EPICS V4 Telecon 13-5-14.


0. Preliminaries
1. Next release, 4.4
  a. I have not received any feature lists. Please forward a list of the features, plus a
     intend to release in 4.4. Please do so in the form of a list, with a short (one sentence or
     short paragraph) explanation of each feature.
  b. Documentation requirements for 4.4.
2. Brief status reports


Present: DH, MS, GW, GS, TK, RL, MK, BD

Scribe: GW

Chair: GW


New Subject: The blocking and non-blocking API of the RPC method of pvAccess:

GW: Yes the rpcClient changes for non-blocking RPC should be in the next release

MK: What is now in the Java pvAccess RPC is what the general RPC API should be. The difference is whether one wants to block or not - say if you want to issue both of them at the same time.

GW: What if one client issues 2 request in parallel

MS: It depends on which methods of the RPC support the server implements. That is, it can use the non-blocking API. There are 2 alternatives in the (java) API, blocking and non-blocking. However, *The C++ wrappers of RPC do not have the non-blocking API yet implemented.* Only the Java wrappers so far do.


AI on MS; 13-6-14: Add non-blocking (parallel) RPC API to C++ pvAccess, implemented on the main branch.

GS: gs doesn’t see an elegant shutdown method to stop an rpc service.

[deferred to mail list]

NEW SUBJECT: Next release, 4.4

AI on Greg by 21-5-14: Make a google doc shared for tracking features of the next release.

RESOLUTION: The next release source code will have complete Doxygen and Javadoc documentation. Specifically, all public and protected method described (using words not in the method name itself), and class and package documents.

DH: What is the intention of the release? Particularly with respect to the EPICS meeting?

NEW SUBJECT: Status Reports

MS: Java side of “changes” for pvAccess API are nearly complete. Nearly ready for merging with main branch.

MK: pvIOCJava and swtShell also (these still necessary for testing)

MS: Now working on C++ code for the API changes.

DH: working on NTNDArray. Porting from areaDetector, and making work with AD 2.0. Now integrating the epics TimeStamp.

GW: is NTNDArray stable?

DH: Stable right now but still isn’t yet in an implementation being used. We won’t be seeing this new processor pipeline in real use until the end of the year.

GW: I’ve been working on 3 fronts: EPICS V4, how to encode errors with values, and things we need to address generally. 2nd a services framework. The services framework is now ready. Murali is implementing an archive service and a channel and device service in it.

3rd a new online model system for LCLS and LCLS-II. Overall excellent progress.