Juan Jose Pablos added a comment - 10/Mar/04 19:05 Have we got any example on how to do it on a XML file? Cheers, Cheche

David Crossley added a comment - 10/Mar/04 20:33 Yes. For an overview of the various tools please see the file: cvs/committers/relicense/README.txt

Nick Chalko added a comment - 11/Mar/04 02:13 Example of license formats area available at http://wiki.apache.org/general/LicenseFormats

Dave Brondsema added a comment - 11/Mar/04 11:56 http://nagoya.apache.org/wiki/apachewiki.cgi?Licensing has descriptions of licensing. based on discussion on licensing@apache.org it seems that if it's ASL it's fine. if it's less restrictive than ASL it's okay but needs to be referenced properly. if it's more restrictive than ASL we shouldn't distribute it (via tarballs or SVN) at all

David Crossley added a comment - 11/Mar/04 15:51 Here is the result of running the script from the "committers" CVS repository called "insert_licence.pl" in practice mode ... The attachment "forrest-license-20040311.log" is the listing of every file processed with statistics about each file. You can use 'grep' to find certain patterns. Below are the high-level stats from today's run ... ------ ./insert_license.pl -p /usr/local/svn/forrest 2002-2004 > forrest-license-20040311.log   Doing practice run. No files will be written   Total 782 text files were investigated. New licenses would be inserted in 367 files. Skipped 195 files with an existing license: (ASL v2.0=10, v1.1=108, v1.0=0) (Other=76, PublicDomain=1) Skipped 11 XML files with missing XML Declaration. Skipped 287 files of unknown type.   List of unknown filename extensions: (Add new fileTypes to this script if you want them to be catered for.) .aart=1 .ai=1 .cat=1 .classpath=1 .cwiki=2 .dcl=1 .dtd=43 .ehtml=2 .ent=55 .g=2 .grm=4 .html=17 .ihtml=1 .jsp=18 .mod=24 .old=1 .org=1 .pdf=1 .pen=6 .pl=3 .project=1 .py=1 .rdf=1 .rnc=1 .rng=5 .rnx=2 .save=1 .tld=3 .txt=50 .types=1 .xcat=2 .xul=1 NoExtension=21 ------

David Crossley added a comment - 11/Mar/04 15:56 Gzipped log file from today's run of committers/relicense/src/perl/insert_license.pl in practice mode.

David Crossley added a comment - 11/Mar/04 16:16 Added a new document at forrest/etc/relicense.txt to list some other notes, and especially list some files that we need to ensure do not get a new license accidently inserted.

David Crossley added a comment - 16/Mar/04 13:06 The new license has now been added to most files. There are still some issues listed in etc/relicense-tree.txt and there are still the other tasks shown above.

David Crossley added a comment - 16/Mar/04 16:47 ./insert_license.pl -p /usr/local/svn/forrest 2002-2004 > forrest-20040316.log   Doing practice run. No files will be written Total 787 text files were investigated. New licenses would be inserted in 12 files. Skipped 646 files with an existing license: (Apache v2.0=586, v1.1=12, v1.0=0) (Other=47, PublicDomain=1) Skipped 0 XML files with missing XML Declaration. Ignored 108 files of specified type (.txt .dcl .ent .pen) Skipped 21 files of unknown type. ------ The 12 insertions will not be done. They are foreign files, e.g. some of the DocBook DTDs do not have a license statement. The 12 files with v1.1 license are those in ./legal directory that still need update. The 21 files of unknown type include some files which have been done manually such as shell scripts with no filename extension, and some other obscure files which have not yet been attended to (see relicense-tree.txt).

David Crossley added a comment - 15/May/04 14:25 This task is not yet finished... A) Removal of author tags and attribution of those people in status.xml B) Review the legal/directory one more time before release. C) There are some other minor things (apart from B) mentioned in etc/relicense-tree.txt

Dave Brondsema added a comment - 15/May/04 15:00 I don't think we should remove names from status.xml. The rationale for removing @authors was to remove implied ownership of code for 1) legal reasons and 2) so people don't feel hindered to work on "someone else's" code. status.xml reflects significant modifications to Forrest, which do not imply any ownership (because it is task-oriented not code-oriented and is stored seperately from the code). It provides us a way to see who has done what so we can go back to that person with questions if we need to, and also to give some credit.

David Crossley added a comment - 15/May/04 15:11 Arghh, sorry. My comments must have misled you. Let us try again ... A) Removal of author tags from code (java, xml, etc) and ensure that those people who were listed there, are properly attributed in status.xml

David Crossley added a comment - 16/May/04 15:19 Created a list of "author" tags in the "forrest" source repository and added it to the private "committers" module. Here is the summary ... 821 text files were investigated. 136 files had some type of author tags ... 77 java-style tags 9 xml-style tags 85 cvs-Id-style tags

Dave Brondsema added a comment - 29/May/04 01:20 Locally I have removed all author stuff. How should we credit these people in status.xml? There's a lot :-)

David Crossley added a comment - 29/May/04 06:56 I too am not sure how to deal with the attribution in status.xml My plan was to just remove all of the tags, then gradually review status.xml using the list of authors that we saved. We will not be able to refer to everything, but we should be able to mention the main stuff.

Dave Brondsema added a comment - 29/May/04 11:23 now it's done... unless I forgot something else

David Crossley added a comment - 29/May/04 13:04 Yippe, this is fun ... close, re-open, ... Sorry, there is still some stuff not yet done. There is more work in the "legal/" directory to ensure that licenses are appropriate and each jar has a matching license.

David Crossley added a comment - 06/Jun/04 18:09 It seems that the re-licensing job is now done. We can run the tools again at any time to ensure that no new issues have crept in (see cvs://committers/relicense/src/perl/) .