Present: AJ, DH, MK, RL
AJ: 2 hours allocated for Thursday afternoon.
MK: Done nothing on this for a few weeks, going over developer guide stuff.
DH: Sent Marty outline. Writing a smaller, more focussed document for training. Dev guide is huge, 30 screen-fulls for just pvData, not usable as a basis of training. New doc creates 1 structure using the field builder, create one … Quick guide on how to build V4, how to run examples, how to use the pv tools pvget, eget etc. pvaDatabase example plus exampleService, how that works. Write your own service example. 2 hr limit, maybe pvAccess API.
MK: Really wary of “training” wording, don’t use that.
DH: OK. Will borrow material from the devGuide and link to sections there as they’re written. We can do a smaller guide in time for FRIB, can’t get a full dev-guide + training material done in time.
MK: OK, show me what you have in a week or so. Lots of material in the devGuide already, just a few sections that are empty.
DH: Sure, but you covered pvData in 30 screens-full, must be much smaller to present it.
MK: Good approach. By next tuesday?
AJ: What format are you developing it in?
DH: HTML document, present that? Easy to develop.
MK: Sounds good to me.
AJ: I have a talk immediately before the V4 Overview on Base releases and plans.
DH: Closing talk?
RL: Traditionally done by the hosting institution. I also submitted another talk that might fit into the future session on developing hard real-time apps.
AJ: Agenda currently shows 2 other talks from SNS (Matt Pearson) & Archiver (Murali Shankar) mentioning V4.
AJ: EasyCPP API
RL: Subscription options (pvRequest)
MK: When is the next release (Release planning)
RL: We should have made 2 releases since January according to our last agreement.
DH: We lost Matej.
AJ: Discuss merging V4 with Base. The parts of Boost we use in pvCommon is still 10s of MBs of source code and is not needed on all arch’s anyway.
MK: Is Boost a problem on Windows?
AJ: We probably don’t want to tell people “Download and install Boost” as a prerequisite for building Base, it’s huge and not friendly.
DH: What we do next.
AJ: That’s one of Greg’s topics that he wanted to talk to Bob about first.
DH: Working with GW on NT doc. “Order matters”, but implementations don’t match the document for some fields (TS, Alarm). Should we adjust the document to match the implementation?
AJ: Does order really matter?
MK: Is the code self-consistent?
DH: Mostly. NTNDArray follows the document, but others pretty much the same. Python can’t distinguish the order anyway.
MK: Change the code? Anything that relies on the order is probably making a big mistake anyway.
DH: Ok, will change NTNDArray and the document to match, making everything consistent.
AJ: WG Page on V4 website is very out of date for links to the Telecon minutes.
DH: Also has Ralph and Guobao still with old labs.
AJ: Has anybody bought lunch tickets for FRIB yet?
RL: Yes, 5 days lunches plus the dinner ticket.
AJ: We should follow your example, I just haven’t got around to it yet.
AJ: Next telecon not next week, discuss frequency at FRIB.
Meeting closed at 10:35.