tell who owns the copyright of the file. In some cases, the person who modified the file does not own the copyright to the file (take for example, employees of certain companies). Above all, this copyright information that you intend to remove from the files will not present in your tarballs, which is eventually what you release to your end-users and downstream distributions. Please just leave the headers intact. You'll create hell for downstreams otherwise. -- Kind regards, Chow Loong Jin (GPG: 0x8F02A411) Ubuntu Developer
Attachment:
signature.asc
Description: OpenPGP digital signature