Closed Bug 285347 Opened 20 years ago Closed 18 years ago

files in java-sdk still licensed under the NPL

Categories

(Directory :: LDAP Java SDK, defect)

Product:
Directory
Component:
LDAP Java SDK
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

(Not tracked)

Status:
RESOLVED FIXED

People

(Reporter: fitzsim, Assigned: mcs)

Details

Attachments

(4 files)

Script for getting CVS authors
172 bytes, text/plain
Details
List of CVS authors
199 bytes, text/plain
Details
Script to scan source for e-mail addresses
1.59 KB, text/plain
Details
List of contributors
221 bytes, text/plain
Details 
User-Agent:       Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.5) Gecko/20041228 Firefox/1.0 Fedora/1.0-8
Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.5) Gecko/20041228 Firefox/1.0 Fedora/1.0-8

Files under java-sdk contain the NPL header.  Is this correct, or were these
files missed during the mass re-licensing?


Reproducible: Always

Steps to Reproduce:
1. open a .java file in java-sdk
Actual Results:  
NPL header displayed.


Expected Results:  
MPL/GPL/LGPL header displayed.
The LDAP C SDK (mozilla/directory/c-sdk) and Java LDAP SDK
(mozilla/directory/java-sdk) code were originally released under NPL, like much
of the Mozilla code.  Relicensing efforts have typically been led by Mozilla.org
leadership in cooperation with module owners such as myself.  The LDAP C SDK
code was relicensed to use the triple license a while ago, but the Java code was
skipped because it does not ship inside any current Mozilla product.

Is relicensing important to you?  Depending on how many people have contributed
to the code over the years and who they are, it will be fairly easy or fairly
painful ;-)
Well, we've been looking around for a while for a JNDI server to add to Kaffe
and other GPLd GNU Classpath runtimes, as the Java 1.4 specs demand having one
around, so the mozilla java LDAP code base looks like a prospective base to
start working with. 

Unfortunately, the NPL is not GPL-compatible, so that would prevent us from
picking up the code and reviving it. Could you guesstimate the number of people
that contributed?

cheers,
dalibor topic
I added Gerv to this bug CC earlier.  I think he has some tools to look at all
of the CVS activity and pull out a list of contributors.  Gerv, is that true?

I do know that most of the contributors worked at Netscape (which makes
relicensing easy) but I am not sure if all contributors did.
Hmm. The pulling-out-of-contributors was done by Dawn Endico back in 2001, using
tools she never gave me. But basically, it's a grep of all the code files and
CVS comments for email addresses.

Dalibor: if you'd be happy to put together the list (and document how you did
it), I'm happy to compare the names on it to my master list of people who've
given relicensing permission. We can then see how many people are left. :-)

Gerv
Sure, sounds good to me. I'll grep through the source, ChangeLog and CVS history
for e-mail addresses, and attach a list to this bug report along with whatever
shell script I used.

cheers,
dalibor topic
Don't forget to include the CVS IDs of the people who actually did the
checking-in :-)

Gerv
Any update on this?  We're doing some work on the java sdk - now would be a good
time to get these files relicensed.
From my point of view, I'm waiting on the work outlined in comment #4.

Gerv
Heya,

I've been spending a lot of time lately kicking the latest Kaffe release out of
the door, and doing the regular "brain/lungs/heart" transplantations on the code
base. With that work largely done, I will have time to dig through the commits
in two weeks, after the dust settles after the debian-installer/GNU Classpath
devjam meeting in oldenburg

cheers,
dalibor topic
Has any progress been made on getting the list of contributors?
Here's a quick Perl script I whipped together to gather a list of authors from CVS log comments.  To use it, you simply run it from the top of your source tree (inside the "mozilla" directory for the LDAP JDK) and redirect the output to a file.
Attached file List of CVS authors — 
Here is the list of CVS authors gathered on a fresh checkout of the HEAD of DirectorySDKSourceJava.  I used the cvslog-authors.pl tool that I previously attached to this bug to generate this list.
That's a pretty short list, and we know almost all of them.  Is there some legal boilerplate document we need to get them to sign?
I created this script to scan for e-mail addresses within a source tree.  It will recursively scan through all non-binary files in a source tree (excluding any CVS directories) looking for e-mail addresses.  It considers an e-mail address to consist of alphanumerics, hypen, underscore and decimal characters with either a "@" or a "%" character used to delineate the account portion of the address from the domain portion.

To run the script, you must supply a "-t <path to tree>" option to tell it what source tree to scan.  Redirect the output to a file to get a list of addresses found.
Attached file List of contributors — 
Here is a list of e-mail addresses of contributors found with the previously attached src-contrib.pl script.  The "ace.com" and "aceindustry.com" addresses are sample addresses used in the code, not actual contributors.
My previous attachments list all contributors found from CVS commits as well as source code scans.  Are there any other things we need to check to see if there are other contributors?

From the two lists that I generated, there are only 8 non Netscape or Mozilla addresses, and one of them is Rich.  This is a pretty small list of people that would need to be contacted.  Are any of these people already listed in the master list that Gerv mentioned in comment#4 of this bug?
Any update on getting the list of contributors cross-referenced with the master list of people who've given relicensing permission?
Of the two lists given, only timeless@mac.com has already given permission. Of course, some of the others may have contributed on behalf of an employer, and so may not be able to give permission themselves.

Gerv
I just sent an email to relicensing@mozilla.org granting permission to relicense any code submitted by me.
OK, I've discussed this with Frank. To head towards changing to the full tri-license, here's what we need:

* From independent developers, permission to relicense to MPL/GPL/LGPL (there are two left who have not yet given it)

* For people who were at Netscape, an indication of whether their managers considered this a true spare-time project (e.g., with no resulting Netscape ownership of code) or not. (Maybe one or more of the contributors got an explicit indication on this point from their management, e.g., via email or whatever.)

* For people who were at Netscape, information on who their managers were, in the event that we need to contact them and try to sort out the situation regarding whether this code was work-for-hire or not.

If whoever is driving this would kindly get in touch with the Netscape people and the remaining independents, that would be great. 

Independents, and Netscape people who consider their contributions to have been solely a spare-time thing, should send in a permissions email as outlined here:
http://www.mozilla.org/MPL/relicensing-faq.html#granting-permission

Other pertinent information, such as manager names, should be posted to this bug, either by them or whoever they get in touch with.

Let me know when you're done :-)

Gerv

As the manager of the LDAP dev team in Mountain View before the team and enterprise products moved to RH, the work on this them was indeed "work for hire" as all contributions were performed by this AOL-paid development team.
Frank Keeney has given the detail on the Netscape contributions.

I am unable to contact the two remaining independent contributors (Joseph A. Rank and R.J. Keller).  Their e-mail addresses are no longer valid.  I tracked down another address for what looks like the same R.J. Keller, but I have received no response from that address either.

How should we proceed on this?  We have a strong business need to get the licensing worked out as we don't want licensing issues to push contributors from the community away from this project (as well as other projects that use the Mozilla LDAP JDK).
R.J. Keller has done other stuff for the Mozilla community, and I believe several people know him. You might want to ask more widely among the California-based Mozilla community as to how to reach him.

As for Joseph A. Rank, you may wish to have a look and see if you can quantify his contribution. If you are lucky, it may well be trivial.

If it turns out that he's written 3/4 of the code, contacting Netscape probably isn't worth it. So give me a ping once you've got some idea of where we are on these two issues, and if it still looks plausible, I'll ask Frank to initiate getting in touch with Netscape.

Gerv
The fix that Joseph A. Rank contributed is a trivial fix.  The fix was for bug#62700.  The fix is made up of an addition of 6 lines of code and the removal of 3 lines of code.  The change was basically fixing a broken enumerator that was only printing out the first value of a multi-valued LDAP attribute when writing an LDIF.

If you are interested in seeing the specific code changes, do a diff between revisions 1.2 and 1.3 of mozilla/directory/java-sdk/ldapjdk/netscape/ldap/util/LDIFWriter.java.
OK. Keep trying to track down RJK, and I'll talk to Frank.

Gerv
While researching the contribution from Joseph A. Rank, I came across a few contributors that I missed.  These contributors are:

glenn@somanetworks.com
kdowney@amberarcher.com
Jeff.Gay@gfnews.net

The reason these were missed is that their e-mail addresses were only listed in the CVS log messages.  I ran a cvs log on the entire LDAP JDK tree, redirected the output to a file, then ran the attached src-contribs.pl script to find these 3 contributors.

I will work on contacting these 3 contributors and I will also look into what exactly they contributed.
I looked into the changes contributed by the contributors mentioned in comment#27.  All of these changes appear to be small bug-fixes.  Here are my findings:

Glenn McAllister (glenn@somanetworks.com) - Bug #74378
    +2, -1 according to the attachment to the bug.  Miodrag added some other
    fixes in the same revision, so cvs reports +29, -7, but a lot of those are
    comment lines.
    (cvs diff -r1.7 -r1.8 
    mozilla/directory/java-sdk/ldapjdk/netscape/ldap/LDAPCache.java)

Kyle F. Downey (kdowney@amberarcher.com) - Bug #?
    +17, -3 according to cvs.
    (cvs diff -r1.3 -r1.4 
mozilla/directory/java-sdk/ldapjdk/netscape/ldap/factory/JSSESocketFactory.java)

Jeff Gay (Jeff.Gay@gfnews.net) - Bug #?
    +96, -93 according to cvs, but most of these were changes in spacing.  If 
    you ignore whitespace changes, you will see that the change is really +5, 
    -2.
    (cvs diff -r1.3 -r1.4 
    mozilla/directory/java-sdk/ldapjdk/netscape/ldap/util/LDIFWriter.java)
Glenn McAllister (glenn@somanetworks.com) has sent an e-mail to relicensing@mozilla.org granting permission.
I am unable to contact Kyle F. Downey.

The amberarcher.com domain is no longer valid.  There are a number of Google hits on Kyle's name as well as amberarcher.com, but I am unambe to locate a new e-mail address for Kyle.  Amber Archer appears to have been his own consulting business, but he must no longer be doing that.

I looked into his contributions closer, and it appears to be a pretty small amount of code.  The changes are +17, -3 according to cvs, but most of those lines are comments.  If I ignore the comments, the actual code changes are +7, -2.  The change is basically just adding a new constructor for the JSSESocketFactory class.
You have my permission to relicense the SDK for code that I've contributed. Sorry for the email confusion. Spam has been a killer for me.
Here's a summary of the status of the individual contributors (Y indicates permission has been granted):

Y ingo.schaefer@fh-brandenburg.de
  joer@trapdoor.net (Joseph A. Rank)
Y lukemz@onemodel.org
Y richm@stanfordalumni.org
Y rlk@trfenv.com (R.J. Keller)
Y timeless@mac.com
Y tony@dahbura.com
Y glenn@somanetworks.com
  kdowney@amberarcher.com (Kyle F. Downey)
  Jeff.Gay@gfnews.net (Jeff Gay)

The code contributions from the three contributors that I have been unable to reach have been shown to be small (details on them are in my above comments).

Gerv - Is there anything else you need from me to help move this forward at this time?
Nope, that's great, thanks. I need to talk to Frank about the best way to approach Netscape/AOL. Now that I've just seen your other bug, we may want to get PerLDAP to the same place first, so we only have to make one approach.

Gerv
The three remaining contributed bug-fixes have been re-implemented in a cleanroom fashion.  This work was done in bug 372194 and bug 372440.  The bug fixes were written by Steve Parkinson (sparkins@redhat.com).  Steve has already followed the correct procedure to allow his contributions to be relicensed.

Gerv - Everything is checked in now, so we are ready to have you run your relicensing scripts on the LDAP JDK!
The LDAP JDK has been relicensed on the trunk. Please make sure I haven't broken anything :-)

Gerv
Status: NEW → RESOLVED
Closed: 18 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: