version 0.1: December 11, 2010
November 9, 2010
Total attendance: 11
Total attendance: 15
|Since 75% of the ME EC was present, that EC was quorate for
Since 75% of the SE/EE EC was present, that EC was quorate for this meeting
The EC statistics were inserted into the record.
There were no personnel changes to report.
Patrick reported the results of the annual EC elections. See the PMO presentation for details. He explained that abstentions are not counted for the purpose of determining whether ratified candidates are elected (they must simply receive more "yes" than "no" votes to be confirmed.)
During discussion of the failure to ratify Hologic John Rizzo suggested a company that was more invested in and committed to Java would have been preferable. Patrick responded that Oracle was looking for such a candidate as a replacement for Hologic. Josh Bloch pointed out that since the seat that needed to be filled used to belong to Doug Lea, a similar individual member would be preferable.
Patrick reported that members could expect to see a significant number of new JSRs being filed during the next few months, with major revisions for all three platforms. (See the PMO presentation for details.) Werner Keil asked whether progress was being made in the Modularity Working Group that was set up after the October meeting, and whether a JSR could be expected from that group soon. Josh Bloch reported that he was no longer leading that group, but that it was meeting, and did expect to file a JSR relatively soon. Werner also reported that the Trusted Computing JSR (#321) would soon be ready to move to the next stage.
Sean Sheedy pointed out that EC members would need to see licensing terms for the upcoming Java ME JSRs, and asked for the opportunity to review and comment on them before they are filed. Calinel Pasteanu promised to socialize the JSRs before filing, and noted that licenses would be disclosed as required by the Process Document.
Patrick discussed the plans for the JCP.next JSRs, suggesting that the simplest approach would be to withdraw JSR 306 (the previous JSR that addressed Process and JSPA changes) and to file two new JSRs to be worked simultaneously: one for simple, non-controversial changes that could be implemented quickly and the second for more complex changes. He asked for volunteers to form a Working Group to focus on refining the proposals and crafting the language for changes to the Process Document and the JSPA. John Rizzo, Steve Wolfe, Mike DeNicola, Michael Bechauf, and Scott Jameson volunteered. (Others volunteered during a later email discussion. The full list of members will be published soon on jcp.org.) It was agreed that the Working Group would operate transparently, and would report back regularly to the full EC. Patrick agreed to call the first meeting of the Working Group soon.
Since one of the primary goals of the JCP.next proposals is to further strengthen the transparency requirements of the process, Patrick presented a summary of what had been done so far to promote transparency. He noted that the "community" pages associated with each JSR are not as widely used as they should be. He also noted that they are accessed, somewhat confusingly, via the tab labeled "Update." It was suggested that this be changed to "Community."
Tim Peierls asked why it is necessary to login in order to view the Update page. Patrick responded that this was part of the effort to ensure that there were sufficient "privileges" for members that there was an incentive for them to join the organization. He agreed that this seemed to run counter to the notion of transparency, and should be reviewed.
Patrick suggested that the primary decision the JCP.next Working Group would have to make in this area would be to decide which of the items on the transparency checklist should be made mandatory. Tim Peierls said that some kind of enforcement mechanism would be necessary, and suggested that EC members should consider Expert Groups' transparency record when voting on JSRs.
Patrick clarified the wording on the second "Branding and Compatibility" slide in the JCP.next presentation. He explained that Sun/Oracle's TCK licenses contained "secrecy" requirements in an attempt to guard against implementers making "relative" or "comparative" compatibility claims such as "my product is 98% compatible" or "my product is more compatible than yours." Josh Bloch argued that compatibility is not really a binary quality. Patrick agreed that such comparative compatibility claims, together with the force of the market, had worked well in the W3C world, with implementations of complex specs such as XML Schema gradually improving in quality and compatibility.
Patrick asked for additional suggestions for the JCP.next list of proposed changes. Scott Jameson suggested adding to the list a clarification of the JSPA's language around licensing requirements and restrictions.
Before discussing the meeting schedule Patrick raised concerns about the difficulties of holding "mixed mode" meetings (where the majority of participants are physically present while some are on the phone) if the phone infrastructure is inadequate or the acoustics in the meeting room are bad. (There had been some difficulties in this area at the recent Bonn face-to-face meeting.) No simple solutions were suggested, but it was agreed that members should consider these factors before volunteering to host a face-to-face meeting.
Patrick then suggested that the ECs should consider meeting in Asia every other year, in order to make it easier for the South Korean members to fully participate.
He asked whether the EC should meet face to face four rather than three times a year, but members rejected this suggestion given the logistical difficulties discussed earlier. Scott Jameson suggested that an additional "meet the community" event at JavaOne could constitute the fourth meeting.
Patrick asked whether members were planning to attend the JavaOne events in Brazil or China. Few were.
Patrick agreed to conduct an online poll to find the most convenient time for the January face-to-face meeting. (This was done - the dates chosen were January 11 and 12.)