EPICS V4 Telecon 07-June-2016

Earlier                                                                                Later

Agenda

  1. Status of Action Items
  2. Release schedule for 4.6.0
  1. Status
  2. Freeze dates, target release date
  3. Telecons (weekly or biweekly)

Minutes

Chair: AJ

Present: AJ, RL, MK, DH, KK

Minutes: RL

1. Status of AIs

No

Action

Assignee

Target Date

Status

2

See who can fund dbGroup/Qsrv going on

GW

assigned

5

Write the one license file (using the MIT license in most cases)

AJ

assigned

6

Update your module to include copy of that one LICENSE. Update your headers to contain the header text.

ALL MODULE OWNERS

wait on 5

7

On your module pages, remove the current/previous links; replace “Working Draft” with version number of module, which has its own numbering scheme, like 5.0.2 for pvDataCPP. Eg “Version 5.0.2 18/Nov/2015”

AJ check status, ALL MODULE OWNERS

5-Mar-2016

assigned

7a

For masarService

GS

5-Mar-2016

assigned

7b

For pvAccessJava

MS

5-Mar-2016

assigned

7c

For pvCommonCPP

MS

5-Mar-2016

assigned

10

Simplify Literature page and home page to help users find novice oriented documentation.

GW

assigned

16

Commit changes to all modules, other than exampleCPP and pvDatabaseCPP, which set the shared library versions

MD

assigned

17

Monitor changes pull request (AI from meeting 15-Mar-2016, http://tinyurl.com/zqa3qyp

MD

5-Apr-2016

assigned

2. Release schedule for 4.6.0

a) Status

MK:

For the last several weeks I have been working on the following (pvaClientCPP and pvDatabaseCPP):

1) looking for memory leaks when a process terminates.

2) As far as possible support RAII (Resource Acquisition Is Initialization)

3) Add more documentation about pvaClient and pvDatabase to developerGuide.html

I have pushed the latest to pvDatabaseWWW.

4) For pvDatabaseCPP support on-line delete of PVRecords.

The current status of pvaClient is:

1) It now fully supports RAII. All destroy methods are now deprecated and do nothing.

2) The following problems were found in pvAccessCPP. Matej has been notified.

   a) caProvider has a problem with monitors.

   b) When ClientFactory::stop() is called there are still memory leaks.

3) exampleCPP/exampleClient has some new tests that help identify memory leak problems.

4) The changes for pvaClientCPP and exampleCPP are only in mrkraimer gitHub

I do not want to issue a  merge request for epics-base until 2) is resolved.

The current status of pvDatabaseCPP is:

1) pvDatabaseCPP itself has no memory leaks.

2) ServerContext has memory leaks when destroy is called.

3) It is not possible it make this fully RAII.

The reason is that destroy can be called by both the remote side of pva and for on-line record delete.

4) pvDatabaseCPP does support on line record delete BUT there is currently no way to

make current clients stop accessing the record.

This was just discovered last week. I notified Matej but he has not had time to think about it.

5) I do not want to issue a  merge request for epics-base until 2) is resolved.

I think that except for on line record delete the other problems should be fixed before release 4.6

RL: Please make the issues issue tickets on GitHub.

AJ:

Changes are ready to move (at least most of) the boost stuff out of pvCommonCPP. What to do with the Micro-Benchmarking code? MD has a branch in Base that provides some of the API the MicroBench stuff needs.

b) Freeze dates, target release date

AJ: Deferred. Probably stretched out because of vacation, anyway.

Aiming for a feature freeze date in July?

DH: Would be nice to have a new version for the Tennessee meeting in September (19..23-Sep-2016).

AJ: Sounds like 4.6 will be time driven more than feature driven.

RL: Should still discuss the list of wanted features before feature freeze, at the next telecon (21-Jun-2016).

AI on AJ: Circulate a straw-man timetable and request to collect a list of features for 4.6.0 before the SNS EPICS meeting.

c) Telecons (weekly or biweekly)

No telecon next week (14-Jun-2016).