Use of JCP site is subject to the JCP Terms of Use and the Oracle Privacy Policy

JCP.next master list: March 3 2011

JCP.next Master List of proposed changes

Updated March 3 2011

TODO for each item

Prioritize.
Process Doc or JSPA change.
Decide whether in JSR1 or JSR2.
Do we have consensus or need more discussion?
Do we already have draft text?
Find volunteer to flesh out the idea

Documents to be modified

The JSPA
- Normative
The Process Document
- Normative
The EC Members' Guide
- Non-normative: referenced from Process Document
The Spec Lead Guide
- Non-normative: referenced from Process Document

Procedural issues

When will the changes in the JSPA take effect?
- We can not require that an existing JSR be ruled by new version of the JSPA.
- Currently there is no requirement that members upgrade to the latest JSPA.
  - Note that section 2B uses the term "expires" but it seems there is no way for a JSPA to expire (merely "terminate") - we should fix this,
- The JSPA also forbids us from applying changes to JSRs that are in-progress:
- Will the new version be version 3 (containing significant legal changes?)
  - Presumably yes - otherwise why make the effort to modify it?
  - If so, we want people to upgrade - all members should be operating under similar IP grants.
- Specify a lengthy transition period (two years?) with a common deadline, allowing ample time for review.

New JSRs must adopt the latest JSPA. This implies that the Spec Lead and EG members must sign it.

When will the changes in the Process Document take effect?
- The JSPA currently prohibits us from applying Process Document changes to JSRs that are in progress:
  - "The Process is described on the JCP Web Site (at http://jcp.org), and may be revised from time to time in accordance with terms set forth in the Process document, provided that no such revisions shall apply to any JSR that has already been approved for development."
- Should we modify this language to give us the power to apply to in-flight JSRs?
  - This wouldn't require us to do so, just make it possible if we wish.
  - If so, the language should specify that at the EC's discretion, the requirements of the Process Document will apply to all JSRs when they next make a formal state-change through the process.

Transparency

Expert Group transparency

JSPA changes (JSR2)
- Eliminate confidentiality language from the JSPA.
  - We have draft language (from the JSR 306 draft)
Process Document changes (JSR1) (draft language required)
- All significant business must be carried out on public mailing lists.
  - EG-private mailing list may be maintained for administrative matters.
- Issues must be tracked through a publicly viewable issue tracking mechanism.
- Expert Groups must respond publicly to all comments before JSRs can move to the next stage.
  - We have language requiring comments to be published, but not requiring responses - this must be tightened up.

Executive Committee transparency

Process Document changes (JSR1) (draft language required)
Add semi-annual EC meetings that all JCP members are free to attend and where the agenda chosen from topics suggested by the membership.
- Meeting schedules are currently not specified in the Process Document. Should they be, or does this belong in the EG Members Guide?
Create a public alias (with archive) for members to provide feedback to the ECs.
Need some language to address Undocumented Practices.
- For example, PMO/Oracle Legal review of license terms.

TCK transparency (needs more discussion)

Permit and encourage publication of TCK test results so users can see who's compatible and who's not.
- Oracle TCK licenses require confidentiality - JSPA change could explicitly mandate openness.
  - TODO: Oracle should own this - draft some language
TCK documentation (including Compatibility Requirements) must be publicly available.
Should we publish the tests too?
- TODO: Oracle to review this

Individual membership

Fix potential issues with JSPA Exhibit B.
- Needs legal review
- TODO: Oracle to start this
- Others should feel free to offer suggestions.
Individual members' votes - how to avoid ballot-stuffing?
- Process Doc contains this language: "All JCP Members are eligible to vote in a ratification ballot subject to the provision that if a Member has majority-ownership of one or more other Members, then that group of Members will collectively have 1 vote." If we change this to “has majority-ownership of or is the employer of one or more other Members” then we should be covered.
  - It might be difficult for the PMO to enforce this in practice, but we would have the authority to do so if necessary.
Modify cost section of JSPA to permit PMO to charge a nominal fee for individuals (if this should prove necessary)
- See Wayne Carr document for justification.
- Simple change. Since JSPA, target for JSR2.
- Fees are not mentioned in either the Process Doc or the JSPA! Should they be?

IP

Non-assertion patent covenant.
- We have draft language. Patrick to circulate this and ask people to get their lawyers to comment
IP grants flow directly to implementers, not via Spec Lead
- Justification? TODO: Patrick to circulate this.

Licensing

Require that when the license terms associated with a JSR are significantly changed from the previous release, this is explicitly called out.
- Require change log for licensing changes.
- Process doc change, JSR1, non-controversial.
- Suggest changing
  - definition - Change Log: An area accessible from the Spec Page that lists all changes made to the Specification after Final Release. There are three sections: PROPOSED (changes not yet made to the Specification), ACCEPTED (changes made), and DEFERRED (change items to be considered in a new JSR).
- To:

Note also that strictly speaking the change log does not list all changes "after Final Release" but only those since the previous release. We should therefore change "after Final Release" to "since the previous release."

Clarify the requirement to disclose licenses, particularly if different licenses are available for various types of implementation or different fields of use.

Can we require every variant to be disclosed? No.

Require that offered licenses must be available to anyone who is willing to sign.
TODO: John Rizzo to check whether existing language guarantees this, and to suggest Process Doc changes if not.
May need to strengthen the language in the JSPA instead?
- After discussion with John Patrick concluded that if we intend to make changes we should do so in the JSPA rather than the Process Doc (that's where the obligation is defined.)
- Patrick suggests changing this:
  - The Spec Lead (including Sun) for a JSR approved by the JCP shall offer to any interested party licenses concerning the RI and TCK -- and also the TCK separately when developed under any JSR submitted hereafter -- on terms and conditions that are non-discriminatory, fair and reasonable.
- To this:
  - The Spec Lead (including Oracle) for a JSR approved by the JCP shall offer to any interested party licenses concerning the RI and TCK -- and also the TCK separately when developed under any JSR submitted hereafter -- on terms and conditions that are non-discriminatory, fair and reasonable. RI and TCK licenses should permit both commercial and non-commercial implementations and should be granted to all who agree to their conditions.

Define a mandatory standard Spec License.

TODO: Patrick to circulate the "recommended" Spec License for comments.

Recommend (but do not require) standard RI and TCK licenses.

Provide separate templates for Independent (open source) and commercial implementors.
TODO: Patrick: circulate examples for discussion

Agility

Time-outs for inactive JSRs (JSR1) (we have draft language)

See JSR Renewal Ballot suggestion from JSR 306
We have language
JSR1

Enable implementations before specs are finalized, to gain real-world experience.
- Must maintain compatibility.

See Transplant JSR proposals from JSR 306
Look for feedback from WG members.
- What mechanism to use to gather feedback?
- If something is critical for one or two companies but not high priority for others - may not be "on the list"

Better procedures to remove or replace non-responsive spec leads and to deal with bankrupt or defunct spec-lead organizations.
- Have no language.
- Process doc only, or should we modify JSPA to permit withdrawal of contributions.
- TODO: Michael to review and suggest changes.

Participation

Make the JSPA less intimidating by refactoring into three layers
- Simple Terms Of Use for non-members who wish to comment on specs in public forums.
  - http://jcp.org/en/home/terms
- A simple membership agreement for those who want voting privileges and the right to serve on Expert Groups.
  - Would the IEPA be a useful starting point?
    - No - it still includes all of the Spec Lead obligations - it's no better a starting-point than the JSPA itself.
- The complexities of IP and compatibility obligations are required only for Spec Leads – factor these out into a separate agreement.
Clarify Terms of Use for collaboration software used by Expert Groups
- Should ensure sufficient IP grants, but not impose requirements more strict than the JSPA

See Wayne Carr document for explanation.

How to police?
Spec lead tells us in submission what collab software will be used
Review Terms of Use (who?)
- Oracle legal to report - EC members can review the report?
If OK - add to approved list. If not, look elsewhere
Deal with possible changes to terms of use over time?
- May need to review each time, even if previously reviewed.
Incorporate this into JSR Submission process
TODO: Patrick to talk to Oracle Legal about this.
JSR2? (more effort - legal work required)

Is this a Process Doc or a JSPA change?
Do in two stages?

Non-controversial, but we have no draft language yet

No barriers to participation in Expert Groups.
- Requests to join EGs and the Spec Lead's responses must be made in public.
- Discuss with EC again.
Define a process for penalizing EG members who do not participate.
- Lose voting privileges. Look at OASIS, W3C.
Section 2.1.3: Uncooperative or unresponsive Expert Group Members seems incomplete. It starts by talking about how to deal with EG members but the proposed remedies then address the Spec Lead role. Clarify.
Clarify requirements for RI and TCK to be posted within x weeks of Final Release (and Maintenance Release?)
- How to "undo" if materials not posted?
  - PMO to draft something
Penalties for EC members who do not attend meetings or do not vote.
- First step: lose voting privileges (which would solve the quorum problem.)
  - Would need to change the definition of quorum to be "of those who are eligible to vote."
  - Suggestion from Eduardo: if you miss two meetings in a row you lose voting privileges. Can restore by attending two meetings in a row.
- Second step?
  - Lose membership if...
    - Miss x meetings in a year...
    - Losing voting privileges more than x times per year...
- If people routinely miss meetings, members who attend only occasionally are "disruptive"

Cleanup

Begin

Change Process, JSPA and IEPA to wind down use of IEPA. Remove IEPA
reference in JSPA. Change IEPA to say it will no longer be offered. Leave the ability to modify the IEPA in the Process document until there
are no IEPA members, but note it is no longer available for new members of any EG. When the
last JSR EG disbands that has an IEPA member in it, revisit this and remove IEPA references.
(IEPA, JSPA – much later when there are no IEPA members, a Process doc change finishes the
clean up)

Begin phase-out of IEPA
- IEPA is no longer used, but it had no termination provision so those under it remain in JCP until the JSR they joined disbands.
Remove Process Doc section 1.1.4 J2ME building blocks
Correct typos and errors in JSPA.
Correct typos and errors in Process Document.

Remove the ambiguity of the term Spec Lead in the Process Document (sometimes it refers to the Member and sometimes to the person representing the Member and leading the EG.)

See Wayne Carr document for explanation

Prohibit major Patform changes in Maintenance releases (see MR2 of JSR 176)

Patrick to follow up

Review Appendix A of Process Document (Eduardo has some concerns?)

Eduardo to review and report back (also on "lack of clarity when things are supposed to happen")

Branding and compatibility

Emphasize the value of compatibility & the brand and the risks of using incompatible implementations.
Relax all-or-nothing compatibility reqiurements?
- TODO: Patrick to give the the compatibility presentation to EC...
Publish lists of compatible implementations and report what optional features are implemented.
Publish lists of incompatible implementations.

No JSPA or Process Doc changes required for these proposals – Oracle/PMO commitment is sufficient?

More formal reporting requirements for conformance claims?
Where would the language be written? In TCK licenses?
- Too hidden?
- How to make this more public?
What would the remedy be if someone makes a false claim?
Oracle to drive this...

Miscellaneous

Clarification of the JSPA as discussed within the EC.
- Members will be asked to sign the new/revised JSPA (resulting from JSR2) - some may insist on this clarification at that point.
- Modified JSPA should not grant rights to one player that others don't have.
Enable and encourage the establishment of more formal relationships with other standards bodies.
- Some disagreement as to whether Confidentiality is an obstacle to liaison with other organizations. Need more discussion.
Transplant JSRs (enable incorporation and standardization of work developed outside the JCP.
- Need some use-cases?
- Early availability was an issue. Permitting implementations before specs go final. IP-flow issues?
Hybrid JSRs (allowing non-Java implementations of a JSR's specification)
Discuss these two next week - notify EC members in advance we'll be doing so - invite those who are interested to participate - review the actual docs/drafts

Additional suggestions

Address the issue of where litigation should be located. California, or elsewhere?
- California is an obstacle to Brazilian state involvement.
- According to Java Champion Douglas Jenssen this is also an issue within the US. He has said in private mail: "The issue is a legal one, where do legal disputes between Sun/Oracle and a member -- whether individual, sponsored by his employer, or corporation -- get litigated. When I last met with the JCP, the answer was it has to be in California. No public institution such as a state university is going to agree to that, and many corporations will not agree either. Hence I was told there are no members from public universities, only from private ones that choose to agree."
- The JSPA says: "Any action related to this Agreement will be governed by California law, excluding choice of law rules, provided, however that neither party has consented to the jurisdiction of any court located in the other party´s country of incorporation."
Incorporate relevant portions of the EC Members' Guide (eg the sepecification of quorum) into the Process Document.
- Eduardo is reviewing
Increase fees?
Deal with the Maintenance Expert Group problem
- Spec Leads want the EG to persist after Final Release. The Process Doc states that the EG must disband.
  - Solution is to replace the Expert Group with a Maintenance Expert Group?
    - Initial membership of the MEG would be recruited (automatically?) from the EG?
- We need to maintain the history of which members were on the EG when it completed its work, but it should also be possible for EG to evolve.
  - We really should "snapshot" the membership at every Release.
Also clean up Maintenance Release requirements. Currently there's no absolute requirement for the Maintenance Lead to actually deliver an updated spec, RI, and TCK. (The process simply says "the maintenance changes will be considered final when the RI and TCK are synchronized with the Specification. We should require a formal Maintenance Release just as we do a Final Release.

TODO

Legal collaboration between Oracle and others' lawyers.
Decide on location for f2f (Redwood Shores or Santa Clara?)
- Assume 9-5 on Tusday, 9-4 on Wednesday
Do the compatibility presentation at the March f2f

Transplant

Do this for agility purposes alone?
Necessary for "co-development" work outside of as well as inside the JCP
- Eg, cloud technologies
Do not permit indefinite shipment - insist that existing products be made compatible within x months?
On the list

Hybrid

Under what conditions would we permit this to be applied to an existing JSR?
Makes JCP more attractive - encourages interoperability with other platforms
On the list.